OXFORD STUDIES IN THEOLOGICAL ETHICS

General Editor
OLIVER O’DONOVAN
 OXFORD STUDIES IN THEOLOGICAL ETHICS

General Editor
OLIVER O’DONOVAN

The series presents discussions on topics of general concern to Christian

Ethics, as it is currently taught in universities and colleges, at the level
demanded by a serious student. The volumes will not be specialized
monographs nor general introductions or surveys. They aim to make a
contribution worthy of notice in its own right but also focused in such a
way as to provide a suitable starting-point for orientation.
The titles include studies in important contributors to the Christian
Tradition of moral thought; explorations of current moral and social
questions; and discussions of central concepts in Christian moral and
political thought. Authors treat their topics in a way that will show the
relevance of the Christian tradition, but with openness to neighbouring
traditions of thought which have entered into dialogue with it.
God’s Command
JOHN E. HARE

1
 3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© John E. Hare 2015
The moral rights of the author have been asserted
First Edition published in 2015
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2015938773
ISBN 978–0–19–960201–8
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 Preface

This book is a defence of the claim that what makes something obligatory is
that God commands it, and what makes something wrong is that God forbids
it. I will call this claim ‘divine command theory’. I describe at the beginning of
the first chapter how the book lays out the argument for the claim. This
preface is a reflection on what kind of book this is, and then an acknowledge-
ment of my debt of gratitude to many people who have helped me write it.
The first thing to say is that the book proceeds almost entirely by way of
conversation. It works out its ideas by reflecting on how they differ from other
people’s ideas, some of them contemporaries and some of them from the
tradition. Sometimes philosophers get impatient with this procedure, espe-
cially with drawn-out engagement with historical figures in the tradition. They
want to get on with the theory. But the danger of starting from scratch or from
the very recent past is that while in fact using ideas and concepts with a
history, one does not know where they come from, and so one does not know
their context in other ideas and concepts that may well be disagreeable.
Nobody can know the whole history. But our discussion now would be
much better focused if we had collectively a better sense of the genealogy of
the ideas we use. I have, however, given a version of the theory right at the end
without historical reference, and without reference to contemporary disagree-
ments. Readers who feel that the main text is losing the forest for the trees
should consult the summary in the final chapter.
What discipline does the book belong to? In particular, is it philosophy or is
it theology? This turns out not to be a straightforward question, because the
terms themselves are in dispute. The book is focused on the Abrahamic faiths,
and I have not tried to write about the religions of the East. I have also not
tried to write about all the Abrahamic faiths. I have used the term ‘the
Abrahamic faiths’ as though it referred just to Judaism, Christianity, and
Islam, but the term should rightly include the Baha’is, the Druze, and many
others. The medieval philosophers from Judaism, Christianity, and Islam who
are covered in this book all did both philosophy and theology. They had
different accounts of the relation between revelation and reason, which is
the issue that lies behind the question about current disciplinary labels. One
way to put the question is to ask what is the relation of this book to the
authority of Scripture. I myself am a Christian, and I have the conviction that
God has spoken decisively in the Scriptures that Christians call ‘the Old and
New Testaments’. But this book does not use the form of argument: ‘This is
the teaching of the Holy Book and therefore it is true.’ Rather, it uses the form:
‘This is the teaching of the Holy Book, on at least one interpretation of it, and
vi Preface

the following philosophical discussion shows how it may make sense within
the framework of Christianity.’ The same is true with relation to the teaching
of Torah for Judaism and Qu’ran for Islam. This makes the book a work of
philosophy rather than theology on one account of what the difference is
between these disciplines.
The answer that the book is a piece of philosophy fits also with its aspiration
to be the next in a recent series of books that have defended forms of divine
command theory within analytic philosophy. The series I have in mind started
with Philip Quinn’s Divine Commands and Moral Requirements (Oxford,
1978), and proceeded, among other titles, through Robert M. Adams’s Finite
and Infinite Goods (Oxford, 1999) and C. Stephen Evans’s God and Moral
Obligation (Oxford, 2013). No one looking at the discipline in the late 1960s,
when I was doing my training, would have predicted the existence of this
series. These titles, and others like them, are a manifestation within ethical
theory of a much larger shift within the discipline, which has become more
pluralistic in its sense of what is allowed as a sensible topic of conversation.
On the other hand, this book strays beyond most works of philosophy. It
engages with a number of Christian theologians, most especially Karl Barth,
and it extends into a discussion of divine command within Judaism and Islam.
It also attempts a sustained discussion with contemporary evolutionary psych-
ology. I am conscious that the scope of the enterprise is larger than any one
person can responsibly undertake. I have had a great deal of help from people
who know much more about the individual parts of the programme than I do.
I will come to thank those people in a moment, but, first, I want to say
something about my intended audience.
The first audience is those who are already engaged in thinking about
theological ethics. There is already a long history of reflection about how
divine command and natural law fit or do not fit together. For the most part
this reflection has been internal to one religious tradition, and one contribu-
tion of this book is to widen the conversation to three. One of the theses of
this book is that the Abrahamic faiths all contain an internal tension between
divine command theory and what we might call ‘natural law theory’, though
there are many different versions of this latter and I am not going to try to
give a set of necessary and sufficient conditions for what counts as natural
law. The project is to examine similarities and differences between the
conversations about this tension in each of the Abrahamic faiths. Because
I have emphasized the commonalities, however, I have given less attention to
the distinctive features of Christianity, especially doctrines of Trinity and
atonement. My earlier book, The Moral Gap, focused theologically on atone-
ment and justification.
The idea that God’s command might lie at the foundation of ethics is an
idea that will still strike many of my colleagues within the discipline of
philosophy as simply incredible, despite my earlier comments about the shift
 Preface vii
in the discipline. I hope that some of them will, nonetheless, try the experi-
ment of reading the book. There are, after all, many people in the world who
do not find this idea incredible. On some estimates, 80 per cent of the world
will belong to one of the major religions by 2050 on current trends.1 This
figure needs to be taken with a grain of salt, but the point remains. If we want
to have a foundation for ethics that makes sense to most people in the world,
we need to take divine command seriously, even if we do not share the
sensibility that makes this a live option for ourselves. This is not to say that
divine command theory is the only way to take divine command seriously. But
it is one way, and it promises to give an answer to a perennial philosophical
problem: Why should we be moral? Divine command theory says that the
answer to this is that we should be moral because God commands it. Far too
often, the dismissals of divine command theory by philosophers have been
ignorant, relying on caricature. Surely it is worth looking at a philosophically
serious attempt to defend the view. Another inducement for philosophers to
read the book is the engagement in the final chapter with contemporary
evolutionary psychology.
On the other hand, religious people who adhere to one of these three faiths
and organize their lives around them tend to be impatient with philosophers.
Part of the problem is the minute distinctions and agonizingly complex trains
of argument that are quite properly endemic to the field. But more than this,
I think the ground of the impatience is the sense many ordinary religious
people have that professional philosophers simply do not take them seriously.
I hope that some of these people within the faiths who have some background
in philosophy will try the experiment of reading this book to see what a friend
of the faith who is a philosopher might be able to contribute to understanding
their faith better. The book aspires to be helpful with the question of how to
live with two beliefs that seem to be in tension with each other. One belief is
that God is sovereign and God’s will and command settle how we ought to live.
The other belief is that we often have to work out ourselves what to do.
Understanding how these beliefs fit together requires a good deal of hard
intellectual work. I should warn that I have not forbidden myself technical
vocabulary, unlike my practice in Why Bother Being Good?, which contained
no word ending in ‘ism’. This is because the technical vocabulary is often
economical, with one word saving a paragraph of explanation. I have, how-
ever, explained each technical term the first time I use it.
The first seed for the book was a series of Gifford Lectures in Glasgow that
I gave in 2005 together with Lenn Goodman, Abdulaziz Sachedina, and
A. C. Grayling. The idea for the series was that we would discuss the biblical
commandment ‘Thou shalt love thy neighbour as thyself ’ from a Jewish,

1
John Micklethwait and Adrian Wooldridge, God is Back (London: Penguin Books, 2009),
p. vii.
viii Preface

Christian, Muslim, and Secular Humanist point of view. We found, not

surprisingly, that, while we agreed on the general principle, there was signifi-
cant disagreement when it came to the detailed working-out of the theory. But
what was surprising to me was that many of the disputes were not so much
between the religions (and the non-religion) as internal to each of the four
traditions represented. This thought provided the germ for the present book.
In 2011 I gave an earlier version of the first four chapters of this book as
the Wilde Lectures in Natural and Comparative Religion at Oxford. The
discussions both after the lectures and privately were most fruitful, and
I remember especially conversations with Richard Swinburne, Terry Irwin,
and Tony Kenny.
I have given versions of portions of the book as lectures in various places.
The danger of this kind of list is that one omits all sorts of people one should
not omit. But I am especially grateful for discussion at Purdue with Michael
Bergmann, Patrick Kain (who has been very helpful on Kant), Jeff Brower, and
Kyla Ebels-Duggan, and in Auckland with Rosalind Hursthouse and Mark
Murphy.
Jennifer Herdt and I taught a class together at Yale Divinity School that was
based on her Warfield Lectures at Princeton Theological Seminary and my
Wilde Lectures. The members of that class, and Jennifer particularly, gave me
continual material for reflection.
Perhaps the most fruitful sessions of all for the final shape of this book have
been with a group of faculty and doctoral students in philosophy and religious
studies who have given generously of their time over a whole semester to
discuss the content of each chapter. The members of this group were Jamie
Dunn, Brad East, Janna Gonwa, Ross McCullough, Evan Morse, Stephen
Ogden, Philip Stambovsky, and Graedon Zorzi (with some other occasional
visitors). It has been a privilege to have their input, and every chapter has
changed significantly as a result of the conversation.
I would like to mention also (in alphabetical order) Neil Arner, Joel Baden,
Brian Besong, Nigel Biggar, Paul Bloom, Andrew Chignell, Sarah Coakley,
Michael Cohen, Anthony DeBonis, Omar Farahat, Paul Franks, Greg Ganssle,
Frank Griffel, Andrew Hare, Caden Hare, Christine Hayes, John Heinze, David
Kelsey, Philip Kitcher, Gerald McKenny, Geoffrey Mosely, Joan Lockwood
O’Donovan, Oliver O’Donovan, Humeyra Karagozoglu Ozturan, John Pittard,
Jean Porter, Ulrich Rudolph, Michael Ruse, Fred Simmons, and Miroslav Volf.
I have had very significant help from three research assistants, Jack Sanchez,
Erik Santoro, and Matthew Vermaire. The book is a great deal better than it
would have been if I had foolishly tried to do the whole thing on my own.
Finally, my wife, Terry, has given unflagging support, even at the end stages
when the book required an absurd number of hours in front of the computer.
My gratitude to her goes beyond words.
 Contents

1. Morality and Religion 1

Introduction 1
1.1. The Argument from Providence 5
1.2. The Argument from Grace 11
1.3. The Argument from Justification 16
1.4. God’s Command and the Scope of Obligation 25
2. What is a Divine Command? 32
Introduction 32
2.1. Five Types of Prescription 33
2.1.1. Prescription in General 33
2.1.2. Precepts 37
2.1.3. Prohibitions 41
2.1.4. Permissions 42
2.1.5. Counsels 44
2.1.6. Directly Effective Commands 48
2.2. Divine Authority 49
2.3. Barth on Divine Command 55
2.3.1. Six Implications of Our Being Commanded by God 55
2.3.2. Three Puzzles 57
3. Eudaemonism 63
Introduction 63
3.1. Does Morality Make You Happy? 63
3.2. The Sources of Motivation 67
3.2.1. A Single-Source View: Aristotle 68
3.2.2. A Double-Source View: Scotus 72
3.2.3. Two Errors of Kant 75
3.3. Four Attempted Defences of Eudaemonism 77
3.3.1. The First Defence: Epicurean 78
3.3.2. The Second Defence: Stoic 83
3.3.3. The Third Defence: Thomist 86
3.3.4. The Fourth Defence: Agent-Transcendent
Eudaemonism 94
4. Can We Deduce Morality from Human Nature? 99
Introduction 99
4.1. Scotus 100
x Contents

4.1.1. The Non-Deducibility of the Law from our Nature 100

4.1.2. The Fittingness of the Law to our Nature 102
4.1.3. The Social Character of Obligation 109
4.2. Consensus Deductivism 112
4.3. Prescriptivism 117
4.3.1. Motivation 118
4.3.2. Moral Properties 121
4.3.3. Ideals 124
4.4. Foot and Hursthouse on Deductivism 128
4.4.1. Too Much and Too Little 128
4.4.2. Good Roots and Good Wolves 130
4.4.3. The Good Promise-Keeper 135
5. Barth on Divine Command 142
Introduction 142
5.1. Barth on Particularity 143
5.1.1. The Specific Individual 143
5.1.2. Haecceity 145
5.1.3. The Positions in a Moral Judgement 147
5.1.4. Barth on Universality 151
5.2. Barth’s Account of Human Freedom 157
5.2.1. Three Pictures of Freedom 157
5.2.2. The Canaanite Woman 164
5.3. Barth and our Access to the Commands 166
5.3.1. Barth and Kant 166
5.3.2. Kant on Conscience 169
5.3.3. The Tradition’s Resources for Assessing the Command 173
6. Divine Command in Some Medieval Islamic Thinkers 184
Introduction 184
6.1. Intrinsic Value 187
6.1.1. ‘Abd al-Jabbar 188
6.1.2. Al-Ash‘ari 195
6.1.3. Al-Maturidi 197
6.2. Human Freedom 203
6.2.1. ‘Abd al-Jabbar 204
6.2.2. Al-Ash‘ari 206
6.2.3. Al-Maturidi 209
6.3. Revelation and Reason 212
6.3.1. ‘Abd al-Jabbar 213
6.3.2. Al-Ash‘ari 215
6.3.3. Al-Maturidi 217
 Contents xi

7. Divine Command in Some Recent Jewish Thinkers 223

Introduction 223
7.1. Marvin Fox on Maimonides 224
7.1.1. An Esoteric Text 224
7.1.2. Maimonides on the Mean 226
7.1.3. Maimonides on the Reasons for the Commandments 230
7.2. David Novak 235
7.2.1. The Current Situation of Judaism 236
7.2.2. Novak on Maimonides 239
7.2.3. Novak on the Noahide Laws 243
7.3. Franz Rosenzweig 247
7.3.1. Introduction 247
7.3.2. Creation: The Disappearance of God 249
7.3.3. Revelation: Initiative and Response 251
7.3.4. Redemption: Revelation and Creation Seen Backwards 256
8. Divine Command and Evolutionary Psychology 261
Introduction 261
8.1. The Story 263
8.2. Evolution and Reducing the Moral Demand 264
8.2.1. Herbert Spencer and Larry Arnhart 264
8.2.2. Jonathan Haidt 267
8.3. Evolution and Anti-Realism 273
8.3.1. John Mackie 273
8.3.2. Michael Ruse 275
8.3.3. Sharon Street 278
8.3.4. Paul Bloom 281
8.4. Transcending our Evolutionary Situation without God 285
8.4.1. Joshua Greene 286
8.4.2. Philip Kitcher 292
8.5. Transcending our Evolutionary Situation with God 300
9. Summary 309

Bibliography 315
Index of Biblical Passages 329
Index of Names and Topics 331
 1

Morality and Religion

I N T R O D U C TI O N

This book is about divine command, and in particular it defends the thesis that
what makes something morally obligatory is that God commands it, and what
makes something morally wrong is that God commands us not to do it.1
Religion, like music, is pervasive across human culture and is very hard to
define. In many societies there is an intimate connection between religion and
the foundations of morality. The Abrahamic faiths have made this connection
through the idea of God’s command. They have had to integrate two kinds of
experience. The first is that God tells us to do something, or not to do
something. The second is that we have to work out for ourselves what to do
and what not to do. None of these faiths has been able to dispense with either
claim. The difficulty has come in reconciling them. The concern of this book is
that we remember to love God’s law and God’s command. Christians, in
particular, sometimes emphasize God’s grace as opposed to the law, and say
that they are not ‘under’ the law.2 But they can learn from Jews and Muslims,
who hold the love of God’s law closer to the surface. Christians should
remember that, despite our repeated failures to live by the law and the
command, and despite the need to be forgiven for this failure, the law and
the command are the groundwork for the rest of the narrative of redemption.
Psalm 119 is an extended expression of the gratitude of a people who would
otherwise, without heeding God’s revelation, go astray like lost sheep. The
relationship between this revelation of the law and command and our human
nature is not that we should deduce how we ought to live from how we are by
nature inclined to act; for our natural inclinations are a thorough mixture of
what we should follow and what we should not. (The term ‘nature’ here needs,

1
Strictly speaking, there is an exception to this principle, discussed and shown not to be
worrying in Section 1.3. God does not make it obligatory, by commanding it, to obey God’s
commands, but this is because the principle that God is to be loved, and so to be obeyed, is
‘known from its terms’.
2
Galatians 4: 5.
2 God’s Command
and will receive, careful scrutiny.) But God’s command to us fits our nature
very well in the sense that it guides us in discerning which of these inclinations
found in our nature we should embrace and which we should not. We also,
however, need some discernment about what to take as a divine command. In
all three Abrahamic faiths this is a key site of internal disagreement. This book
will argue that it makes sense in this discernment to look first not at abstract
principles independent of religion, but at the narratives internal to these faiths
about what God and humans are like.
In Christian reflection on this, two main traditions have emerged, divine
command theory and natural law theory. Chapter 6 of this book argues that
there is a similar division within Islam, and Chapter 7 argues this within
Judaism. These terms, ‘divine command theory’ and ‘natural law theory’, are
conventional, but this book for the most part conducts its argument in
reference to the theories of particular philosophers and theologians rather
than using these terms. The reason is that we can be clear about what we are
accountable to when we are discussing, say, Duns Scotus, namely, the texts of
Duns Scotus. It is not at all clear what we would be accountable to if we were
discussing ‘divine command theory’, unless by stipulation, as in the first
sentence of this chapter. There is no canonical text for the theory. It is better
to be content with building up an understanding of how the various thinkers
in these two traditions have held views partly similar to each other and partly
different. That said, this book is an attempt to defend a version of divine
command theory, but it also makes use of some themes more familiar from
natural law theory.
The first chapter proceeds by identifying three arguments by which we can
establish various kinds of dependence relation of morality upon religion.
These three arguments are not original, and versions of them are to be
found pervasively in the literature; the first chapter takes versions directly or
indirectly from Kant, and then describes how divine command features in this
relation. In order to understand this, we have to know what kind of thing a
divine command is, and what its species are. This is the topic of the second
chapter. The third chapter is about one typical disagreement between divine
command theorists and natural law theorists. This is a disagreement about
eudaemonism, the view that all our choices and actions are properly aimed at
our own happiness.3 This is relevant for divine command theory because, if we

3
‘Eudaemonism’ is Immanuel Kant’s term: see The Metaphysics of Morals (henceforth MdS),
in Practical Philosophy, ed. and trans. Mary J. Gregor (Cambridge: Cambridge University Press,
1999), vi. 378. I will make reference to Kant’s texts by the volume and page number of the Berlin
Academy Edition (Berlin: George Reiner, later Walter de Gruyter, 1900–) and I will use
abbreviations of the German names for the works, as given in that edition. The other texts are
these: Natürliche Theologie Volckmann [Lectures on Natural Theology (Volckmann)] (Berlin
Academy Edition); The Conflict of the Faculties, trans. Mary J. Gregor and Robert Anchor
(henceforth SF), in Religion and Rational Theology, ed. and trans. Allen W. Wood and George
 Morality and Religion 3
make our moral choices for the sake of happiness, we do not need divine
command as an answer to the question why we should choose what is morally
right; we should do so in order to be happy. The third chapter is mostly about
Aquinas on eudaemonism, but it starts from some versions to be found in the
Epicureans and Stoics, and ends by rejecting a revision of Aquinas’s view made
to accommodate some earlier objections. The topic of the fourth chapter is
what I call ‘deductivism’, the view that we can deduce our moral obligations
from facts about human nature. This is relevant for divine command theory
because, if we can deduce our moral obligations from facts about human
nature, we do not need divine command to give us the content of the moral
law. The fourth chapter has three sections. The first discusses Duns Scotus
and his denial of deductivism, the second rejects a form of deductivism in
Robert M. Adams, and the third enters into the dispute about deductivism
between R. M. Hare and Philippa Foot and Rosalind Hursthouse. In sum, the
first half of the book is largely concerned with laying out a version of divine
command theory and defending it against alternative theories.
The second half of the book relates the theory outlined in the first half to
four new areas, the first three to theological accounts in the three main
Abrahamic faiths. Chapter 5 is on a Christian theologian, Karl Barth, focusing
on three themes: his particularism (the view that the paradigmatic divine
commands are to particular people at particular times and places), his account
of human freedom, and his discussion of how we know what divine command
is being addressed to us. Chapter 6 is on divine command theory in Islam,
focusing on the work of al-Maturidi, who was a contemporary of al-Ash‘ari,
and who has a striking resemblance to Scotus. Chapter 7 is on divine com-
mand theory in Judaism, focusing on recent work about Jewish natural law
theory and its relation to Maimonides. Finally, Chapter 8 is on recent work in
evolutionary psychology, and defends the claim that thinking of our moral

di Giovanni (Cambridge: Cambridge University Press, 1996); The End of All Things, trans. Allen
W. Wood, in Religion and Rational Theology; Speculative Beginning of Human History, in
Perpetual Peace and Other Essays, trans. Ted Humphrey (Indianapolis: Hackett, 1983); Lectures
on the Philosophical Doctrine of Religion (Pölitz) [Religionslehre nach Pölitz], trans. Allen
W. Wood, in Religion and Rational Theology; On the Miscarriage of All Philosophical Trials in
Theodicy, trans. George di Giovanni (henceforth M.), in Religion and Rational Theology; Religion
within the Boundaries of Mere Reason, trans. George di Giovanni (henceforth Rel.), in Religion
and Rational Theology; Lectures on Ethics (Collins) [Moralphilosophie Collins] (henceforth MC),
in Lectures on Ethics, ed. Peter Heath and J. B. Schneewind, trans. Peter Heath (Cambridge:
Cambridge University Press, 1997); Critique of Pure Reason, ed. and trans. Paul Guyer and Allen
W. Wood (Cambridge: Cambridge University Press, 1998) (henceforth KrV); Critique of Prac-
tical Reason (henceforth KpV), in Practical Philosophy; Groundwork of the Metaphysics of Morals
(henceforth Gl.), in Practical Philosophy; Critique of the Power of Judgment, ed. Paul Guyer,
trans. Paul Guyer and Eric Matthews (Cambridge: Cambridge University Press, 2000) (hence-
forth KU); Anthropology from a Pragmatic Point of View, trans. Robert B. Louden (henceforth
Anth.), in Anthropology, History, and Education, ed. Günter Zöller and Robert B. Louden, trans.
Mary J. Gregor et al. (Cambridge: Cambridge University Press, 2007).
4 God’s Command
obligations as produced by divine command helps us see how a moral
conscience could develop in a way that is evolutionarily stable. This requires
returning to the three ways in the first chapter of establishing various kinds of
dependence relation of morality upon religion.
The programme of the book as described is absurdly ambitious, covering
such a widely disparate set of topics. What holds them together is the notion of
God’s command. What will emerge, so far as the argument of the book is
successful, is that divine command theory and natural law theory are closer
than one might expect. There remain differences between them, but the two
are in many respects complementary. There is nothing incongruous in a
divine command theorist saying that God’s commands fit human nature, or
in a natural law theorist saying that God’s commanding is a necessary
condition for a moral obligation. Nonetheless, the form of divine command
theory defended in the book remains different in some key respects from the
most familiar forms of natural law theory in the literature.
The first topic, then, is three arguments (three ‘ways’) by which we can
establish that morality depends upon religion. We can give these three ways
the names, ‘the argument from providence’, ‘the argument from grace’, and
‘the argument from justification’. The first two arguments come directly from
Immanuel Kant. The third is only indirectly from Kant, from his translation
enterprise in Religion within the Boundaries of Mere Reason, and my argument
is independent from him. Because I am starting from Kant, and this might
suggest that this is a book about Kant, I need to stress that this is not so. The
Sage of Königsberg is not a major topic in the remaining chapters, though his
influence is detectable throughout. For this reason discussion of disputes
about Kant’s text has gone into the notes. Arguments from Kant are here for
two reasons. The first is that they are good arguments. The second is that Kant
continues to be a vital figure in moral philosophy. This is shown by the fact
that even the philosophers who disagree with him feel that they have to say
why they disagree. Since Kant is not usually associated with the idea of divine
command, it is important to see the role that this idea plays in his moral
theology.4 Kant describes his own work as ‘philosophical theology’ as opposed
to ‘biblical theology’.5 This distinction reappears in Chapter 5, in relation to

4
He says throughout his published corpus that we have to recognize our duties as God’s
commands (e.g. Kpv v. 129, Rel. vi. 154). But he also says, MC xxvii. 277: ‘It is true that any moral
law is an order, and they may be commands of the divine will, but they do not flow from such a
command. God has commanded it because it is a moral law, and His will coincides with the
moral law.’
5
Rel. vi: 9: ‘Over against biblical theology, however, there stands on the side of the sciences a
philosophical theology which is a property held in trust by another faculty. This theology must
have complete freedom to expand as far as its science reaches, provided that it stays within the
boundaries of mere reason.’ The fact that Kant uses at the end of this quotation the very title of
his book suggests strongly that he has his own project in mind.
 Morality and Religion 5
Karl Barth, who deliberately takes up Kant’s description of the role of the
biblical theologian. Where possible, Kant’s arguments have been stated in
terms that do not depend upon the details of his system as a whole. But this
is not in fact possible much of the time, because he is a highly systematic
philosopher, and the terms he uses in an apparently non-technical way in one
text are given a technical sense in another.

1.1. THE ARGUMENT F ROM P ROVIDENCE

The first argument is that morality becomes rationally unstable if we do not

have a way to assure ourselves that morality and happiness are consistent, so
that we do not have to do what is morally wrong in order to be happy; it
concludes that we need belief in God to give us this assurance.6 The term
‘unstable’ here is Kant’s, and I am repeating in this section the argument he
gives in the Dialectic of the Critique of Practical Reason, but also at the
beginning of Religion within the Boundaries of Mere Reason, and at the end
of the first and third Critiques.7 Kant is arguing not that a life committed to
meeting the moral demand is impossible without belief in God, but that there
is a certain kind of rational instability in such a combination.8 We know many
more people than Kant did who combine a morally good life with unbelief in
God, and indeed the lives of some of them put the lives of many believers in
God to shame. Nonetheless, this combination betrays a lack of rational fit.
What is the moral demand? What is moral obligation? I will start with a
brief account of Kant’s view. I will, however, later make two important
qualifications to Kant, to his account of morality in Section 5.1.3, and to his
account of happiness in Section 3.2.3. Kant gives us, in the Groundwork
of the Metaphysics of Morals, various formulations of what he takes to be
the supreme principle of morality, namely the Categorical Imperative. It is
important not to think that any one of these, and in particular the first, is self-
sufficient.9 Here are two of these formulations or formulas. The first states:

6
R. M. Hare (the author’s father, henceforth RMH) gave a version of this argument very
briefly in ‘Theology and Falsification’, his contribution to the so-called University Discussion,
repr. in New Essays in Philosophical Theology, ed. Antony Flew and Alasdair MacIntyre (London:
SCM Press, 1955), and at greater length in ‘The Simple Believer’, in Essays on Religion and
Education (Oxford: Clarendon Press, 1992), 1–36.
7
The term ‘unstable’ comes from Volckmann’s notes on Kant’s Natürliche Theologie,
xxviii. 1151.
8
At KU v. 452, Kant describes Spinoza as a famously good person, but one who does not
believe in God. This may not be fair to Spinoza, but that is not my present concern, which is to
argue that Kant thinks there are people who combine moral aspiration and religious unbelief.
9
See Allen Wood, Kant’s Ethical Thought (Cambridge: Cambridge University Press, 1999),
186–90.
6 God’s Command
‘Act only in accordance with that maxim through which you can at the same
time will that it become a universal law.’10 A maxim prescribes an action
together with the reasons for, or the end to be produced by, that action. The
first formula requires that I test any maxim by asking whether I can will it
universally. Kant gives an alternative version of the first formula to make this
point clearer: ‘Act as if the maxim of your action were to become by your will a
universal law of nature.’ Since he is talking about the actions of free agents, he
does not mean that the maxim will become a law of physical nature, which
would imply that humans lose their freedom. But nature has one feature that
makes the analogy useful: nature is a system in which the same kind of cause
produces the same kind of effect in a lawful way wherever and whenever it
occurs. Kant is asking us to imagine a similar system, but a system of moral
permissions, in which the end incorporated in our maxim is included. If we
will as a universal law the maxim of a false promise to repay some money in
order to secure a loan, we have to be able to say: ‘Whenever someone is in
need, he may make a false promise to repay some money in order to secure a
loan.’ Kant thinks that, when we consider a system including this principle, we
will see that it displays a kind of incoherence. We can call any obligation that
passes this universalizing test ‘a universal obligation’. The qualification I am
going to make in Chapter 5 to Kant’s account of morality is that there are also
what I will call ‘particular moral obligations’, which are obligations to par-
ticular people who are not, even in principle, eliminable from the maxim of the
action by willing it as a universal law. But that qualification does not affect the
argument of the first four chapters, because particular obligations are stand-
ardly accompanied by universal obligations, and these are usually the focus of
the discussion.
The second formula of the Categorical Imperative is the formula of the end-
in-itself: ‘So act that you treat humanity, whether in your own person or in the
person of any other, always at the same time as an end and never merely as a
means.’11 Humanity should never be treated as a mere means. This is true,
Kant says, about humanity in myself as well as humanity in other people. I am
not allowed to treat myself as merely a means to other people getting what they
want. The word ‘merely’ is important. Kant is not forbidding using people, but
we must never merely use. To treat another person as an end in herself is to
share as far as possible her ends.12 If we return to the example of the false
promise, this gives us another reason to think it morally forbidden. By making
a false promise to a person, I am preventing our sharing ends. Her end
includes getting the money back, and mine is to get the money without
repaying the loan. False promising is a form of deception and what is being

10
Gl. iv. 421. I have said more about some of the details of this formulation in my God and
Morality (Oxford: Blackwell, 2009), 145–56.
11 12
Gl. iv. 429. Gl. iv. 430.
 Morality and Religion 7
concealed is the end I actually have in making the promise. The same is true in
cases of coercion, where, again, there is no sharing of ends, but one person’s
end is in this case imposed by force on another.
There is something common to the positions on the moral demand held
by the Kantian, the Consequentialist, and any Virtue Theorist who takes
impartial benevolence to be a virtue. The moral demand is that we treat
each person as one, and no person as more than one, and we try to make
the other’s purposes our purposes as far as we can, namely as far as the moral
law itself allows. This account itself includes reference to the moral law in its
final clause, and therefore does not explain the moral demand in a non-
circular way.
Sometimes people who know Kant’s moral theory but do not know his
moral theology wonder why he would bring in happiness at all, as the
argument from providence requires. Is he not committed to the view that
morality is to be pursued for its own sake, and would not requiring a
connection with happiness be a pollution of this kind of purity? To reply to
this worry, it is helpful to see how he distinguishes his position from the views
he attributes to the Stoics and the Epicureans. The Stoics, he says, held that
happiness is simply being virtuous (and knowing that one is virtuous), which
has the effect of reducing happiness to virtue. The Epicureans held that virtue
is simply what leads to happiness, and so in effect reduced virtue to happiness.
But Kant objects to both. We humans are not merely rational, but also
creatures of sense and creatures of need. If we were merely rational, perhaps
our highest good would be purely a life of virtue. But, because we are this
combination, our highest good is a union of virtue and happiness, which are
two different things. Virtue is the disposition to live by duty or the moral law,
and happiness is the satisfaction of our inclinations as a sum, or where
everything goes the way we would like it to.13 Since these are different, the
Stoics are wrong to try to reduce happiness to virtue. But the Epicureans, in
reducing virtue to the means to happiness, are also wrong, because they fail to
give us morality at all; morality requires that we seek to do our duty for its own
sake, not for the sake of happiness. It might be helpful here to consider an
example of a case where duty and happiness seem to conflict. Suppose my aged
parent is no longer capable of independent living, and I consider that she will
soon die if sent to a nursing home. But suppose I also predict that, if she comes
to live with me, I will get anxious and depressed, though to a still manageable
degree. If I conclude that my duty is to invite her to my home, it may be with a
clear-eyed recognition that the effect on my happiness will be negative. One
key to the argument from providence is to see that duty and happiness are

13
MdS vi. 480, and see KpV v. 22. From this definition we can see that complete happiness is
unlikely on earth for beings with our limited capacities, given our tendency to misidentify what
will in fact give us satisfaction.
8 God’s Command
different and not reducible the one to the other. This irreducibility is the main
topic of Chapter 3. The qualification I am going to make in that chapter to
Kant’s account of happiness is that not all the components of happiness are
satisfactions of what Kant calls ‘inclinations’, and they are not all properly
classified under the general heading of ‘pleasure’. This means that the differ-
ence between duty and happiness is not as stark as Kant pretends, but I will try
to show that there are still basically two kinds of motivation for action, and not
(as the Aristotelian proposes) finally only one.
Since we are both rational beings and creatures of sense and of need, our
highest good, Kant says, requires a union of virtue and happiness. This union
is not merely for us as individuals; our morality gives us the end or goal of the
happiness of all, proportional to the virtue of all. This end is the combination
of our own happiness and the happiness of others, along with our own virtue
and the virtue of others.14 But, since our morality gives us this end, the highest
good, we must, if we are to pursue the morally good life in a way that is
rationally stable, believe that this highest good is really (and not merely
logically) possible. Real possibility has to be founded on what is actual. But
we do not see that we have the capacity to bring this highest good about. What
we see, on the contrary, is a world in which people who are not committed to
the moral law get large amounts of what they wish and will, and those who are
committed to it often end in misery and frustration. Nature, Kant says, is
indifferent to our moral purposes, as far as we can tell from our sense
experience.15 In order to sustain our belief in the real possibility of the highest
good, we therefore have to postulate the existence of a ‘supersensible author of
nature’, who can bring about the conjunction of happiness and virtue, and
thus ‘morality inevitably leads to religion’.16 This argument is implicit already
in the Groundwork of the Metaphysics of Morals, in which Kant distinguishes
between the ordinary members of the kingdom of ends, to which all humans
belong, and the head of this kingdom, who, unlike us, has ‘unlimited resources
adequate to his will’.17 Kant is not, in the sense defined in this book, a ‘divine
command theorist’. But he does say throughout the corpus that we have to
recognize (erkennen) our duties as God’s commands.18 He is thinking primar-
ily of this argument from providence. We have to recognize our duties as
God’s commands, because it is only if they are God’s commands that we can
rationally believe in the real possibility of the highest good, which is the end
that morality itself gives to us. When Kant defines religion as ‘recognizing our

14
See MdS vi. 391–2, where the matter, as opposed to the form, of the duties of virtue is my
own perfection and the happiness of others. The pursuit of my own happiness is taken to be
natural and not a duty, and the virtue of others is the matter of their duties and not mine.
15
KU v. 452.
16
Rel. vi. 6. It is important to bear this sentence in mind when reading the first sentence
of Religion. I have discussed this in my God and Morality, 162–5.
17 18
Gl. iv. 434. See MC xxvii. 274, 283; KpV v. 129; Rel. vi. 154.
 Morality and Religion 9
duties as God’s commands’, the notion of religion is of a moral faith that how
things ought to be is sustained by how things fundamentally are, by the
governance of the universe.19
Kant thus subscribes to the scholastic picture of the three roles of God as
sovereign, distinguishing God’s legislative, executive, and judicial authority.20
On this picture God promulgates the law by command, runs the universe in
accordance with this law, and then judges our success in keeping this law. The
notion of divine command is fundamental to this picture. But Kant employs
the picture from the standpoint of the practical, not the theoretical, use of
reason. These different uses and their implication for the argument from
providence are discussed in more detail in Section 5.3.2.
Not only Kant, but also the classical authors of the utilitarian tradition, have
endorsed a version of the argument from providence. J. S. Mill, in his Essays on
Religion, said that we need hope with respect to the government of the
universe, if we are to sustain the moral life. Otherwise we are kept down by
what he called ‘the disastrous feeling of “not worth while”’.21 Henry Sidgwick,
in Methods of Ethics, recognized that the only way to reconcile enlightened
self-interest with aiming at the maximum balance of happiness for all sentient
beings present and future, whatever the cost to oneself, was to bring in a god
who desires the greatest total good of all living things and will reward and
punish in accordance with this desire.22 Belief in such a god is necessary,
Sidgwick thought, to restore coherence to our moral beliefs, but he did not
commit himself one way or the other as to whether this was sufficient reason
to believe. He did, however, recognize that incorporating this belief would be a
return to the utilitarianism of William Paley, which preceded Bentham’s.
It will always be possible to escape the force of this argument if we deny that
morality gives us the end of the highest good, and so requires us to believe in
its real possibility. We can always think of the moral demand, as Camus’s
Sisyphus thought of the command to roll the rock up the mountain, as absurd,
and shake our fists at the gods who have given us this task. But if we really
thought morality was absurd, we would surely find it hard to sustain our
attempt to live morally. Consider the possibility of an evil demon, rather like
the evil demon Descartes imagines, who makes it impossible for us, roughly
70 per cent of the time, to carry out what is morally good. Would we sustain
the moral life in such a world?23

19
See Andrew Chignell, ‘Rational Hope, Moral Order, and the Revolution of the Will’, in
Nancy Cartwright and Eric Watkins (eds), Divine Order, Human Order, and the Order of Nature
(Oxford: Oxford University Press, 2013).
20
See Suarez, De Legibus, 2. 15. 20.
21
See J. S. Mill, Three Essays on Religion (London: Henry Holt, 1874), 249–50.
22
Henry Sidgwick, The Methods of Ethics (Indianapolis: Hacker, 1981), 509.
23
The novels of Camus contain heroic characters who do persevere in this way, but the
question is about life outside the novel.
10 God’s Command
One reason one might think the world is absurd in this sort of way is that it
contains too much evil to make a stringent moral demand practicable. This is
not the right occasion to launch into a sustained discussion of the problem of
evil, which has its own enormous literature. But Kant himself condensed his
thoughts about the problem into a short monograph about Job.24 The title,
‘On the Miscarriage of All Philosophical Trials in Theodicy’, and much of the
content are negative. In this way this small work is like a small version of
the first Critique. And, as in the first Critique, we need to persevere past the
negative to the positive content. Kant is in fact giving a kind of theodicy, what
we might call a ‘transcendental theodicy’, even though it is not what he calls
‘theodicy proper’. A theodicy ‘proper’ would be within the limits of reason in
its theoretical employment; that is to say, it would have to be based on what we
can perceive with the senses, because our concepts are empty unless they are
given content from our sensory experience. But as in the first Critique, Kant’s
objective is to ‘deny knowledge so as to make room for faith’.25 The faith that
he wants to make room for is faith in God as legislator, ruler, and judge, the
threefold sovereign role that we have already encountered.26 The problem of
evil is a problem for the claim that there is a God like this. Kant goes through
three traditional theodicies ‘proper’ for each of these three roles, and shows
that all nine fail. Here he resembles the ‘all-destroyer’ of the first Critique,
dismantling the ontological and cosmological and physico-teleological proofs
for the existence of God.27 But then Kant makes the key move. It is true that
we have no proof within the limits of theoretical reason for God’s wisdom in
the three roles; but we also have no disproof. To attempt a disproof would
transgress the limits of our insight just as much as the attempted proofs. By
acknowledging our ignorance in these matters, we can show that the attempted
disproofs fail. But if there is no disproof within theoretical reason, the need of
practical reason for the postulation of this divine wisdom prevails.
This brings us to Job and his friends. Kant compares the spirit in which Job
speaks with the spirit of his so-called friends. The friends urge that Job’s
sufferings must be a result of God’s punishing Job’s wrong-doing, even if
Job does not know what that wrongdoing could be. The friends speak, says
Kant, as though they were ingratiating themselves with God. Job alone is frank
and sincere. He does not hide his doubts, but he also does not deceive himself
about his own guilt. What God does in the story is to reveal (out of the
whirlwind) the wisdom of the creation, and especially its inscrutability. God

24 25 26
M. viii. 253–71. KrV Bxxx. M. viii. 257.
27
Note, however, that, in the case of the third, the physico-teleological proof, Kant is much
less destructive. He thinks we can show the existence of a highest architect of the world, though
not a creator of the world. See KrV A627 = B655. This is rather like Philo at the end of Hume’s
Dialogues Concerning Natural Religion, part XII. In M. Kant acknowledges similarly that we can,
from physico-theology, get to an ‘artistic wisdom’ in the designer, even within our speculative
faculty of reason (M. viii. 263).
 Morality and Religion 11
shows to Job the beautiful side of creation, but also the fearsomeness of it.28
Using Otto’s language, we might say that God shows the divine wisdom as
both fascinans and tremendum.29 Job comes to see, and Kant quotes him:
‘I have uttered what I did not understand; things too wonderful for me, which
I knew not.’30 This is, Kant says, a ‘vexing resolution’, and it could count as a
resolution at all only in a man who could say, even in the midst of his doubts:
‘Till I die I will not remove mine integrity from me.’31 In other words, Job
founded his faith on his commitment to the moral life. He is like the naturally
honest people described at the very end of Religion, who go from virtue to
grace, not from grace to virtue.32 Grace is the topic of the next section of this
chapter. But, before we get to that, we should consider how Kant sums up what
Job’s kind of sincerity is like. It does not mean taking the possibility of error so
seriously that one refuses to believe anything one cannot prove.33 Rather, it
means being aware always of the limits of our own understanding, and not
transgressing them in either a positive or a negative direction. If we are sure
that we are under the moral law, then we are entitled to believe in the existence
of a ruler of the world who makes the evil in the world (which we cannot deny)
subordinate to the good.

1.2. THE ARGUMENT F ROM GRACE

The second way of establishing a dependence relation of morality upon God is

by means of the argument from grace, again an argument from Kant’s
Religion. To understand this argument we need to see that Kant recommends
that we see revelation as two concentric circles, with historical revelation (the
revelation to particular people at particular times and places) in the larger
circle and in the smaller the revelation to reason (available at least potentially
to all people at all times and places). The project of Religion within the
Boundaries of Mere Reason is then to determine whether doctrines in the
outer circle can be translated into the language of the inner circle by means of
the moral concepts. In fact, Kant attempts this translation with the four
doctrines of Creation, Fall, Redemption, and Second Coming, and this project
dictates the structure of the work as a whole. Each of the four parts of the work
ends with a General Remark on a topic that Kant says belongs in the larger

28
M. viii. 266. God reveals the beauty of the horse and the hawk, but also Behemoth and
Leviathan.
29
Rudolf Otto, The Idea of the Holy (Oxford: Oxford University Press, 1923), 12–40.
30 31 32
Job 42: 3. Job 27: 5–6. Rel. vi. 202.
33
M. viii. 268. See William James, ‘The Will to Believe’, in Pragmatism and Other Essays
(New York: Pocket Books, 1963), 205: ‘But I can believe that worse things than being duped may
happen to a man in this world.’
12 God’s Command
circle, but in the area of the larger circle that borders upon the smaller one. At
the end of part one, the topic is ‘effects of grace’. Kant is here discussing a
problem he refers to elsewhere as ‘Spener’s problem’, after the great Lutheran
Pietist.34 We humans are born, Kant says, under the evil maxim, which
subordinates duty to happiness. Evil is not, though we are sometimes tempted
to think so, simply the product of our sensory inclinations. Rather, it is a
choice in the will to rank happiness over duty. Kant is here in the tradition of
Luther, who denies that the source of evil is in the ‘lower and grosser
affections’ and locates it instead in ‘the highest and most excellent powers of
man, in which righteousness, godliness, and knowledge and reverence of God,
should reign—that is, in reason and will’.35 Since we are born under this
ranking of happiness over duty, we cannot reverse the ranking by our own
devices, for this would require a choice that was already under the opposite
ranking. Kant says that the propensity to evil is ‘not to be extirpated through
human forces, for this could only happen through good maxims—something
that cannot take place if the subjective supreme ground of all maxims is
presupposed to be corrupted’.36 Here we have the problem that I have else-
where called the problem of ‘the moral gap’, a gap between how we ought to
live and how we can live by our own devices; ‘ought’ implies ‘can’, but in this
case we ought to give duty the priority ranking, yet we seem to have a radical
incapacity to do so. To talk about a radical incapacity or about being ‘under
the evil maxim’ is not to say that we are fundamentally evil. We are born, Kant
thinks, with both the predisposition to good and the propensity to evil, and of
these two only the predisposition to good is essential to us. Nonetheless, it is
natural to us, when forced to rank the two, to put our own happiness first and
duty second. This means that legislators should try to set up laws so that we are
forced to rank the two as seldom as possible.
By presenting the problem in terms of a ranking of incentives, Kant puts
himself again in the tradition of Luther and Augustine. Augustine says that
God bids us do what we cannot, in order that we might learn our dependence
upon God. In On Free Choice of the Will, he says both that we have lost our
freedom to choose to act rightly and that we do have the ability to ask God for
assistance: ‘In the midst of their ignorance and difficulty he leaves them the
free will to ask and seek and try. He will give to those who ask, show himself to
those who seek, and open to those who knock.’37 The key to a solution to the

34
SF vii. 54.
35
Martin Luther, The Bondage of the Will, trans. J. I. Packer and Olaf Raymond Johnston
(New York: Fleming H. Revell, 1957), 280.
36
Rel. vi. 37.
37
Augustine, On Free Choice of the Will, trans. Thomas Williams (Indianapolis: Hackett,
1993), iii. 20. But this is an early text of Augustine, and his later work is different on this topic.
See Retractations 4: ‘And unless the divine grace by which the will is freed preceded the act of
will, it would not be grace at all.’
 Morality and Religion 13
problem of the moral gap is to see that, while ‘ought’ implies ‘can’, ‘ought’ does
not imply ‘can by our own devices’. There are things we can do, but only with
assistance from outside. Kant thus appeals to God’s assistance in accomplish-
ing what he calls ‘a revolution of the will’, by which the ranking of happiness
over duty is reversed. This divine assistance is an effect of grace.38 Kant says
that ‘we can admit an effect of grace as something incomprehensible, but we
cannot incorporate it into our maxims for either theoretical or practical use’.39
We cannot make theoretical use of effects of grace because they go beyond the
limits of the understanding, and Kant thinks we need to confine the theoretical
use of reason within these limits. We cannot make practical use of effects of
grace because they are things God does and not things we do. Nonetheless, the
appeal to effects of grace is the solution to what would otherwise be a
contradiction in practical reason; we both ought to and cannot live by the
moral law. We can add, though this goes beyond Kant, that one of the effects
of grace that makes the moral life livable is that grace makes forgiveness
possible, in cases where we cannot forgive ourselves for moral failure, because
we do not have the right moral status to do so. The view of the effects of grace
defended here is that God intervenes in our situation, and enables us to live by
the high moral demand placed on us by divine command.
There is a large theological problem here. Does not God put all human
beings under the moral demand? But does not this mean that God gives all
human beings the means to comply with it? It is, after all, incoherent to put
someone under a demand that he or she cannot reach. This is the meaning of
‘ought implies can’. There is a tendency in this thought towards universalism,
the view that all human beings are saved. Kant may have been a universalist.
At least he thought that the doctrine of double election (of some to salvation
and others to damnation) was the salto mortale (the death leap) of human
reason.40 There is also a tendency in Barth towards universalism.41 In
Chapter 6, we will discuss the reaction of al-Ash‘ari to the set of questions
here. He concedes that, if we are to have the power to comply with the divine
commands, God has to give us this power. But he denies that there is any
injustice in God’s declining to give such power to unbelievers; injustice, for
al-Ash‘ari, is disobedience to a command, and there is no one in a position to

38
One difficulty here is that Kant also says, at Rel. vi. 44: ‘The human being must make or have
made himself into whatever he is or should become in a moral sense, good or evil.’ I have discussed
this in my ‘The Place of Kant’s Theism in his Moral Philosophy’, in Mark Timmons and Sorin
Baiasu (eds), Kant on Practical Justification (Oxford: Oxford University Press, 2013), 300–14. The
key is to see that Kant’s idiosyncratic use of the term ‘moral’ means that moral improvement, but
not improvement in general, requires self-production.
39 40
Rel. vi. 53. Rel. vi. 121.
41
See Bruce McCormack, ‘So That He May Be Merciful to All: Karl Barth and the Problem of
Universalism’, in Bruce McCormack and Clifford Anderson (eds), Karl Barth and American
Evangelicalism (Grand Rapids, MI: Eerdmans, 2011), 227–49.
14 God’s Command
command God. Christians may, in a weak sense, hope that all are saved. They
should not, however, have hope in the strong sense of expectation that this is
so, for two reasons. The first is that there are too many contrary texts in the
Scriptures. There are passages (mainly from Paul) that suggest universalism,
but the texts from the Gospels are predominantly opposed to it. The second
reason is that Christians have the experience of people who die denying God,
or at least their picture of God; ‘at least their picture’, because in many cases
the God they deny is a God that believers would also deny, but this is not
always the case as far as can be discerned. The best conclusion is that God does
make the moral demand of all human beings (though there are also divine
commands and counsels, as Chapter 2 will urge, that are not the same for all
human beings) and God does offer assistance to all human beings to meet this
demand. But this assistance, though sufficient for all, is not efficient for all;
that is to say, it does not bear the fruit of obedience in every human life.42
Chapter 5 says more about human freedom. For now, it is enough to say just
that the grace offered by God to meet the divine command is not irresistible.
Not only Kant, but also many of the theologians and philosophers who
preceded him, recognized the presence of the gap between the moral demand
and our natural capacities. The picture is basically Augustinian, one Kant
inherited through Luther and the Lutheran Pietists.43 It can also be found
outside Christianity, for example, in Aristotle and in the Neo-Confucian Chu
Hsi.44 However, some of the details in Kant’s filling-out of the moral gap
picture are peculiar to Kant. He thought he had identified the supreme
principle of morality, which he called ‘the Categorical Imperative’, and he
thought that this imperative was part of the content of the religion of reason,
the inner circle. He sometimes suggested that this conception of the moral
demand is present to all human beings at all times and places.45 This is no

42
One way to put this is that God gives the capacity either to accept or to resist the divine
assistance. We will look at al-Maturidi’s suggestion along these lines in Chapter 6. Another
possibility is that God offers a post-mortem opportunity to everybody for salvation, but we have
no evidence of this in Scripture and the suggestion seems ad hoc.
43
See John E. Hare, ‘Augustine, Kant, and the Moral Gap’, in Gareth B. Matthews (ed.), The
Augustinian Tradition (Berkeley and Los Angeles: University of California Press, 1999), 251–62.
44 John E. Hare, Why Bother Being Good? (Eugene, OR: Wipf and Stock, 2002), 31. Aristotle

said (Nicomachean Ethics (henceforth NE), X. 7. 1177b26f) that ‘we should not agree with those
who exhort us, because we are human, to think of human things, or because we are mortal, think
of mortal things; we ought rather to take on immortality as far as possible’. Chu Hsi held that we
should think of human beings (at least most of them) as prevented from a clear view of right
principle (of which heaven is the source) because they are like pearls lying in muddy water.
45
This is suggested by his remarks about the predisposition to good at Rel. vi. 28. But the
assumption that Kant makes the descriptive claim that all humans are aware of this ‘revelation to
reason’ is probably wrong. In the last part of the Pölitz Lectures on the Philosophical Doctrine of
Religion, xxviii. 1122 ff., Kant says that what is universal is the capacity to form a morally
determinate concept of God, but that the exercise of this capacity can be blocked in various
ways—for example, among the Greeks and Romans, by the fact that ‘they knew so little of
morality that was pure and certain’.
 Morality and Religion 15
doubt wrong, though empirical evidence does suggest that at the beginning of
human society hunter-gatherer groups had some sense of fairness, and this
topic will return in Chapter 8. This book, nevertheless, is using a conception of
morality that is fundamentally Kantian, with the two significant qualifications
already mentioned. Without a high sense of the moral demand, the argument
from grace is much less plausible.
All this doctrine about the moral gap and the need for God’s grace has a
tendency that should be repudiated. The tendency is to reduce God’s grace to
help in meeting the demands of the moral law. Kant is prone to this tendency,
though the end of this chapter will argue that we have to be careful to note
where he is, so to speak, in his translation project, whether he is talking about
the doctrines before or after he has translated them into the language of the
revelation to reason using the moral concepts. But Kant makes morality too
important a part of the godly life. In the next life, contrary to Kant, the moral
law will probably not be relevant any longer. We will not be under the
constraint of a proscription on lies or murder or theft. Even the command
to love our brothers and sisters will not have any longer the character of
obligation. Finally, even in the present life, there is something unwholesome
about the focus on the moral gap. While being grateful for God’s commands,
we should live as those forgiven, and delight in all the other blessings that God
has given us through grace.46
To return, however, to the argument of this section, it is always possible to
evade the conclusion of the argument by denying the premiss about the
stringency of the demand. There are many ways to make this demand less
stringent by distancing it from the two formulas of the Categorical Imperative
already described.47 For example, the Kantian formulation requires us to treat
humanity in every person as creating obligations for us, so that we are to share
the morally permitted ends of all those in need, wherever they are in the world,
whose lives we affect by our actions. One way to reduce the demand is to say
that, unlike the Good Samaritan in Jesus’s parable, we should consider that we
have obligations only to the people we know, who are related to us in special

46
Calvin talks as though there are two motivations for obeying God: the law can be ‘to the
flesh like a whip to an idle and balky ass, to arouse it to work . . . a constant sting that will not let
him stand still’ (Institutes of the Christian Religion, ed. John T. McNeill, trans. Ford Lewis Battles,
2 vols (Philadelphia: Westminster Press, 1960), 2. 7. 12); but on the other hand Christians should
act with the freedom of children: ‘sons, who are more generously and candidly treated by their
fathers [than servants are], do not hesitate to offer them incomplete and half-done and even
defective works, trusting that their obedience and readiness of mind will be accepted by their
fathers, even though they have not quite achieved what their fathers intended’ (Institutes, 3. 14. 5).
Perhaps the way to bring these two different ideas together is to think that as time goes on
Christians should act less from fear of the whip and more in filial trust. I owe these references to
Matthew Vermaire.
47
I have described some of these ways in my The Moral Gap: Kantian Ethics, Human Limits,
and God’s Assistance (Oxford: Clarendon Press, 1996), 142–69.
16 God’s Command
relations of family or friendship or community.48 But this reduction will end
up with an ethics that is unacceptably parochial (though there may be no way
to establish this without begging the question).
Christianity does not require Kantian morality, and this is true also of the
Abrahamic faiths as a whole. But there is, nonetheless, a congruity between
Kant’s formulations and the Lutheran pietism in which he was raised. To the
extent that we lower the demand, we also lower the need for grace.

1.3. THE ARGUMENT F ROM J USTIFICATION

The third way of establishing a dependence relation of morality upon God is

by means of the argument from justification. We can ask what Christine
Korsgaard has called ‘the normative question’, which is ‘Why should I be
moral?’ or ‘Why should I accept morality as a proper demand upon me?’49
Here we cannot rely upon an argument from Kant, because Kant does not
think he can give a justification of the moral demand. At least in the second
Critique he simply starts from what he calls ‘the fact of reason’, that we are
under the moral law. In an earlier work, the Groundwork of the Metaphysics of
Morals, he was more sanguine, attempting a proof from freedom. But he seems
to have given up the attempt, on the grounds that it was viciously circular. The
present chapter will end, however, by claiming an indirect connection between
a divine command justification of the moral demand and Kant’s view that our
dignity as humans resides in our responsiveness or our potential to respond to
the moral law. In the meantime, instead of Kant, we can reflect on what
Thomas Aquinas says in Summa Theologiae I. 5. 1, not so much following
him as giving an account in response to his, and in response also to some
thoughts of Duns Scotus. The argument from justification is weaker than the
first two arguments because, even though the divine command theorist has an
answer to the normative question, it would take more work, going beyond
the scope of this book, to show that it is the only possible answer or even the
best one.50
A divine command theorist will say that the answer to the normative
question is that I should accept morality as a proper demand upon me because
it is God who places this demand. This statement is, however, incomplete.

48
Luke 10: 25–37.
49
Christine Korsgaard, with G. A. Cohen, Raymond Geuss, Thomas Nagel, and Bernard
Williams, Sources of Normativity, ed. Onora O’Neill (Cambridge: Cambridge University Press,
1996), 16.
50
I have done some of this work in a less formal book, Why Bother Being Good?, chs 6–9.
C. Stephen Evans gives a good account of other possible answers in God and Moral Obligation
(Oxford: Oxford University Press, 2013), ch. 4.
 Morality and Religion 17
A justification of a normative claim cannot be derived from a factual claim
alone. To say that I ought to live a certain way because God tells me to do so
requires, for completeness, the claim that I ought to do what God tells me to
do. This feature of justification has led some philosophers, indeed Korsgaard
herself, to think that a divine command justification begs the question. They
ask: ‘Why should I do what God tells me to do?’
The difficulty they raise can be put in terms of a dilemma: either obedience
to God is itself a moral obligation, or it is not.51 If it is, then to justify moral
obligation by appealing to it is viciously circular. If it is not, then again it seems
no justification is available by this route. For it seems impossible that we could
justify the claim that we have an obligation by appealing to something that is
not itself a higher obligation. Consider Socrates’ attempt to show Crito (in the
eponymous dialogue) that there is an obligation to obey the laws of the city.
He tried various justifications. Simplifying the case, he tried arguing that it is
wrong to disobey because disobeying harms the city, and it is always wrong to
harm someone even in return for a harm we have received ourselves. He tried
arguing that we have to obey the city because we have the obligation of
gratitude to what has formed us, in the way that we have the obligation
of gratitude to our parents. He tried arguing that we ought to obey because
it would be unfair to take the benefit of general obedience by the citizens
without paying our share of the cost. All these arguments have the form of
deriving a justification of a claim that we have an obligation of a certain kind
by deriving it from a higher obligation. It is hard to see what other kind of
justification there could be. But then the project of justifying our moral
obligations as a whole seems hopeless.
There is a reply to this difficulty, however. Here we can use the distinction
Scotus draws between natural law strictly speaking and natural law in an
extended sense. He thinks that the command to love God given in the first
table of the Ten Commandments (the law brought down by Moses on two
tables or tablets from Mount Sinai) is natural law strictly speaking. It is
known to be true just by knowing its terms (or follows from propositions
known in this way). But he thinks the second table, which concerns our
various duties to the neighbour, is natural law only in an extended sense. It is
true, but only contingently so. Scotus’s view about the second table is one of
the main topics of the fourth chapter. For our present purposes we need to
focus on the first table. It is necessarily true, Scotus holds, that God is to be
loved. We know this just by knowing the terms ‘God’ and ‘to be loved’. This
is because we know that, if God exists, God is supremely good, and we know

51
Stephen Darwall raises this problem in The Second-Person Standpoint: Morality, Respect,
and Accountability (Cambridge, MA: Harvard University Press, 2009), 104–9. We will return to
his specific form of the objection (from Cudworth) in Section 2.3.2.
18 God’s Command
that what is supremely good is to be loved.52 It is also true that we know that
to love God is at least to obey God. There is scriptural warrant for this. Jesus
says, in John 14: 23–4: ‘If anyone loves me, he will obey my teaching. . . . He
who does not love me will not obey my teaching.’ But independently of
Scripture it is plausible to say that to love God is at least to will what God
wills for us to will.53 One way to show this is to appeal to the Kantian
principle discussed at the beginning of this chapter that to treat another
person as an end in herself (Kant calls it ‘practical love’) is to share as far as
possible her ends.54 To love God is to share God’s ends as far as we can. But
our love for God is different in this way from our love for other human
beings, where we share joys and sorrows, and so the friend is, in Aristotle’s
term, ‘another self ’. In the case of our love of God there is a disproportion,
since we are finite beings and God is infinite. Loving God is not simply to
repeat God’s will in our will, because there are things God wills that God does
not will for us to will. So what we are to repeat in our wills is God’s will for
our willing. But willing what God wills for our willing is obedience. So it is
necessarily true not just that God is to be loved, but that God is to be obeyed.
If I justify the claim that the moral demand is a proper demand upon me by
saying that God’s command makes things obligatory, I am not terminating
the justification in something that itself requires justification, except in as far
as I have to justify the claim that God exists.55 This means that divine
commands do not generate all our obligations, because there is one import-
ant exception, namely the very obligation to obey divine commands. But this
is not a troubling exception once one accepts the necessary truths (if God
exists) that God is to be loved and that God is to be obeyed.
With this conclusion in place, we can make progress in answering a difficult
question about divine command theory: ‘What is the relation between the
divine command and the obligation?’ Various answers have been proposed.
Philip Quinn suggests that the relation is causal; the command causes the

52
See also Ockham, In Sent., I. 4: ‘It is because God is the greatest good that He is to be loved
above all [Dico quod solus Deus est summe diligendus, quia est summum bonum].’ I will give
shortly an account of what ‘supremely good’ means that does not assume a whole Thomist
metaphysics about the coextension of degrees of goodness and being. I should note also that
Scotus says strictly not that we know from its terms that God is to be loved, but that we know that
God is not to be hated. This is because he is treating love and hate as occurrent actualities and not
dispositions, and this means that the command to love God would, in the absence of temporal
limitation, require us to be thinking about God all the time. If we take love as a disposition, we
can say that we know from its terms the proposition that God is to be loved.
53
See also Ockham, In Sent., II. 7: ‘In the act of charity I love God and everything that God
wills me to love [objectum caritatis sit totum istud: ‘Deus et omne quod Deus vult diligi].’
54
KpV v. 83.
55
See Maimonides, Mishneh Torah: Yesodei ha-Torah, 1.1, ed. Rabbi Eliahu Touger (Jerusa-
lem: Moznaim, 1989), who says that the law itself is to be obeyed because God is God, and to
deny that is to deny ‘the foundation of all foundations’.
 Morality and Religion 19
obligation to exist.56 Robert M. Adams suggests that the relation is one of
constitution; the command constitutes the obligation.57 But causation seems
wrong because it suggests two different events, in logic separable (though one
always comes after the other under physical law). And constitution seems
wrong because obligation is not a natural kind like water, which is constituted
by two hydrogen atoms and an oxygen atom. So consider a king who makes a
law by declaring: ‘The king wills it.’ This is an ‘explicit performative’, in John
Austin’s terminology, like ‘I promise’.58 J. R. Searle claimed that the case of
promising showed that it is possible to derive an ‘ought’ from an ‘is’.59 Searle
argued that from the statement that Jones uttered the words ‘I hereby promise
to pay, you, Smith, five dollars’, it follows, given the institution of promising,
that Jones has promised to pay Smith five dollars, and that therefore Jones has
placed himself under an obligation to pay Smith five dollars. R. M. Hare
replied that, if I conclude that Jones ought to pay the money, I must be
endorsing the institution, and that this endorsement is itself a kind of pre-
scription.60 We do not have a case here of a prescription (‘he ought’) following
from a description (‘he said’) without any intervening prescriptive premisses.
But now consider the case ‘God commanded Jones to pay Smith five dollars’. If
God’s command makes something obligatory, it will follow that Jones ought to
pay Smith five dollars. Is this deriving an ‘ought’ from an ‘is’? No, because the
truth that God is to be obeyed plays the same role in the argument as the
institution of promising. If I affirm that God is to be obeyed, I am prescribing
obedience, and so prescribing that, if God commands Jones, Jones ought to
do what God commands. It is odd to say that God’s command causes the

56
Philip Quinn, ‘Divine Command Theory’, in Hugh LaFollette (ed.), Guide to Ethical Theory
(Oxford: Blackwell, 1999), 54–5.
57
Robert M. Adams, Finite and Infinite Goods: A Framework for Ethics (Oxford: Oxford
University Press, 1999), 250. Another version of this thesis is that of Evans, God and Moral
Obligation, 26.
58
J. L. Austin, How to Do Things with Words, ed. J. O. Urmson (Oxford: Oxford University
Press, 1962), 69: ‘If someone says, “I shall be there”, we might ask: “Is that a promise?” We may
receive the answer “Yes”, or “Yes, I promise it” (or “that . . . ” or “to . . .”), whereas the answer
might have been only: “No, but I do intend to be” (expressing or announcing an intention), or
“No, but I can foresee that, knowing my weaknesses, I (probably) shall be there.” ’ He goes on,
‘The situation in the case of actions which are non-linguistic but similar to performative
utterances in that they are the performance of a conventional action (here ritual or ceremonial)
is rather like this: suppose I bow deeply before you; it might not be clear whether I am doing
obeisance to you or, say, stopping to observe the flora or to ease my indigestion. Generally
speaking, then, to make clear both that it is a conventional ceremonial act, and which act it is, the
act (for example, of doing obeisance) will as a rule include some special further feature, for
example . . . tapping my head on the ground. . . . And so it is with putting in the expression
“I promise that”. It is not a description, because (1) it could not be true or false; (2) saying “I
promise that” (if happy, of course) makes it a promise, and unambiguously a promise.’
59
J. R. Searle, ‘How to Derive “Ought” from “Is” ’, Philosophical Review, 73 (1964), 44. We
return to this in Section 4.4.3.
60
R. M. Hare, ‘The Promising Game’, in Essays in Ethical Theory (Oxford: Oxford University
Press, 1989), 130–44.
20 God’s Command
obligation to exist, just as it is odd to say that the words ‘I promise’ cause the
obligation to exist. Given the institution, we do not have two separate events in
a causal relation (like throwing the stone and breaking the window). Also, it is
odd to say God’s command constitutes the obligation, as though the command
were somehow still in the obligation in the way in which oxygen is physically
in the water. It is better to see the relation as resembling that of performative
acts like a priest’s baptizing, or a president’s signing a bill. In all these cases,
there is an internal conceptual connection between the ‘producing’ and what is
‘produced’, but making the inference requires the endorsement of an institu-
tion or of a necessary truth.
There is a reply here to an objection to divine command theory proposed by
Nicholas Wolterstorff.61 He denies that all moral obligations are generated by
God’s commands; some are generated by human commands, such as a
parent’s command to a child to clean up his room. An unsuccessful reply to
this objection is to say, like Robert M. Adams, that these obligations generated
by human commands are merely pre-moral obligations, and only God’s
command makes something morally obligatory.62 This reply is unsuccessful
because human commands do sometimes make things morally obligatory. But
we should see divine command theory as operating in answer to the normative
question why we should hold ourselves under those obligations. Granted, for
example, that, if I have promised to take my children out for lunch, and I have
an obligation to keep my promises, then I have an obligation to take them out
for lunch. There is still the question why I should keep my promises.63 To
draw the implication from my having said ‘I promise’ to my obligation, I need
to endorse the institution of promising, and the fact that God requires this
faithfulness of me gives me a reason for this endorsement.
The view just described from Scotus answers the normative question by
referring back to something that is good in itself. More needs saying about
what is meant by ‘good’. Here we can go to what Aquinas says in Summa
Theologiae I. 5. 1:
Goodness and being are really the same, and differ only in idea; which is clear
from the following argument. The essence of goodness consists in this, that it is in
some way desirable. Hence the Philosopher says: Goodness is what all desire.
Now it is clear that a thing is desirable only in so far as it is perfect, for all
desire their own perfection. But everything is perfect so far as it is actual.
Therefore it is clear that a thing is perfect so far as it is being; for being is the
actuality of every thing, as is clear from the foregoing. Hence it is clear that

61
Nicholas Wolterstorff, Justice: Rights and Wrongs (Princeton: Princeton University Press,
2008), 271–3.
62
Adams, Finite and Infinite Goods, 242–5.
63
Chapter 4 will return to this in discussing Philippa Foot’s view of promising in her book
Natural Goodness (Oxford: Oxford University Press, 2001) (henceforth NG).
 Morality and Religion 21
goodness and being are the same really. But goodness expresses the aspect of
desirableness, which being does not express.64
I am not going to try to explain everything in this complex passage. In fact,
I have made a lifetime habit of not attributing views to Aquinas, because I have
found that Aquinas scholars have such different views and hate each other so
much that any attribution is likely to occasion deep animosity. I will rely on
the magisterial volume by Eleonore Stump, who explains this passage by
saying that ‘being’ and ‘goodness’ are the same in reference and differ in
sense, and that the sense of the good is that what is good is desirable.65 She
adds: ‘Goodness supervenes on the natural property of the actualization of
a specifying potentiality.’ So, for example, if we take rationality to be the
specifying potentiality of human beings, that is, the potentiality that makes
something belong to the species ‘human’, this means that a human being is
good to the extent that the human is rational. Probably this is not a good way
to specify what makes something human, as opposed to belonging to another
species, because it leaves too many people out. Probably also goodness is not to
be confined in its reference to degrees of natural-kind actualization. Some
of the things we properly call good are to be given this kind of account.
A good downy woundwort is one that displays well the characteristic features
of downy woundworts. But there are many things that we properly call good
that are not good because they actualize a natural kind. Pleasure is good, for
example, and so is aesthetic beauty. But neither of these forms of goodness is
amenable to this kind of analysis.
Aquinas’s account of the sense of ‘good’ is separable, however, from this
account of its reference. ‘Desirable’ is a term of notorious difficulty. Something
can be desirable in the sense that it is able to be desired, or that it tends to be
desired, and in those senses its desirability follows from the fact that it is very
often desired. But in a different sense something is desirable if it is worthy to be
desired, and in that sense a thing can be very often desired but still not
desirable. A thing can also be desirable in this sense, and not desired. Aquinas
refers in the passage I quoted to the first sentence of Aristotle’s Nicomachean
Ethics, but Aristotle fails to make this important distinction in this sentence.
From the fact that all people desire something, it does not follow that it is
desirable, in the sense of ‘worthy to be desired’. On the other hand, to say that
something is good is not simply to say that it is worthy to be desired or loved,
but also to express one’s desire or love for it. It is controversial in the literature
whether to be ‘internalist’ or ‘externalist’ about evaluative judgements (whether

64
Thomas Aquinas, Summa Theologiae (henceforth ST), I. 5. 1, trans. English Dominicans
(London: Burns, Oates, and Washbourne, 1912–36; repr. New York: Christian Classics, 1981).
65
I have discussed this question at greater length in my ‘The Supervenience of Goodness on
Being’, in Kevin Timpe (ed.), Metaphysics and God (New York: Routledge, 2009), 143–56. The
account by Eleonore Stump is in Aquinas (New York: Routledge, 2003), 71.
22 God’s Command
to think there is an internal relation between making the judgement and
desiring or loving). Section 4.3.2 will suggest that we should be ‘prescriptive
realists’, holding that evaluative properties like goodness are independent of
whether a person makes a judgement about them, but that such a judgement
nonetheless expresses a state of the desire or will.66 There is something odd
about the combination of saying sincerely of a thing that it is good and being
indifferent to it.67 This commitment to a form of internalism that we could call
‘judgement internalism’ is going to be unacceptable to some of those who
would otherwise accept the version of divine command theory proposed in
this book. They should be reassured that much of the theory would survive
subtracting that element from it.
The good is the desirable in the following sense: if I say that something is
good, I express the fact that I desire or love it, and I claim that it merits such
desire or love.68 To say sincerely that something is good is to express that one
is drawn by it, and to endorse the claim that it deserves to draw one in that
way.69 If we accept this proposal, we will have an account of the supervenience
of goodness upon being: To say something is good—for example, a
strawberry—is to say that it is good because of its natural properties—for
example, sweetness, redness, ripeness, firmness, and so on (these are my
criteria for goodness in strawberries); but this description is not entailed by
the evaluation that the strawberry is good, without the prior endorsement of
this set of descriptive criteria.70 This account of the supervenience of evalu-
ation upon description is consistent with the account given by Stump, but is
more accommodating in what can count as criteria. Stump’s more plausibly
Thomist account is that the criteria for goodness are degrees of being in the
special sense of degrees of actualization of specifying potentialities. As stated
earlier, this more plausibly Thomist account seems too narrow an account of
the criteria for goodness as a whole.

66
I have given a fuller account of this view in my ‘Prescriptive Realism’, in Andrew Moore
and Michael Scott (eds), Realism and Religion (Aldershot: Ashgate, 2007), 83–101.
67
Section 4.3.1 discusses the dispute about whether a person who does not feel any sense of
being drawn can nonetheless use moral language in a ‘full-blooded’ way.
68
See A. C. Ewing, The Definition of Good (New York: Macmillan, 1947), ch. 5, who defines
‘good’ as ‘a fitting object of a pro attitude’ and ‘bad’ as ‘a fitting object of an anti attitude’. See also
‘Abd al-Jabbar’s definition of ‘good’, described in Section 6.2.1, as what deserves praise.
69
See my Why Bother Being Good?, ch. 6. Note that on this account of goodness it does not
follow from a thing’s being good that anyone desires it, or from the fact that anyone desires it that
it is good.
70
See Alvin Plantinga, ‘Naturalism, Theism, Obligation and Supervenience’, Faith and
Philosophy, 27/3 (2010), 247–72. It was in fact RMH who reintroduced the term ‘supervenience’
into the discussion of moral philosophy. His view was that the decision of principle explains the
asymmetry of implication between supervening property and subvening base. The use I am
making of the term is consistent with Plantinga’s, but I do not hold that divine command theory
requires that for every obligation there is ‘a command such that it is essential to God to issue it to
all rational creatures’.
 Morality and Religion 23
We can now return to the divine command answer to the normative
question, that the demand of morality is a proper demand upon me because
it is God who makes the demand. This answer is incomplete, and depends
upon the addition of the claim that the principle that God is to be loved is
known from its terms, and is natural law strictly speaking. We can tie this to
the sense of ‘good’ as follows. When I say that something is good, I am
expressing my desire or love for it and I am claiming that it is worthy to be
desired or loved. Note that this account of the meaning of ‘good’ itself contains
the evaluative terms ‘deserve’ and ‘worthy’. The account thus escapes Moore’s
objection (in the ‘open question’ argument) to accounts of evaluative terms
that attempt to reduce them to descriptions, such as the account of ‘good’ as
‘conducive to pleasure’.71 The theist claims that what is worthy and deserves
our love is, supremely, God, and in saying God is good she expresses that love.
She may add that, secondarily, what is worthy of love is her own love of God,
and others’ love of God. The connection is that, if something is good, the love
of it (which is the recognition of its goodness and the internal movement
towards it) will, other things being equal, be good. Scotus suggests that loving
God is itself not only good but our destination as final end.72 She may say,
thirdly, that what promotes this love of God or draws us toward it is good.
These references to God are not part of the meaning of ‘good’, but are claims
about what satisfies the criteria for worthiness to be desired or loved.
If this account is correct, we can see that there can be two seemingly opposite
priority relations between what is obligatory and what is good. On the one
hand, the good has priority over the obligatory, because the justification
relation is as just explained: I should try to meet my moral obligations because
God gives them to me, and I prescribe that God is to be obeyed because God is
to be loved as the supreme good. There is a separate point that the commands
God gives me are to do good things, and their goodness gives me a reason for
doing them. On the other hand, the obligatory has priority over the good,
because there is an enormous number of good things possible for me to do, and
God, in prescribing some obligation, selects some of these goods and neglects
others. Only the ones God selects for prescription are obligatory.73 Chapter 5
discusses Karl Barth on particularity, and defends the claim that God’s pre-
scriptions are most importantly to particular people at particular times. This
makes it easier to see how God, in prescribing, is selecting some goods and

71
G. E. Moore has different versions of the open question argument in Principia Ethica (1903;
Cambridge: Cambridge University Press, 1993). One comes on pp. 176–7. But see RMH, The
Language of Morals (Oxford: Oxford University Press, 1961), 86–93.
72
Scotus, Ordinatio (henceforth Ord.), III, suppl. dist. 27.
73
This is what C. Stephen Evans calls ‘the discretion thesis’, and he says that, even though
most divine command theorists hold it, it is not necessary for divine command theory; see God
and Moral Obligation, 34–5. I am not trying, however, to give conditions for a theory’s being a
divine command theory. The theory I am defending does hold that God has this discretion.
24 God’s Command
neglecting others, for the goods central for one person at one time may be
different from those that are central for another person, or for the same person
at a different time.74 The two priority relations are the opposite way round, but
there is nothing contradictory in this, because we have two different kinds of
priority. The first kind of priority is what Aristotle calls priority ‘in account’,
and the second is something more like ‘veto’ or ‘overridingness’. The good has
priority to the right because everything that is right is good, though not vice
versa. The right has priority to the good because the goods that God selects as
mandatory for us are, so to speak, trumps.
In this account of the two different priority relations we can see already a
reply to one typical objection to divine command theory, that it makes
morality arbitrary. This objection is sometimes tied to Plato’s account in
the Euthyphro of Socrates’ question ‘Is the holy holy because it is loved by the
gods, or do they love it because it is holy?’75 Socrates is clear that the answer
to this question is the second alternative, that the gods love the holy because
it is holy. But this answer has seemed to many philosophers to be fatal to
divine command theory. If the gods love the holy because it is already holy,
then they do not produce its holiness by loving it. I have elsewhere tried to
show that Socrates does not actually give an argument for the key premiss
(at 10d4) that the gods love the holy because it is holy.76 He simply gets
Euthyphro to agree to this without an argument, and then he proceeds with
an elaborate argument that from this premiss we can see that the holy and
the god-loved are not the same. If we try to find in the dialogue an argument
for the key premiss, the best candidate is from the earlier discussion about
why the gods fight with each other. Socrates and Euthyphro agree (at 8e4)
that dispute among the gods (and also between humans) is because some see
justice in an action and some see injustice. There is an implicit appeal here to
some kind of Theory of Forms. But the dialogue is weak support for the
opponent of divine command theory, because the opponent needs an explicit
argument that the gods love the holy because it is holy, and the dialogue does
not in fact provide one.
Socrates has a truth here, however, but one that is consistent with divine
command theory. We know that God’s commands are not arbitrary, because
we know that what God commands is good, and the goodness is not produced

74
See Sren Kierkegaard, Works of Love, ed. and trans. Howard V. Hong and Edna H. Hong
(Princeton: Princeton University Press, 1998), 230. Robert M. Adams has a good discussion of
this in Finite and Infinite Goods, 292–3, in relation to Bonhoeffer and vocation.
75
Plato, Euthyphro 10d–f.
76
See my Plato’s Euthyphro (Bryn Mawr, PA: Bryn Mawr Commentaries, 1985). The term
‘arbitrary’ has different senses. It can have a negative evaluative force, meaning that a decision
ignores some relevant consideration, and this is how the term is used in the objection from
arbitrariness to divine command theory. But it can also mean that a decision is within the agent’s
arbitrium or discretion, and in that sense I want to agree that God’s commands are arbitrary.
 Morality and Religion 25
by the command. This does not, however, make God’s command redundant,
because only those good things that are commanded are obligatory.77

1 . 4. G O D’S COMMAND AND THE SCOPE

OF OBLIGATION

The justification of moral obligation by God’s command is more intimate than

has yet been explained. God’s command after all produces not only moral
obligation, but obligations of other kinds; in Judaism, for example, ceremonial
and dietary obligations; in Christianity, obligations about baptism and Euchar-
ist; in Islam, obligations about pilgrimage and daily prayer. But with moral
obligation, we might say that God’s command not only lays the obligation upon
us, but also gives us the scope of the obligation (in that we have an obligation to
respect all those who receive the command.)78 This needs more explanation.
Kantian morality requires that we give equal moral status, or dignity (as
opposed to price), to all human beings. But it has proved hard to justify this
status. We can start with some brief remarks about Kant’s own view, though
this is not a book about Kant. But we can then go beyond Kant and locate
human dignity in our call by God, where a call is a kind of prescription.79 If
this is right, then divine prescription will not merely give us a justification for
the claim that we are under obligation, but it will ground the particular kind of
obligation that is peculiar to morality.
Kant scholars disagree about how Kant grounds his views about human
moral status. He says in the Groundwork, in the second formulation of the
Categorical Imperative given above, that we are to treat humanity, whether in
our own person or the person of any other, always at the same time as an end,
and never merely as a means.80 Moreover: ‘Morality is the condition under
which alone a rational being can be an end in itself, since only through this is it
possible to be a lawgiving member in the kingdom of ends. Hence morality,
and humanity insofar as it is capable of morality, is that which alone has
dignity.’81 But there are different schools of interpretation of this point.82 One

77
This solution to the ‘Euthyphro problem’ is already worked out in J. L. Mackie, Ethics:
Inventing Right and Wrong (Harmondsworth: Penguin, 1977) (henceforth E.), 229–32.
78
I am not here discussing the question about whether we also have moral obligations to non-
human animals. I think that we do, but I am here interested in defending the claim that we have
moral obligations to all human beings. I am assuming, not arguing for, the premiss that we need
an account of dignity that has all humans in its scope.
79
I will describe what kind of prescription in Section 2.1.5.
80 81
Gl. iv. 429. Gl. iv. 435.
82
I am following here Patrick Kain, ‘Kant’s Defense of Human Moral Status’, Journal of the
History of Philosophy, 47/1 (2009), 59–102. He identifies as members of the first group Otfried
Höffe, Reinhard Brandt, and Ludwig Siep. In the second group he places Tom Regan, Allen
26 God’s Command
group takes an inclusive interpretation, according to which only persons have
moral status, all human beings have moral status, and therefore all human
beings are persons. The problem with this is that Kant’s criterion for person-
ality (‘the susceptibility to respect for the moral law as of itself a sufficient
incentive to the will’) seems to rule out some human beings, such as two-
month-olds.83 A second group holds that, on Kant’s criterion for personhood,
many human beings, including normal human infants but also adults with
Alzheimer’s disease, must lack moral status. This would be true even if Kant
held that humanity (the capacity for rationally setting ends) was the basis
rather than personality (the capacity for respect for the moral law). On this
interpretation, then, Kant ends up disallowing important subgroups of human
beings from having moral status.
If we take Kant’s language about the predisposition to the good seriously,
however, we have a partial answer to this difficulty. In his biological and
psychological writing, ‘the characteristic features and capacities of members of
each species, including their capacity to breed with one another, are under-
stood in terms of an underlying common specific nature, understood as a set of
“predispositions” and “germs” or “seeds”, that each member of a given species
shares as a result of its origin’.84 The predisposition to the good cannot be,
strictly, biological, given Kant’s distinction between nature and freedom. But
the language of ‘predisposition’ suggests strongly that he has reference to the
species in mind. This is why he can say in Religion that the predisposition to
good is essential to us and the propensity to evil is not. This would explain why
in The Metaphysics of Morals he says:
Children, as persons, have from procreation [aus der Zeugung] an original innate
(not acquired) right to the care of their parents until they are able to look after
themselves, and they have this right directly by law (lege), that is, without any
special act being required to establish this right. For the offspring is a person, and
it is impossible to form a concept of the production of a being endowed with
freedom through a physical operation. So from a practical point of view it is a
quite correct and even necessary idea to regard the act of procreation as one by
which we have brought a person over into the world without his consent and on
our own initiative, for which deed the parents incur an obligation to make the
child content with his condition so far as they can.85

Wood, Jeff McMahan, Mary Anne Warren, and Reinhard Merkel. A third group holds that
Kant’s decision to include all humans was a pragmatic decision, responsive to the moral ideal of
promoting a ‘kingdom of ends’, not a decision itself determined by theoretical or empirical
investigation. But this seems to beg the question about what the grounding for membership in
the kingdom of ends is supposed to be. Kain gives Christine Korsgaard as an example of the
third group.
83
Rel. vi. 28.
84
Kain, ‘Kant’s Defense of Human Moral Status’, 75. 85
MdS vi. 280–1.
 Morality and Religion 27
Kant holds that we have obligations only to persons, and he here commits
himself to the view that humans are persons from conception. This means
that what makes something a person is not the manifestation of respect for
the law, nor even what Aristotle calls a ‘second potentiality’ for this, such as
the capacity to think about something I know even when I am not at the
moment thinking about it. Rather, it is membership in a species in which
some members have this kind of second potentiality for responding to the
moral law.
If this is Kant’s view, he can overcome some of the objections already
mentioned. Two-month-old infants and adults with Alzheimer’s disease
belong to the human species, and so have moral status. There is a difficulty,
however. It is unclear why we should give status to members of a species who
do not themselves have the relevant capacities (the second potentialities), for
example, infants born with severe mental retardation, if it is the existence of
just those capacities in some of its members that is supposed to make the
species valuable in the particular way that moral status implies. I myself do not
see how to overcome this difficulty. Are we then left without a good way to
ground human dignity? Within the Abrahamic faiths we do have a way to do
this, starting from the premiss that humans are created in the image of God.86
Two ways of understanding this are not successful, but there is a third that
works, and that takes us back to divine command.
The passages that mention the image of God tell us very little about what
this image in a human being amounts to. Speculation has been continuous and
manifold, referring to our rationality or our freedom or our capacity for
dominion or our capacity for relation (as between male and female). The
problem with all these accounts is that they are based on the capacities we can
exercise in this life. This is the first way that is unsuccessful, because it is not
clear how any such capacity-based account can cover all human beings and
give them the same basic dignity.87 Consider, for example, the capacity for
dominion. Even if we take this to mean something like ‘stewardship’ rather
than ‘mastery’, there are many humans who do not have any significant
capacity to look after or steward creation.

86
There are two relevant passages from Genesis. According to Genesis 1: 26–7: ‘Then God
said, “Let us make humankind in our image, according to our likeness; and let them have
dominion over the fish of the sea, and over the birds of the air, and over the cattle, and over all
the wild animals of the earth, and over every creeping thing that creeps upon the earth.” So God
created humankind in his image, in the image of God he created them; male and female he
created them.’ The other passage is Genesis 9: 6: ‘Whoever sheds the blood of a human, by a
human shall that person’s blood be shed; for in his own image God made humankind.’
Maimonides discussed this idea in the first chapter of The Guide of the Perplexed, trans. Shlomo
Pines (Chicago: Chicago University Press, 1963) (henceforth Guide).
87
See Wolterstorff, Justice: Rights and Wrongs, 342–61. However, Wolterstorff takes a
different view of Kant, following Allen Wood, and he also proposes that we can get dignity
from ‘conferred value’, and I am going to deny this.
28 God’s Command
One response to this point is to look for a theistic account of the basis of
human dignity not in human capacities but in God’s activity of conferring or
bestowing value. This is the second way that is unsuccessful. Nicholas Wolter-
storff says:
What we need, for a theistic grounding of natural human rights, is some worth-
imparting relation of human beings to God that does not in any way involve a
reference to human capacities. I will argue that being loved by God is such a
relation; being loved by God gives a human being great worth. And, if God loves
equally and permanently each and every creature who bears the imago dei, then
the relational property of being loved by God is what we have been looking for.88
Wolterstorff goes on to give an analogy. My friend shows me a particularly
decrepit stuffed animal, a rabbit, but tells me that this is the animal loved by
his son, Nathan. Nathan ‘may acknowledge that lots of others are “nicer”. But
this is the one he loves, not any of those. This is the one he is attached to; this is
the one he is bonded with.’89 Wolterstorff suggests that God loves every
human being equally and permanently with the love of attachment, and that
this is just what respect for human worth requires.
There is a problem with this account, however. We want an account of
human value that makes it intrinsic to us.90 In the Genesis account, God
created over the six days, and after each day God looked at what had been
created and ‘saw that it was good’. After the creation of human beings, God
saw everything that God had made, ‘and indeed, it was very good’.91 God is
portrayed here not as reflecting upon the divine attachment, but as seeing
something good in the created order, and especially in the human life. Paul
Weithman makes this objection to Wolterstorff ’s analysis, focusing on the
analogy of the stuffed rabbit. If an adult abused the stuffed animal, she ‘would
do something very hurtful to Nathan. In performing the act, the adult would
be failing to give appropriate consideration to Nathan’s love for the rabbit and
to the feelings to which that love makes Nathan liable. In performing the act,
the adult would be under-respecting Nathan, and failing to value Nathan
highly enough.’92 But would the adult be failing to respect the stuffed animal
itself? It is hard to see that she would be. The analysis of human value as
imparted value makes this value too transparent, as though we see through it
to God’s value without any value added. A successful theistic account of
human value needs to accommodate both the relation to God, who is the
ultimate source of all value, and the intrinsic value of what God creates.93

88 89
Wolterstorff, Justice: Rights and Wrongs, 352. Ibid. 359.
90
For a discussion of what ‘intrinsic’ means here, see Section 6.2.1.
91
Genesis 1: 31 (emphasis added). In the Septuagint, kai idou kala lian.
92
Paul Weithman, ‘God’s Velveteen Rabbit’, Journal of Religious Ethics, 37/2 (2009), 256.
93
See Wolterstorff ’s response to this objection in ‘Justice as Inherent Rights: A Response to
my Commentators’, Journal of Religious Ethics, 37/ 2 (2009), 274–5.
 Morality and Religion 29
There is an account that meets these conditions. I take it from David Kelsey,
but it descends from Karl Barth.94 Kelsey writes:
Human beings’ inherent accountability for their response to God provides the
theological basis on which the peculiar dignity of human creatures is to be
understood. Dignity inheres in human creatures’ concrete actuality by virtue of
the fact that the triune God has directly related to them as their creator. . . .
Human dignity is thus ex-centric, grounded and centered outside human
creatures.95
But if dignity is centred outside human creatures, how can it be intrinsic to the
human creature? Kelsey asks what is the justification for ascribing this kind of
value to human beings, and he answers that the justification is not from our
capacities, but from God’s calling us to a certain vocation. This calling is
particular, different for each concrete human person. In this way the ground is
not something abstract or universal, like Kant’s ‘personality’. There is a good
reply here to the charge that locating the ground of human value in God’s
attachment to us makes our value extrinsic. On the conception defended in the
rest of the book, there is a call by God to each one of us, a call to love God in a
particular and unique way. Revelation 2: 17, in the instructions to the church
in Pergamos, refers to a name about which God says, ‘and [I] will give him a
white stone, and in the stone a new name written, which no one knows except
the one that receives it’. If we think of this name, like ‘Peter’ meaning ‘rock’
(the name Jesus gives to Simon), as giving us the nature into which we are
being called, and if we think of this nature, as Scotus does, as a way of loving
God, then we can think of the value of each of us as residing in us, in our
particular relation to God. What we have here is an intrinsic good in a slightly
odd sense; not that we have value, each of us, all by ourselves (which is one
thing the phrase ‘intrinsic value’ might mean), since we have our value in
relation.96 But the value is not reducible to the valuing by someone outside us,
on this account, but resides in what each of us can uniquely be in relation
to God.
Is dignity based on the call of God, or in the destination towards which God
is calling us? Dignity in this discussion is incommensurable worth. This is how
Kant distinguishes dignity from price, such as the price of a pen that can be
exchanged for something else of commensurable or equivalent worth.97 But
this means that we can return to an argument at the beginning of this section.
If we grant that God has incommensurable worth, we should grant that, other

94
See Gerald McKenny, The Analogy of Grace (Oxford: Oxford University Press, 2010), 229,
on Barth’s notion of responsiveness (Verantwortlichkeit): ‘The decision of God on our conduct
gives our conduct the character of a reply or response to the question posed to us by and in the
command of God.’
95
David Kelsey, Eccentric Existence (Louisville, KY: John Knox Press, 2009), 275–7.
96 97
Moore, Principia Ethica, 187. See Section 6.2.1. Gl. iv. 434.
30 God’s Command
things being equal, the love by a particular person of God (which is her
destination) shares in that worth, though it will not be of the same value.
And what leads her to that love (namely, God’s call) will also have value. The
idea that this love is of incommensurable value is suggested by texts like, ‘Or
what will they give in return for their life?’98 To answer the question at the
beginning of the paragraph, we should say that it is the destination that gives
the final value, but it is the call that leads to the destination, and it is the call
that we already have; we are not yet there, even though it is already our
destination.
In order that this account can escape the objection to a this-world-capacity
account, we have to be able to believe that God proportions or fits the call to
each human being, and there may not be much we can recognize as cognitive
capacity in this life that is a precondition of such proportioning. We can
believe that God heals us of the impediments we encounter here to following
the divine call, and that many of the conditions that we assume are impedi-
ments will turn out to be paths to a unique way of loving God. To non-theists
this belief will no doubt seem merely an attempt to escape from harsh reality.
The beginning of Section 1.3 claimed that this theist reply to the request for
a justification was an indirect use of Kant’s view that our dignity resides in our
potential to respond to the moral law. What kind of use of Kant is this? Kant’s
view throughout his published corpus is that we should recognize our duties as
God’s commands. But in Religion he undertakes a translation project as a
philosophical theologian, translating historical revelation (given to particular
people at particular times) into the revelation to reason (available at least
potentially to all people at all times), using the moral concepts. On my view, he
means, not to reduce historical revelation to the revelation to reason, but to
leave what he calls ‘biblical theology’ as it is.99 In the course of this translation,
he proposes to talk about the Trinity in relation to the problem of our falling
short of the life we ought to lead, the problem of the moral gap: ‘The law says:
“Be ye holy (in the conduct of your lives) as your Father in Heaven is holy,”’
and Kant translates this as the requirement of ‘the conformity of the conduct
of one’s life to the holiness of the law’.100 With this translation in mind, we can
see Kant’s language about our dignity residing in our potential to respond to
the moral law as a translation of more traditional language about our potential
to respond to God’s command or call. Kant indeed is willing to use the
language of ‘call’, as when he talks of ‘the call of human beings to be citizens
of an ethical state’, though he insists that we do not understand how beings
can be both created and free.101 Kant does not have the idea of the particularity
of the call. He does, however, have the idea that what gives us our dignity is
our potential to respond, and not our actual response. As argued earlier, he ties

98 99 100 101
Matthew 16: 26. Rel. vi. 9. Rel. vi. 66. Rel. vi. 142–3.
 Morality and Religion 31
this potential to our membership in the human species. The basic idea of
locating our dignity in our potential to respond to God’s call is already in Kant,
and is part of his inheritance from the Lutheran catechisms of his youth. We
get valuable help in answering the normative question by returning to the pre-
translated version that he does not discuss, but takes for granted.
To sum up, briefly, what was done in this chapter. Three arguments were
distinguished for the dependence of morality upon religion (the argument
from providence, the argument from grace, and the argument from justifica-
tion), and the chapter spent most time on the third. The justification of
obligation, that it is obedience to God’s commands, was shown, if Scotus is
right, not to rely on a basic premiss that itself requires justification. The
justification also does not fall prey to the Euthyphro objection, if we make a
separation between the good and the obligatory in the way suggested. Finally,
this justification gives us a way to ground the basic Kantian morality (that
gives the same dignity to every human being) in the notion of a unique call by
God to each individual to love God in a unique way.
 2

What is a Divine Command?

INTRODUCTION

What kind of thing is a divine command, and so, by way of introduction to this
first question, what kind of thing is a command? The term ‘divine command’ is
the standard term in the literature. But God addresses us in all sorts of ways
that are, in a broad sense, prescriptive, but are not, in a narrow sense,
commands. I will be relying here on work on the nature of prescribing that
was done fifty years ago by RMH, though I will disagree with him at some
points. In applying the notion of prescription to God’s addresses to us (a topic
RMH did not discuss), some scholastic distinctions are useful, conveniently
summarized in a Latin dactylic hexameter, ‘Praecipit et prohibet, permittit,
consulit, implet’. This was a mnemonic device (in the metre of Homer and
Virgil) for remembering the varieties of God’s revealed will.1 The revealed will
is distinguished from the disposing will, which is what God actually brings
about; God does not always bring about what the revealed divine prescriptions
would lead us to expect. Praecipit means ‘gives precepts to’. Prohibet means
‘prohibits’. Permittit means ‘permits’. Consulit means ‘counsels’. Implet means
‘fulfils’. We can call these five types of prescriptions ‘precepts’, ‘prohibitions’,
‘permissions’, ‘counsels’, and ‘directly effective commands’.2
There is an antecedent question that goes beyond the ambitions of this
book. Can God use human language at all? An influential critique of the very
notion of divine revelation holds that ‘violence belongs structurally to, indeed

1
See Ockham, Ordinatio dist. 46, q. 1, B: ‘It is commonly said that the will of God is twofold—
viz., the disposing will [voluntas beneplaciti, what God is pleased to bring about] and the revealed
will [voluntas signi, what God has made known of his will]. [The revealed will] is distinguished
into these five: prohibition, precept, counsel, fulfilment, and permission.’ Later in the same
question Ockham calls the revealed will a kind of ‘antecedent will’ (that is, a divine will
antecedent to our willing) and the disposing will a kind of ‘consequent will’. We return to this
distinction in discussing divine permissions in Section 2.1.4, and in discussing al-Maturidi in
Section 6.2.3.
2
This is intended not as an exhaustive list, but as a summary of the main headings.
 What is a Divine Command? 33
constitutes, language itself ’.3 The problem is supposed to be that our language
seeks to fit objects within previously established categories, and thus does
violence to them. Perhaps using a violent language is inconsistent with an
essentially loving God. But the premiss is doubtful, and the conclusion does
not follow from it.4 Sometimes things do fit concepts, and, even if they do not,
God could no doubt work in the Spirit to purge our usage by bringing to our
attention (in a way we could resist) a meaning that does not violate the
integrity of what is talked about. But this is a topic too broad for the project
of this book.

2.1. FIVE TYPES OF PRESCRIPTION

2.1.1. Prescription in General

We start with prescription as a genus, labelling the contrast term, somewhat

artificially, as ‘description’. The contrast can be explained using Elizabeth
Anscombe’s articulation of the distinction between wanting and belief:
For you cannot explain truth without introducing as its subject intellect, or
judgement, or propositions, in some relation of which to the things known or
judged truth consists; ‘truth’ is ascribed to what has the relation, not to the things.
With ‘good’ and ‘wanting’ it is the other way round; as we have seen, an account
of ‘wanting’ introduced good as its object, and goodness of one sort or another is
ascribed primarily to the objects, not to the wanting.5
Note that, in both cases, both the wanting and the believing, there is a possible
relation of fit with the world. The term ‘desire’ is convenient as a genus term
for the mental state expressed in a prescription, although it is not completely
appropriate for this role.6 A desire is, we can say, satisfied when the world
comes to be in conformity to it. A belief is true when it is in conformity to
the world. The same distinction can be made between prescriptions and

3
John D. Caputo, ‘How to Avoid Speaking of God: The Violence of Natural Theology,’ in
Eugene Thomas Long (ed.), Prospects for Natural Theology (Washington: Catholic University of
America Press, 1992), 142.
4
A strong argument to both these results is given in Kevin Hector, Theology without
Metaphysics: God, Language and the Spirit of Recognition (Cambridge: Cambridge University
Press, 2011), esp. ch. 1.
5
G. E. M. Anscombe, Intention (Oxford: Blackwell, 1957), para. 40. Michael Smith explains
the idea of ‘direction of fit’ in terms of different kinds of counterfactuals that are true of the
subject who believes and desires. See The Moral Problem (Oxford: Blackwell, 1994), 111–25.
6
Attic Greek has the term orexis here, but there is no good English translation. Aristotle
distinguishes ‘desire’ in a narrower sense (in Greek epithumia) from, for example, rational wish
(in Greek boulesis), aspiration, fear, and love of risk. But in what follows I use ‘desire’ broadly to
cover the whole family.
34 God’s Command
descriptions. A prescription is satisfied when what is prescribed occurs.
A description is true when what is described obtains. But, though both kinds
of mental act have their own peculiar kind of fit with the world, Anscombe
points out there is a difference between them. If the fit fails, we might say, in
the case of prescription or desire, it is a problem for the world; in the case of
description, it is a problem for the mental act or belief. Or we could put this
in Anscombe’s terms: for descriptions, the term ‘true’ (which indicates the
proper fit) marks primarily the mental act or its content; for prescriptions,
(when fitting) the term ‘good’ marks primarily the object.
It is no accident that this difference between prescription and description
mirrors the difference Anscombe draws between desire and belief. We have
prescription in our language as a family of speech acts because we want to be
able to express desires or will in a certain way. This is different from what
seems to be the view of Aquinas that commanding (in Latin, imperare) is an
act of the reason but not of the will, though the will can receive the command.7
To command, on this conception, ‘is to set in motion, not anyhow, but with an
emphatic declaration pressing in on another, and this is an act of reason’.8
There is a complex and unfamiliar conception here of the relation of intellect
and will. Writers after Aquinas, such as Scotus, often use a more robust notion
of the will, like the one described in Section 3.2.2, and it becomes more natural
to think of the will being the locus of commanding. In Scotus, the will already
contains a ranking of the affection for justice and the affection for advantage,
where the affection for advantage is an inclination towards one’s own happi-
ness and perfection, and the affection for justice is an inclination towards what
is good in itself, independently of any relation to oneself. The more robust
conception of the will can be found even in later Thomists such as Suarez.9
Command as a relation of divine and human wills is the topic of the final
section of this chapter.
We have imperatives because we want to be able to express our desires. But
this does not mean that we want to be able to say that we have the desires; that
is simply another form of description. We want, rather, to be able to effect
change in the world, to make it fit our desire, by communicating the desire. To
say this is also not to give what John Austin called the ‘perlocutionary force’ of
an utterance, namely, what the speaker is intending to bring about through
speaking.10 Someone who gives a command may intend to bring it about that

7
ST I–II. 17. 5, ad 3. But Aquinas is not completely consistent about this. See ST I–II. 1. 1,
ad 2, ‘as being commanded by the will’.
8
ST I–II. 17.1, ad 1.
9
God wills to ‘bind’ his subjects by commanding them. See Francisco Suarez, A Treatise on
Laws and God the Lawgiver, trans. Gwladys L. Williams, Ammi Brown, and John Waldron, in
Selections from Three Works of Francisco Suarez, ii (Oxford: Clarendon Press, 1944), 55.
10
Intending to produce an obligation through a command needs to be distinguished here
from intending to produce the action that fulfils the obligation. I am differing here from Paul
 What is a Divine Command? 35
her audience does the thing commanded, but this is not necessary. RMH gave
the example of a person who knows that the child to whom she gives the
command is counter-suggestible, and who therefore intends the child to do
the opposite of what she commands. A different sort of example comes from
the biblical story of Abraham and his son. God commands Abraham to kill his
son, but God does not intend to bring it about that Abraham obeys this
command. So it is not necessary, in order to have a fully functioning com-
mand, that the person giving the command intend that the addressee carry it
out. Nonetheless, these are outlier cases, and we have imperatives as a gram-
matical form in our language because we want to be able to effect changes in
the world through the communication of our desires.
RMH divided imperatives into singular and universal, and he held that
moral prescriptions are universalizable in the sense that they imply universal
imperatives. The difference between singular and universal imperatives can be
made clearer by thinking of an imperative sentence as having different pos-
itions within it, of which we can separate out four: the positions of addressee,
subject, action, and recipient.11 A singular imperative sentence is one in which
the term in at least one of the positions is singular. A universal imperative
sentence is one in which the terms in all the positions, including the subject
position, are universal. Proverbs are often like this, but they leave the term in
the subject position unspecified, as in the proverb ‘Don’t spoil the ship for a
ha’porth of tar’. The distinction between the positions is clearest in an
example: ‘John, Andrew is to take the package tomorrow to Sarah.’ (a) The
addressee is the person to whom the imperative is addressed. The example is
deliberately chosen to show that the addressee (John) may not be the same as
the person who is commanded to do the action (Andrew).12 Sometimes, the
addressee is, by implication, supposed then to communicate the command to
the person in subject position.13 (b) The subject is the person commanded to
do the action. Note that the example is a third-person imperative. We are
more familiar with second-person imperatives like ‘John, take the package to
Sarah’. An example of a third-person imperative is God’s command at the
beginning of Genesis: ‘Let there be light.’ This is what I called a ‘directly
effective command’, where God effects the result just by commanding it. It is a
prescription in the broad sense, but not a command in the narrow sense in
which commands generate obligation. (c) The action is what is commanded.

Grice, ‘Meaning’, repr. in Studies in the Way of Words (Cambridge, MA: Harvard University
Press, 1989), 219–21, discussed in Adams, Finite and Infinite Goods, 267.
11
See my The Moral Gap, ch. 6.
12
RMH went wrong by not making this distinction, since he suggested that universal
commands are addressed to everyone. See John Cronquist, ‘Hare and Prescriptivism’, disserta-
tion, Stanford University, 1972, 55.
13
Joshua 6: 2–4: ‘And the Lord said unto Joshua, “. . . And seven priests shall bear before the
ark seven trumpets of rams’ horns.” ’
36 God’s Command
Very often there will be multiple sub-positions within the action position (in
the example, ‘to take’ and ‘tomorrow’). (d) The recipient is the person or object
to whom the action is commanded to be done (in the example, there is a direct
object, the package, and an indirect object, Sarah).
The example of the Ten Commandments raises a complex exegetical and
theological question about whether the prescriptions to ‘you’ (for example,
‘You shall not murder’) are just to the people of Israel, or to all human beings.
Perhaps the right thing to say is that they are initially given to the people of
Israel (both in addressee position and in subject position), as part of the
covenant (see Deuteronomy 9: 4–6, Exodus 34: 28), but that they are eventu-
ally intended by God to be commandments given to all human beings (both in
addressee position and in subject position). If this is right, we have an example
of what was a singular imperative becoming a universal imperative. ‘The
people of Israel’ is a singular term because it makes reference to a particular
region of space and time, even though there are many human beings included
in the reference of the term. Chapter 5, in the discussion of particularity in
divine commands, returns to the distinction between the different positions
within an imperative sentence, and argues that, for a command to produce a
moral obligation, it is necessary that at least the term in the action position
be universalizable, but that this is not necessary for the terms in the other
positions.
Ordinary logical relations such as entailment and negation apply to pre-
scriptions as much as descriptions. RMH tried to show this by making another
distinction, between what he called ‘phrastic’, ‘tropic’, and ‘neustic’.14 The
phrastic (from the Greek phrazo, meaning ‘I tell’ or ‘I counsel’) is the content
common to the command ‘Andrew is to take the package tomorrow to Sarah’
and the prediction ‘Andrew will take the package tomorrow to Sarah’. We can
say that this content is the state of affairs of Andrew’s taking the package
tomorrow to Sarah. The tropic (from the Greek tropos meaning ‘mood’) is the
mood indicator, distinguishing indicative (in America, ‘declarative’) from
imperative and thus distinguishing statement from prescription. The various
speech acts we are considering in this section all have the imperative tropic. The
neustic (from the Greek neuo, meaning ‘I nod assent’) is the sign of assent to the
combination of phrastic and tropic. When the teacher says in a class, as a
philosophical example, ‘The cat is on the mat’, she withholds the neustic. If the
phrastic picks out no identifiable state of affairs, both statement and command
are meaningless. Using this new vocabulary, we can say that the phrastic can be
the bearer of the same logical relations in both imperative and indicative
sentences, and we can talk about ‘satisfaction conditions’ for imperatives in
the same way that we talk about ‘truth conditions’ for indicatives. For example,

14
In The Language of Morals, RMH does not distinguish ‘tropic’ and ‘neustic’, but he does so
in ‘Meaning and Speech Acts’, in Practical Inferences (London: Macmillan, 1971), 74–9.
 What is a Divine Command? 37
if I command Andrew to take the package tomorrow to Sarah, I am command-
ing him not to not take it. If I state that he will take it, I am stating that it is not
the case that he will not take it. We return to the neustic in Section 4.3.1.
The example of a command is only one kind of use of an imperative.15
Sometimes ‘command’ is taken for the sake of convenience to cover a whole
family of speech acts, but this chapter uses ‘prescription’ as the name for the
family, and reserves ‘command’ for the narrower sense in which commands
generate obligation (and, by extension, ‘directly effective commands’). Since
there is a whole family here, it is worth going through some of the differences,
and working out an account of what commanding in the narrower sense
involves. RMH distinguished ‘military orders, architectural specifications,
instructions for cooking omelets or operating vacuum cleaners, pieces of
advice, requests, and entreaties, et al.’.16 Some of the other members of the
family of prescription are admonitions, exhortations, warnings, invitations,
and calls. There is no need for present purposes to divide the whole family into
species by genus and difference. But it is important to distinguish certain
members of the family from each other, since God can do many of them, and
what God is doing in one case is different from what God is doing in another.
In all of them, however, the use of the imperative suggests that God wants to
effect change in the world by the communication of the divine will.

2.1.2. Precepts

The scholastics distinguished between five different forms of God’s revealed

will. The first of these was precepts (praecipit). In a broad sense, precepts all
tell people to do something. Precepts in this sense include warning, admon-
ishment, and exhortation, as well as other kinds. For present purposes we can
focus on just one kind of precept: commands that generate obligation. In
traditional Roman Catholic moral theology, a precept is universal; it is neces-
sary for all to obey, and disobedience carries with it eternal damnation. But it

15
F. R. Palmer, in Mood and Modality (Cambridge: Cambridge University Press, 1986), 23–6,
distinguishes within the ‘imperative’ sentence type, three types of illocutionary force, ‘directive’,
‘commissive’, and ‘expressive’. There are twelve sub-types of modal function under ‘directive’,
including ‘compulsive’ (‘he has to go’), ‘directive’ (‘he must go’), ‘prescriptive’ (‘es soll gehen’),
‘advisory’ (‘you should go’), ‘permissive’ (‘he may go’), ‘precative’ (‘Go, please’), and ‘hortative’
(‘let us go’). There are two sub-types under ‘commissive’, including ‘promissive’ (‘I will go’) and
‘threat’ (‘I will kill you’). There are two sub-types under ‘expressive’, namely, ‘volitive’ and
‘evaluative’, and four sub-types under ‘volitive’, including ‘optative’ (‘May he still be alive!’)
and ‘desiderative’ (‘Would he were alive!’), and four sub-types under ‘evaluative’, including
predictions/warning, positive doubt/scepticism, surprise and regret. For an interesting use of
these categories of speech act to distinguish types of God’s speech to us, see Andy Warren,
‘Modality, Reference and Speech Acts in the Psalms’, dissertation, Cambridge University, 1998.
16
RMH, The Language of Morals, 4.
38 God’s Command
is better to say that divine commands that generate obligation can be singular,
and, although they contain internal references to God’s authority and some
kind of divine condemnation or chastening following disobedience, this is not
necessarily eternal damnation.
We can start with some examples of speech acts using imperatives that are
precepts in the broad sense, but are not commands that generate obligation.
Take the case of admonishing. 2 Thessalonians 3: 15, in the King James
Version (henceforth, KJV), reads: ‘Count him not as an enemy, but admonish
him as a brother.’ Brothers do not, in any straightforward way, have authority
over each other simply as brothers. This is certainly true also of warning,
which is another way to translate the same text (noutheteite). Someone can
warn me of a danger (‘Watch out for the curb’) without having any authority
over me. In some cases of admonishing and warning, there is authority
presupposed, and in some cases not.17 The same is true with exhorting.
There is usually no presupposition of authority in the speaker. Books of a
certain kind by political scientists (‘musty’ books?) are full of exhortations
about what ‘must’ be done by the political authorities. Concerned citizens can
write exhortations in letters to the press. Titus 2: 6 tells Titus to exhort
(parakalein) the young men to be self-controlled, and then as an addition
tells him (2: 15) to do it with all authority (meta pases epitages), and not to let
anyone despise him. It is false that all uses of the imperative to tell someone to
do something (precepts in the broad sense) have internal reference to the
authority of the speaker. Consider, ‘Shut the door, please. It’s freezing in here’,
called out to my friend who is about to leave the front door open. Or consider
our imperatives addressed to God: ‘O Lord God of hosts, hear my prayer: give
ear, O God of Jacob’ (Psalm 84: 8). These are imperatives used for the
‘precative’ speech act of entreating. We do not have authority over God.
When God admonishes or warns or exhorts, it might seem that there must
be an internal reference to God’s authority in the meaning of the speech acts.
But this is not the case. Section 2.2 discusses what an internal reference to
authority requires. But for now the project is to distinguish commands in the
narrow sense from other prescriptions that God utters, other speech acts with
the same tropic, and then collect these differences together in order to delimit
the kinds of divine command that generate obligation.
Consider the remarkable passage (Deuteronomy 29–30) in which Moses
gives his last address to the people of Israel, including the divine exhortations
that the author calls ‘the words of the covenant’.18 The address culminates in

17
The Oxford English Dictionary gives, as just one sense of ‘admonish’, ‘to charge (a person)
authoritatively’, and gives as an example Caxton, Chronicles of England (1520): ‘We admonish
you first in the Pope’s behalf that ye make full restitution.’
18
The phrase ‘divine exhortations’ is Walter Raleigh’s term for what Moses gives to the
people of Israel: History of the World, ii (London: Walter Burre, 1614), 326.
 What is a Divine Command? 39
the prescription: ‘Choose life, so that you and your descendants may live’
(Deuteronomy 30: 19). Is this prescription a command? The passage suggests
not, for God says that, if they choose life, they will then love the Lord and obey
his voice and hold fast to him. The passage supports the relation claimed in
Chapter 1 between loving and obeying God. But the present point is that
entering into the relation of hearing command and obeying it comes as a
consequence of the initial choosing of life. God does not present this initial
choice as one in which they are acknowledging the divine authority. God is,
properly speaking, exhorting them; setting before them two options, and
urging one of the two. The relation in which they hear the command and
obey is the goal they will obtain through this choice.
To explain this further, we need to see more clearly the relation between
command and covenant.19 Deuteronomy 28–31 is a form of covenant that is
reciprocal, in the sense that it involves both God’s promise of life and the
people’s required obedience, and it mentions punishment by God for the
people’s disobedience. Not all divine covenants are like that. For example,
the covenant after the Flood, quoted in Chapter 1 in discussing the divine
image, is self-imposed by the deity: God will not again bring upon the earth
this kind of total devastation of all flesh, and God seals this covenant with the
rainbow.20 The covenant with Noah and his sons is also a covenant with every
living creature, and does not seem to be conditional on the living creatures in
some way endorsing it. Again, the covenant with David seems to be promis-
sory, rather than reciprocal, though it includes God’s warning that, if David’s
seed commits iniquity, he will be chastened; nonetheless God says: ‘I will not
take my steadfast love from him.’21 In the case of the words of the covenant in
Deuteronomy, the commands (in chapters 12–26) precede the promise of life,
but that does not mean that obedience to them precedes the making of the
covenant. We can construe the relation between the covenant and the com-
mands by analogy with the relation between love and obedience described by
Scotus. To love God requires us to repeat in our wills God’s will for our willing,
and such a repetition is obedience. In the same way, entering the covenant
is entering into a relation that is expressed, on our side, by obedience. God
becomes our God and we become God’s people. If this is right, it means
that the divine commands that have internal reference by the meaning of
the speech act to God’s authority may be within a covenant; but they also
may not be. Covenants with a particular people are not the only way that
God’s commands generate obligation. Section 2.2 discusses the idea of God’s

19
I am grateful to Andrew Forsyth for bringing some of this material to my consideration, in
particular George E. Mendenhall, Law and Covenant in Israel and the Ancient Near East
(Pittsburgh: Biblical Colloquium, 1955), and Moshe Weinfeld, ‘The Covenant of Grant in the
Old Testament and in the Ancient Near East’, Journal of the American Oriental Society, 90/2
(1970), 184–203.
20 21
Genesis 9: 8–17. 2 Samuel 7: 12–17.
40 God’s Command
authority directly, and claims that all people are under this authority, whether
they recognize it or not.
There are other imperative sentences, ascribed to Jesus, that are not com-
mands generating obligation. Jesus says (Matthew 11: 28, KJV): ‘Come unto
me, all ye that labour and are heavy laden, and I will give you rest.’ Is this a
command? Surely it is more like an invitation. The same is true of much of the
biblical language of ‘call’, which belongs with counsel rather than command.22
We will return to the topic of call in Chapter 5, in relation to Barth, who says,
‘the vocation of man is always in fact the terminus a quo of his obedience’.23
That is, the call is the point starting from which we are obedient. The
command is to lead a life worthy of the calling (Ephesians 4: 1), but the calling
itself is not exactly a command; it is, when answered, the context of command.
I propose that we say that ‘command’ (in the narrow sense) has internal
reference to authority as part of the meaning of the speech act, with some
kind of condemnation envisaged for failure. This does not mean denying that
the God who calls us and invites us is also authoritative, or that the call and
invitation are different because of this authority. But, when God calls and
invites, the speech act is not itself, by its meaning, tied to authority or
condemnation for failure.
The discussion of these types of divine prescription has been somewhat
abstract, and for the sake of vividness each subsection of the present section
ends with a concrete example of the type it describes. The examples are all of
the ‘inner prompting’ type, because this is the type that occasions the most
scepticism, and the examples are intended to make it more plausible. But this
does not mean that this is the only way God communicates prescriptions to us.
C. Stephen Evans gives a useful list of nine ways.24 It is important to see, as
Evans emphasizes, that there is no need to posit that we always perceive a
divine command to come from a divine source, and there is no need to assume
that God always uses extraordinary means to communicate with us. Here is an
example of a singular precept (of type four on Evans’s list). A graduate student
is just getting into bed at the end of the day. He has changed into his pyjamas.

22
I wrote a book called God’s Call (Grand Rapids, MI: Eerdmans, 2001), in which I preferred
the term ‘call’ to ‘command’ for cases in which God’s love and God’s drawing us to Godself are
the focus, and not God’s authority and constraint.
23
Karl Barth, Church Dogmatics, ed. and trans. G. W. Bromiley and T. F. Torrance, 14 vols
(London: T & T Clark, 2009) (henceforth CD), III/4. 56 (274).
24
Evans, God and Moral Obligations, 39–45, gives the first way as special revelation through
Scripture, and then takes the next five from Richard Mouw, The God Who Commands (Notre
Dame, IN: University of Notre Dame Press, 1990), 8: (1) Scripture, (2) natural law, (3) the
magisterium of an ecclesiastical body, (4) specific commands of God to an individual, (5) examining
our natural inclinations, (6) listening to our conscience, (7) teaching from other humans as God’s
requirements, (8) teaching from other humans who do not recognize them as God’s requirements,
(9) human social requirements such as legal obligations, family obligations, and obligations of other
socially defined roles.
 What is a Divine Command? 41

He hears as it were a voice in his head telling him to get back into his clothes
and go and visit a friend who also lives in the quad. He can think of no reason
other than this voice to do so. But the voice carries authority. He gets dressed,
and goes to visit his friend, and discovers that his friend has just broken up
with his girlfriend and is in serious, indeed dangerous, depression, and needs
someone to talk to. Why did he take the voice as a divine voice prior to obeying
it? Here are a few pointers. First, the voice did not present itself as a construc-
tion of his own imagination. This does not show, to be sure, that it was not
such a construction. But, second, he recognized the voice as one he had
followed in the past, and it had told him the truth. Another pointer is that
the voice was unshakeable, and that it silenced the usual objections. He has
found that, when he follows it, he has a certain kind of peace. Afterwards, he
had the additional reason that he had discovered that the voice was telling him
to do something unexpected that turned out to be good. None of this is a
demonstration of the claim that it was God’s command; the mental hospitals
are full of people who could say these sorts of things about the voices they
hear. But this is a brief account of the phenomenology, and there will be
considerably more about this in Section 5.3.3.

2.1.3. Prohibitions

A prohibition (prohibet) is a command not to do something. Neither prohib-

itions nor precepts need to have imperative sentences for their expression.
‘Spitting is forbidden’ is a prohibition, but the sentence is grammatically
indicative. Imperatives are, however, a typical form of expression for both. If
I say, ‘Don’t go too close to the edge or you may fall’, I am warning, and in a
broad sense prohibiting. But there is a narrow sense in which warning and
prohibiting are different, and I suggest the difference resides in the presence
(in prohibiting) or absence (in warning) of an internal reference in the nature
of the speech act to the authority of the speaker, and to some form of
condemnation envisaged for failing to comply.
Examples of prohibitive commands will be like examples of preceptive
commands, except in the negative. But there will often be a positive command
going along with the prohibition. In experience, it is sometimes the negative
that has the focus, and sometimes the positive; perhaps with God’s commands
it is more often the negative. Socrates reported that his voice only told him what
not to do. He thought it important that the voice was silent during his conduct
of his defence, and he interpreted this as a divine invitation to leave his earthly
life. But there may be positive implications of prohibitions. The Heidelberg
Catechism, to take just one example, acknowledges that the second table of the
Ten Commandments consists almost entirely of prohibitions, but it insists that
there are positive correlates to all the negatives, and they are equally enjoined.
42 God’s Command
Examples of prohibition are easy to find, but the attribution to a divine source
will often seem indicated only when the prohibition is unexpected or unusually
vivid. This restriction comes from our own natural caution, not wanting to
ascribe to God what could be just our own mental processing. But, as with
precepts, there is no need to posit that we always perceive the divine source, or
that God always uses extraordinary means of revelation, so that God’s prohib-
itions may in fact be much more frequent than we are inclined to credit. Here is
an example of a prohibition. A person has found his sibling difficult, and decides
finally to send her a book, which he thinks will do her good. At the time of
sending it, he hears a still small voice in his head telling him that this is a bad
idea. But he ignores it, and puts the book in the mail. Somehow or other, the
book gets lost and is never delivered. At this point, the voice in his head gets
more insistent, telling him to leave well enough alone. But he is stubborn, and
buys another copy of the book, and sends it off. This time, the book is delivered,
and it is a disaster. He and his sibling have a row, after she has read the book,
which very nearly destroys their relationship. He realizes that he had in fact
known all along that God was telling him not to send it.

2.1.4. Permissions

Another kind of speech act that characteristically uses an imperative tropic is

‘permission’ (permittit).25 A permission is not a command because, if a person
is commanded, he is permitted to comply but he is not permitted not to
comply. If a person is permitted, he is permitted both to do the thing and not
to do it. God tells Adam, ‘You may freely eat of every tree of the garden: but of
the tree of the knowledge of good and evil, you shall not eat’ (Genesis 2:
16–17), following a permission with a prohibition. What is the difference? An
illuminating comparison is with necessity and possibility. God permits me if
God does not command me not to, in the same way that what is possible
is what is not necessarily not the case. But, even though the comparison is
illuminating, it is not exact. In the cases we are interested in, permission is not
simply the absence of a prohibition, or negative command. When God permits
Adam to eat of the fruit of all the other trees in the garden, God expresses
consent to this eating.26
In the cases we are interested in, there needs to be a mental act of permit-
ting, not just the absence of the mental act of prohibiting, and there needs to be

25
F. R. Palmer categorizes ‘you may go’ as a permissive modal function under ‘directive
imperative’, Mood and Modality, 23–4; see n. 15.
26 Some cases of permission are weaker than consent, but still stronger than merely the

absence of a prohibition. A book of photographs of the New Zealand locations of the Lord of the
Rings films says at the beginning, ‘This book has the permission but not the consent of the estate
of J. R. R. Tolkien’, where presumably consent has to be explicitly granted.
 What is a Divine Command? 43
a speech act. There is no doubt a sense of the verb ‘permit’ or ‘allow’, though
perhaps not of the noun ‘permission’, in which a person can, for example,
allow a whole field to be taken over by thistles just by doing nothing, not even
thinking about it. But we are interested in cases where there is a mental act,
and indeed a speech act. God’s speech acts of permission express a divine
mental act, but, to explain what mental act, we need to return to the scholastic
distinction between God’s revealed and God’s disposing will (as in the quota-
tion from Ockham at the beginning of this chapter). The revealed will is a type
of antecedent will (that is, a divine will antecedent to our willing), and the
disposing will (which actually brings something about) is a type of consequent
will. Regarding the consequent will, Ockham says that it is that ‘by which God
wills efficaciously in positing something in being’, but in antecedently willing
from eternity that a given created will should act in a certain way God does not
determine the created will to act in that way.27 This distinction between
antecedent and consequent will is useful in making the point that, even though
everything that happens is in accordance with God’s consequent permissive
will, it is not necessarily in accordance with God’s antecedent will. When
Adam ate the apple, he was frustrating God’s antecedent will, even though he
was accomplishing God’s consequent will. Most of the divine prescriptions we
have been considering should be taken as expressions of God’s antecedent will.
But now we need to make another distinction, within the antecedent divine
will. Consider tragic cases in which there does not seem to be anything good to
do.28 The classic case is the Dutch householder hiding a Jew in the Second
World War, who lies to the Nazi officer. Some ethicists have held that it is
never right to lie, even in such circumstances. It is better to say that lying in
such a case can be the least bad thing to do. But does this mean that God
permits it? If so, divine permission seems to be different from divine com-
mand as described in Section 1.3, where God’s command was said to select
which good things to require. If we allow the existence of tragic cases (as
experience seems to compel), we should say that God may command saving a
life even in a case where this requires and God permits a lie. We should then
distinguish the antecedent revealed will in the Ten Commandments, which
give prohibitions (and thus negative obligations) that are not absolute, from
the antecedent divine permission that may be revealed to a particular person
in a tragic situation. It is better to call this prescription ‘permission’ than
‘command’, because it may still be necessary to repent of the lie, even though it
was the least bad thing to do.
Here is a more ordinary example of a divine permission. A person has
been going to a church for eight years, and has been happy there, though

27
Ockham, Ordinatio dist. 46, q. 1, c.
28
There is a good discussion of this in David Baggett and Jerry L. Walls, Good God: The
Theistic Foundations of Morality (Oxford: Oxford University Press, 2011), 133–6.
44 God’s Command
recognizing that the congregation is not well integrated racially. He sometimes
makes deprecating remarks about a ‘country-club church’. Then one day he
takes an African-American friend to church, and realizes that she is the only
black person in the room, and he starts to hear the whole service with different
ears. After the friend has left, he is overcome with a sense of grief while driving
on the highway, to the point that he is unable to drive and pulls over to the
side. Then he hears as it were a voice in his head telling him that it is all right
for now to go on worshipping in that place, with those people he is fond of; but
that there are some changes in the place that need to be made. This he
interprets as a divine permission. Divine permissions are very often in situ-
ations where human defect has made a mess of things, but obedience is still
possible even though purity is not.

2.1.5. Counsels

Another speech act on the scholastic list is ‘counsel’ (consulit). As with all the
first four items on the list, we understand the speech act first in human use. In
Paradise Lost (ix. 1099), after Adam and Eve have eaten the apple, and
discovered that they were naked, Adam proposes they should both go into
the wood and hide themselves from God: ‘So counsel’d hee and both together
went | Into the thickest Wood.’ God too can counsel. God can use imperative
sentences to give us advice, instruction, or invitation. What is the difference
between command in the narrow sense and these other speech acts? The most
salient point of difference is that commands generate obligation, and there is
standardly some expectation of condemnation if the command is not carried
out.29 One clear case of this is military orders, where the condemnation will
often involve some form of punishment. With advice, this is not so, though
there may be an expectation of adverse consequences. This is a point empha-
sized by Stephen Darwall, who talks about the accountability internally con-
tained within a second-person demand, and we return to this shortly.30
Traditional Roman Catholic moral theology teaches that there are three
‘evangelical counsels’, or ‘counsels of perfection’: poverty, chastity, and obedi-
ence. This formalization into just three is doubtful. But the idea of counsels as
a separate category of divine prescription seems right. One story that raises the
issue of the nature of counsels comes in Matthew 19: 16–17, when the rich
young man asks Jesus what he should do to obtain eternal life, and Jesus tells
him to keep the commandments. When the young man presses further, Christ

29
See Aquinas, ST I–II. 90. 3, ad 2 and I–II. 108.4: ‘The difference between a counsel and a
commandment is that a commandment implies obligation, whereas a counsel is left to the option
of the one to whom it is given.’
30
Darwall, The Second-Person Standpoint, esp. ch. 5.
 What is a Divine Command? 45
says: ‘If you wish to be perfect, go sell your possessions, and give the money to
the poor.’ There are different ways to take this story. Perhaps Christ’s word is a
precept to this particular young man, because he was enslaved to his money,
and this impediment had to be removed if he was to be perfect. But perhaps
also the same words would not be a precept but a counsel to someone who was
not so enslaved.
Jesus tells us to be perfect as our Father in heaven is perfect.31 If this is a
command, does it not follow that, when Jesus says to the rich young man, ‘If you
wish to be perfect, go sell’, this is also a command, a command to carry out the
means to the commanded end? Here it is useful to make some distinctions.
First, the difference between perfect (confusingly) and imperfect duties is
helpful in seeing that the command to be perfect is in a certain way indeter-
minate. The distinction has been made in various ways, but one way is through
Kant’s examples in the Groundwork, where the duties not to lie and not to
commit suicide are perfect, and the duties to help others and develop one’s
talents are imperfect.32 The difference is that in the first case you are in a bad
situation, you have an inclination to do some act to get out of it, and the perfect
duty intervenes to stop that particular act. In the second case you are in a good
situation, you have an inclination not to do anything to remove yourself from it,
and the imperfect duty intervenes to tell you to do something, although it does
not tell you what in particular to do. But, while this distinction goes some way
towards explaining the imperative ‘Be perfect’, it is not enough. It captures the
indeterminacy of how the imperative is to be carried out. But it does not explain
the way in which ‘Be perfect’ gives us an ideal. The word ‘ideal’ here does not
imply that we are given merely an ideal, in the sense that the prescription is to be
regarded as itself unattainable, and the realistic goal is not attainment but
merely trying to be more like what is prescribed. An example of this might be
that, if I am a mediocre singer, I might be told to sing like Caruso, where all that
is meant is that I should sing a bit more like Caruso. But Christian doctrine
standardly sees Jesus as giving us in his own life a model for what perfection
would be like. This has to be qualified by what Chapter 1 said about the
uniqueness of each person’s perfection. As Gerard Manley Hopkins puts it:
For Christ plays in ten thousand places,
Lovely in limbs, and lovely in eyes not his
To the Father through the features of men’s faces.33

31
Matthew 5: 48. See also Luke 17: 10: ‘So you also, when you have done all that you were
ordered to do, say “We are worthless slaves; we have done only what we ought to have done!” ’
But this text is perhaps limited to commands in the narrow sense, and therefore does not settle
the question whether God counsels beyond the divine commands.
32
Gl. iv. 422–3.
33
Gerard Manley Hopkins, ‘As Kingfishers Catch Fire, Dragonflies Draw Flame’, in Gerard
Manley Hopkins: Poems and Prose, ed. W. H. Gardner (Harmondsworth: Penguin, 1953), 51.
46 God’s Command
But this name into which we are called to live (Section 1.4) is not merely what we
should try to reach, or get closer to reaching; it is our destination. Imperfect
duties do not all give us ideals in this way, though they give us indeterminacy
about how they are to be realized. Thus ‘Eat more spinach’ would meet the
criterion for the prescription of an imperfect duty, but it does not give us an ideal.
What more do we need to say in order to capture the special nature of the
prescription to be perfect? One point is that the calling towards our own
perfection, which is itself a perfection of the common nature ‘humanity’, is
continued in the next life. But obligations do not continue in the next life. Why
is this, and what does it tell us about the nature of obligation? Here again Kant
is useful to the extent that he sees that God does not have obligations, since
God does not have any contrary inclinations that have to be disciplined, and
the same is true of finite holy beings.34 In the case of both perfect and
imperfect duties, the prescription is most often to do something other than
what inclination is prompting one to do. But the process of sanctification and
then glorification is one in which the inclinations come to be more and more
in line with duty, so that there is less and less disciplining to be done. Kant did
not think that we humans can ever be holy, but he did not give good
justification for this claim within either the theoretical or the practical use of
reason. It is better to say that there is a call to be holy, and that in this state we
would no longer be under obligation. This suggests that even in this life, where
there are competing inclinations, the call is to become somebody who does not
feel resistance that has to be overcome. This makes the term ‘obligation’
inadequate also for what the call (unlike a command) creates.
There is another feature of obligation that will be more central when we
discuss the nature of authority in the next section of this chapter. Obligation is
accountability to someone. But there are different types of accountability.
Accountability brings with it the envisaging of a sanction of some kind for
non-compliance, even if it is only the sanction of blame. Darwall puts the
point by saying that accountability makes blame for non-compliance ‘appro-
priate’, and quotes Pufendorf, who says: ‘An obligation forces a man to
acknowledge of himself that the evil, which has been pointed out to the person
who deviates from an announced rule, falls upon him justly.’35 But, for calls
and counsels, there is not the conceptually implied envisioning of condemna-
tion and punishment. We can still be answerable, but not accountable in the
sense that there is a sanction in the offing. In Chapter 1, we used Calvin’s
distinction between the ‘balky ass’ that is afraid of the whip and the son who

34
Gl. iv. 414. See also MdS vi. 383: ‘For finite holy beings (who could never be tempted to
violate duty) there would be no doctrine of virtue, but only a doctrine of morals, since the latter is
autonomy of practical reason whereas the former is also autocracy of practical reason, that is, it
involves consciousness of the capacity to master one’s inclinations when they rebel against
the law.’
35
Darwall, The Second-Person Standpoint, 111–12.
 What is a Divine Command? 47
loves the generous and candid father. The idea was that we are supposed to
move from the first kind of relation with God to the second.
Now we can return to the rich young man, and ask whether Jesus is giving a
command or a counsel. There are different possibilities depending on details
of the situation. Perhaps, as suggested earlier, he was enslaved by his wealth,
and he had to give up his possessions if he was to follow Jesus. Perhaps this
was the only thing still holding him back. This is suggested in the text by the
end of the story: ‘Jesus said, “Go, sell your possessions, and give your money to
the poor, and you will have treasure in heaven; then come, follow.” When the
young man heard this word, he went away grieving: for he had many posses-
sions.’36 On this reading, Jesus’s prescription is best construed as a singular
precept. But there is another possibility, perhaps fitting a slightly different
scenario. We can imagine a young man testing the waters, and trying to find
out what the cost would be of discipleship.37 He had made the mistake of
thinking that the demand was confined to commands like those in the second
table of the Ten Commandments. Jesus shows him that there is much more
than these commands, or than any commands in the narrow sense of ‘com-
mand’. There is a call to a destination beyond this life, treasure in heaven. In
any case, I propose that we accept the general point that God can address
imperative sentences to us that have two features different from commands in
the narrow sense; they give us an ideal or a picture of how our lives might be
that is not itself, in the stage of its final implementation, conceived to be at
odds with our inclinations, and they do not envisage, within their meaning,
condemnation or punishment for failure.
Here is an example of a divine counsel. A person who already has tenure at a
secular university is asked to consider leaving tenure to teach at a religious
college. The notepaper of the college, inviting him to an interview, has on it the
words: ‘My heart I offer to you, Lord, promptly and sincerely.’ He considers
there is an invitation here, a call that will in God’s providence bring blessing,
even though it is hard to justify compliance from a common-sense point of
view. He senses that for him to offer his heart to the Lord means to accept this
invitation. On the other hand, he does not have the sense that there is
something wrong with his current place of employment, or that he has some
kind of obligation to go.
God’s prescriptions to us are often counsels of this kind. If we were to
attribute emotions to God, we would say that God was disappointed when we
decline, or not as happy as God could have been, and not that God was angry.
But this is anthropomorphic language. Perhaps an ingredient in the picture is
that a call is often accompanied by a gift. In the Roman Catholic picture of the

36
Matthew 19: 21–2.
37
Compare the lawyer who tested Jesus in Luke 10: 25–37, reciting the love commandment,
to whom Jesus tells the story of the Good Samaritan, and says ‘Go and do thou likewise’.
48 God’s Command
three evangelical counsels, for example, the call to chastity is accompanied by a
gift from God to make that way of life livable in an excellent way. The refusal
of the call is in such a case the refusal of the gift, and an appropriate human
response to the refusal of a gift is disappointment.

2.1.6. Directly Effective Commands

The final item on the list of five types of prescription is ‘directly effective
commands’ (implet). These are still a species of divine prescription (still
voluntas signi), a species of revealed will and not yet disposing will. But, unlike
the species we have considered so far, they do not need to have any language-
using human recipient. God says, ‘Let there be light’, and there is light.38 It is
tempting to say that this is not really a command at all, but that is too swift.
The importance of the Genesis account here is that we are told that God
accomplished creation through speech, in Greek logos. For Christians this
suggests the role of the Second Person of the Trinity in creation, and John 1:
1–3 takes up this suggestion, with explicit reference back to the first chapter of
Genesis: ‘In the beginning was the Word, and the Word was with God, and the
Word was God. All things came into being through him, and without him not
one thing came into being.’ The idea of effecting something directly by
commanding it may seem strange. But it is not unique to the original creation.
When the Psalmist says (Psalm 85: 8), ‘Let me hear what God the Lord will
speak, for he will speak peace to his people, to his faithful’, he is imagining God
saying ‘Shalom’, ‘Peace be to you’, as the Lord says, for example, to Gideon in
Judges 6: 23. When God pronounces peace on us, that is a directly effective
command, and a work of the Holy Spirit.39 Perhaps the sense in which
creation by directly effective command is a ‘communicative act’ is attenuated.
It might be a communication either within the Trinity, or to angelic beings, or
to potential (but not yet actual) human recipients, or perhaps the implication
of the doctrine of creation by speech is just that the creation is in principle
intelligible. In any case, there is still the significant distinction to be made that
this last kind of command, unlike precept and prohibition, does not presup-
pose the existence of human recipients and does not imply sanction or
punishment for failure to comply. It is important to see that this category of
prescription nonetheless places creation in the category of something com-
manded. The claim of Chapter 4 is that God’s commands that produce our
obligations are themselves constrained by the human nature that God created,

38
Genesis 1: 3. See Psalm 33: 9 (about the earth): ‘He commanded, and it stood firm’, and
Genesis 1: 22, where God commands the non-human animals ‘Be fruitful and multiply’.
39
In the passage from Deuteronomy already discussed (chs 29–30), God pronounces both
blessing and curse.
 What is a Divine Command? 49
but that this does not take us outside God’s commands to something else
constraining God. Rather, God creates by command and sustains creation by
command, and then commands us with one of the other types of divine
prescription in a way that is consistent with that creative command.
We can now collect together these results and say that a divine command
that generates obligation is a prescription with which the person commanded
is not permitted not to comply, and a prescription in which there is an internal
reference, by the meaning of this kind of speech act, to the authority of the
speaker, and to some kind of condemnation if the command is not carried out.

2.2. DIVINE AUTHORITY

What does ‘authority’ mean here? We cannot simply take an account of

human authority and apply it to God without qualification. ‘Authority’ is in
this respect like any other term from human life that we try to apply to God.
We have to proceed by analogy, and recognize its limitations when they get in
the way.40 The term ‘authority’ in its use within human life is a ‘thick’ value
term, like ‘polite’. To say that someone is polite is usually to commend that
person, but at the same time it is possible to be too polite.41 Thick value terms
take up the criteria for their application into their meaning, and combine them
unstably with the evaluation, and the result is that with most such terms it is
possible to find cases where the criteria are approximately met, but the
evaluation is the opposite of the usual. Suppose the criteria for politeness
include deference to the opinions of others. It is possible to find people who
are too deferential. In the same way, to say that a person has authority is
usually to express approval of the way she exercises power. But it is also
possible to say that a person has wrongful or too much authority. What,
then, is rightful authority?
As with most value terms, we need to distinguish objective and subjective
uses of the term ‘authority’. In general, values can be present and operative
even if a person does not acknowledge or recognize them. In the subjective use
of ‘value’, I cannot sincerely say that something is a value for me unless I value
it. But in the objective use of ‘value’, there can be values relevant to my choices
that I do not acknowledge. Two qualifications are needed to this general point,
before we apply it to the case of ‘authority’. The first qualification is that

40
I do not mean to try to settle here the dispute between Aquinas and Scotus about univocity.
For one controversial treatment, see Richard Swinburne, The Coherence of Theism (Oxford:
Clarendon Press, 1993), 74–82.
41
See Rosalind Hursthouse, On Virtue Ethics (Oxford: Oxford University Press, 1999)
(henceforth OVE), 13: ‘So, it would appear, being generous, honest, benevolent, or courageous
can also be faults; or they are not always virtues, but sometimes faults.’
50 God’s Command
objectivity does not require independence of human beings as a whole; it
requires independence of the preferences of the person to whom the value
applies. A medical treatment can be good for a person whether he recognizes it
or not. John McDowell has argued convincingly that most of the values we
operate with, like secondary qualities such as colour, are relative to human
dispositions in general, and that this does not mean they are not objective.42
The second qualification is that, according to the ‘prescriptive realism’ out-
lined in Chapter 1, the full-orbed value judgement using the term will always
express a subjective state, though this does not impugn the objective reality of
the value properties picked out in the judgement. Sincerely assenting to a
statement involves believing something, and sincerely assenting to a command
(on the appropriate occasion, if it is within our power) involves doing some-
thing or at least wanting to do it.43 The consequence is that, when a person
makes this kind of value judgement, there will be a ‘value’ in the subjective
sense, an acknowledged pull towards some good or away from some evil
together with an endorsement of that pull, but the objective value can be
there whether it is acknowledged in such a judgement or not.
‘Authority’ has been distinguished from mere ‘power’ by the fact that a
person or thing that has authority properly influences me, whereas a person or
thing that has power merely influences me. This distinction between authority
and power comes from Bishop Butler, but it is not yet adequate.44 Not all
properly exercised influence is authority, because influence (even properly
exercised) can be merely causal, and not all authority is properly exercised. We
need a distinction between power and authority that appeals to the idea that
the person with authority gives (by commanding) reasons of a certain kind for
compliance to the person over whom she has the authority. But the term
‘reasons for action’ is itself value-laden, and this can again give rise to
confusion. In terms of the distinction between authority and power, a person
can objectively have reasons for action that are not acknowledged or even
psychologically accessible, so that these are not subjectively her reasons for
action. If we are talking in terms of subjective reasons, it is easy to see how one
person with undue influence over another person’s thought processes might
wrongly give the other person reasons for action. But this is also true with
objective reasons for action. As with other ‘thick’ value terms, there can be
cases where one person has authority over another wrongly. The reasons for
action may really be there (‘objectively’), produced by the person’s command,

42
John McDowell, ‘Values as Secondary Qualities’, in Ted Honderich (ed.), Morality and
Objectivity: A Tribute to J. L. Mackie (London: Routledge & Kegan Paul, 1985).
43
Section 4.3.1 returns to the discussion of this ‘judgement internalism’.
44
The distinction is made by Joseph Butler in his ‘Second Sermon upon Human Nature’, in
Fifteen Sermons (Charlottesville, VA: Ibis Publishing, 1987), 37–47. ‘Properly’ is a value term,
and so we need to distinguish the case where the influence is objectively proper and the case
where I subjectively perceive it as proper.
 What is a Divine Command? 51
but they should not be there. If someone has too much authority over me (for
example, a dominating therapist), what she tells me to do may nonetheless
give me reason for compliance if I genuinely feel better when I comply. We
cannot, therefore, simply define ‘authority’ in terms that make it always by
definition true that the person who exercises it does so legitimately.
God has, in the tradition of all three Abrahamic faiths, authority as sover-
eign. But now we need to make the distinction between ‘objective authority’
and ‘subjective authority’. God is sovereign, and in this sovereignty God has
both of these kinds of authority. But God has the first kind, objective authority,
for all human beings, whether they acknowledge it or not. Subjective authority,
however, is what the subject acknowledges as authoritative, and so God has
subjective authority only for those who acknowledge it or consent to it. God’s
sovereignty we can further distinguish into functions, by analogy with human
sovereignty, and Chapter 1 made a distinction into legislative, executive, and
judicial functions.45 God makes the law by commanding it, and runs the
kingdom in accordance with that law, and judges all human beings, whether
they have acknowledged God’s authority or not, by their compliance to that
law. In all three cases, God has objective authority but may not have subjective
authority. In all three cases God has rightful authority because the reasons for
a person’s compliance given by God’s commands are reasons that person
ought to have. This is because God’s commands make obligatory the good
things that God prescribes, all of which take us to our proper end by the path
God has selected for us, and our obedience is an expression of our love for God
that is good in itself. There is circularity here. God has rightful authority
because God’s commands give us the reasons for action that we ought to have,
and we ought to have them because God’s commands have authority. But the
circularity is not vicious, because the chain of justification terminates in the
principle known from its terms that God is to be loved and hence God is to
be obeyed.
We have been implicitly using Joseph Raz’s account of the concept of
practical authority to help explain divine authority: one person is practically
authoritative over another in a certain domain only if the first person’s dictates
bring about to some extent the reasons for action that the other has in that
domain.46 Objectively, when God commands a person to do something, this

45
Kant uses this analysis, e.g. at M. viii. 257, and it was the basis for the Kantian arguments in
Chapter 1 for the dependence of morality upon religion. Also see Suarez, De Legibus 2. 15. 20.
46
Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986). See Mark
Murphy, An Essay on Divine Authority (Ithaca, NY: Cornell University Press, 2002), 8–16, and
my review, ‘An Essay on Divine Authority’, Mind, 113/450 (2004), 375–9. I am indebted
throughout this section to private communication from him. Raz says that the authority
‘constitutes the reason’ for acting, and Murphy works out with great elegance what this means
for God’s authority. A thug who holds me up on the street with a gun gives me a reason to
comply with his demand, but his demand does not constitute such a reason; it merely causes it.
52 God’s Command
brings about a reason for acting in accordance with the command, even if,
subjectively, the person does not acknowledge this reason. There is an analogy
with the authority of the state. Kant holds that we have a reason to obey the
laws of the state because the state upholds external freedom, which is the
freedom of action that expresses in the world our internal freedom of choice
(which Kant thinks is valuable in itself). The state provides what he calls ‘a
hindrance to the hindrances’ to this freedom, by providing those who would
interfere with our external freedom a motive for refraining from this interfer-
ence.47 Even an anarchist who does not acknowledge the authority of the state
has an objective reason, on this view, to obey the laws; she has the value of
internal and external freedom. To apply the analogy, we can say that, when
God commands something by legislative authority, implemented in God’s
executive and judicial authority, there is an objective reason for obedience
from the union of wills (divine and human) that is both expressed in such
obedience, and that is good in itself.
We can take God’s executive authority first, and then God’s judicial au-
thority. In Kant’s language, God is the sovereign of the kingdom of ends, of
which we are mere members. The sovereign, unlike the other members, ‘is a
completely independent being, without needs and with unlimited resources
adequate to his will’.48 One example of what this means is that God can put us
next to the people God wants us to help. We can call this ‘the principle of
providential proximity’. There are cases where just because you are the sort of
person you are, and the person you are with is the sort of person she is, you are
able to help her or she is able to help you. God, by exercising the power to put
us next to particular people, makes possible the actualization of the ethical
commonwealth, the work, Kant says, of ‘a higher moral being through whose
universal organization the forces of single individuals, insufficient on their
own, are united for a common effect’.49 The parable of the Good Samaritan is
an example.50 The Samaritan, the priest, and the Levite were all placed next to
the man wounded by the side of the road, but only the Samaritan was a
neighbour (in Latin proximus) to him. Proximity does not need to be under-
stood geographically here. If I have a son who goes off to Zambia to work in a
village (or if my congregation has a member who does this), that village becomes
my village. There may be a village just next door with the same needs, but it is
not the one I have been given to help. The principle of providential proximity is

Murphy says (p. 12) that an authority actualizes reasons through commanding ‘by actualizing a
state of affairs that is the only non-actual element of a reason-candidate’. I do not think we
should confine authority-bearing acts, as Murphy does, to speech acts, but I am going to agree
with him, in Section 4.1.2, that God by telling us to do something makes something already good
(and so a reason-candidate for action) into something obligatory.
47
MdS vi. 396, and see 231: ‘[Whatever] counteracts the hindrance of an effect promotes that
effect and is consistent with it.’
48 49 50
Gl iv. 434. Rel. vi. 98. Luke 10: 25–37.
 What is a Divine Command? 53
helpful in overcoming the despair that comes from seeing the scale of the
world’s need as compared with my own pitiful resources. It is not, however,
simply obvious whom one has been placed next to. It takes discernment, as
God’s commands usually do. But the principle is still helpful. Note that God’s
executive authority gives a reason that is not simply from what Raz calls
‘theoretical’ authority. It is not simply that God sees better than I do (because
of God’s privileged epistemic position) what the reasons are, and transmits
them to me. Rather, God’s choices actually create or produce the reasons.
God puts me next to the person I am supposed to help, and I am to see being
placed next to her as preparing me for a divine command to pay special concern
to her interests.
God’s judicial authority gives me a reason because I am accountable to God.
This is true independently of whether I acknowledge this authority.51 A rival
view is Mark Murphy’s position that God does not have this objective author-
ity over those who have not acknowledged it, but that ‘one is required in
reason to subject oneself to God’s rule’, because there is decisive reason so to
do; to fail to subject oneself is to be guilty of practical irrationality.52 But this
account cannot accommodate the traditional view of all three Abrahamic
faiths about God’s judicial authority even over the non-believer. Christopher
Dodsworth makes this point well, relying on Stephen Darwall.53 Dodsworth
imagines the case where I watch the weather report every morning when
I wake up, and the meteorologist makes recommendations on how to dress.
Suppose I reasonably want to dress appropriately for the weather, but, since
this principle is underdetermined, I decide to dress however the meteorologist
suggests. I now have a reason to do as she says, grounded in the ‘authority’ she
now possesses in virtue of my consent. But Dodsworth says, surely rightly, that
she does not in fact have authority over me at all, because she has no special

51
This point is independent of whether God’s authority is properly speaking a perfection.
Suppose we say, like Murphy, that God’s perfections have to have intrinsic maxima—so that
God’s omniscience, for example, is understood in terms of God knowing of every proposition
whether it is true or false. We might say that God’s authority is not a perfection, because it
increases indefinitely as its scope or domain increases. But Murphy thinks God does not have
objective authority over all rational beings, whether this would be, if it existed, a perfection
or not.
52
Murphy, An Essay on Divine Authority, 175. This is what he calls ‘the compliance thesis’:
that for any commands God gives, those commands are backed by decisive reasons.
53
Christopher Dodsworth, ‘Understanding Divine Authority’, Faith and Philosophy, 28/2
(2011), 190–208. The case of the meteorologist is not exactly parallel, because she does not know
that her recommendations make any difference to what I do. But, even if she did, Murphy thinks
that practical authority does not bring with it a right to demand, just as the theoretical authority
of an expert in the field does not; so the fact that the meteorologist does not have the right to
demand my clothing compliance does not mean that she does not have practical authority over
me. But practical authority of the kind implied in command is different in this way from
theoretical authority, though perhaps we should allow that there is a kind of practical authority
(possessed, for example, by some wise figure I respect) that is not different in this way.
54 God’s Command
standing to demand that I do what she says. I am not accountable to her.
Darwall argues that the reason created by a valid authoritative demand is
‘second-personal’. This ‘is the perspective you and I take up when we make
and acknowledge claims on one another’s actions and will’.54 Darwall gives as
examples orders, requests, claims, reproaches, complaints, demands, prom-
ises, contracts, givings of consent, and commands. A second-personal reason
is ‘grounded in authority relations that an addresser takes to hold between him
and his addressee’.55 Not all the items on Darwall’s list are grounded in
authority in this way, but commands in the narrow sense are, and this is the
sense in which God’s command generates obligation. The difference between
commands from human beings and divine commands is that, whereas we are
accountable to God, it is odd to say that God is accountable to us. Even if we
think God has obligations to us (which is doubtful), God is not accountable to
us, because the obligations (if there are any) would be self-imposed. But the
point is that an account of God’s authority needs to make reference to the fact
that, when we fail to obey God’s commands, this makes us rightly liable to
God’s rebuke and punishment, as in the first Psalm, ‘therefore the wicked will
not stand in the judgment . . . but the way of the wicked will perish’.56 This is
true for believers and non-believers alike.
It is doubtful that God has obligations to us, and it is also doubtful whether
God has permission-rights with respect to us. Nicholas Wolterstorff says that
our obligations to obey God depend upon God having rights to our obedience,
and that God has obligations towards us produced by divine promises and
covenants.57 God’s rights are, according to Wolterstorff, a permission-right to
give us commands and a claim-right to our obedience. But the question seems
inescapable: ‘Who is granting this permission?’ Chapter 1 gave the Scotist
response to the demand for a justification of the principle that God is to be
loved, and thus to be obeyed, and we return to this question in the final section
of this chapter. Scotus says that we know this principle from its terms. But, if
we try to provide a justification in terms of the language of rights and
obligations, we will find that it fits the ordinary members of the kingdom of
ends, but not its sovereign, ‘who is not subject to the will of any other’.58 The
question of God’s obligations is especially acute. Surely, when God makes a
promise to us or makes a covenant with us, God is obligated to comply with it?

54
Darwall, The Second-Person Standpoint, 3. He thinks that accountability is mutual, so that
God would be accountable to us. I deny this in what follows, and I argue for this denial in
Section 2.3.2.
55
Darwall, The Second-Person Standpoint, 4.
56
Psalm 1: 5–6. Murphy can have something resembling God’s judicial authority—namely,
God’s sanctioning, rewarding, and threatening. But he cannot have divine judicial authority
proper.
57
Wolterstorff, Justice: Rights and Wrongs, ch. 12.
58
Gl. iv. 433.
 What is a Divine Command? 55
Scotus is useful again at this point. He makes a distinction between God’s
absolute power and God’s ordained power (which God exercises within an
ordinance divinely established).59 An agent can act in conformity with some
right and just law in accordance with his or her ordained power. When that
upright law—according to which an agent must act in order to act
ordinately—is not in the power of that agent, ‘then its absolute power cannot
exceed its ordained power in regards to any object without its acting disorderly
or inordinately’. But God can act otherwise, ‘so that he establishes another
upright law, which would be right, because no law is right except insofar as the
divine will accepts it as established’. So God is acting within the divine
ordained power by keeping promises, but it is always possible for God by
the absolute divine power to establish a different upright order. God’s obliga-
tion thus does not, we might say, go all the way down. God is unlike us in
this respect.

2 . 3. BA R TH ON DI V I N E CO M M A N D

The final section of this chapter lists briefly six points we can learn about
human beings and about God from the premiss that God gives us commands,
and then discusses three puzzles about these points. These points are all taken
from Barth’s discussion in Church Dogmatics III/4.60 These themes will re-
emerge in a number of places in the chapters that follow, but it is helpful to get
them listed at the beginning. They are not original with Barth, and we will see
as we go various other authors, not influenced by him, who have made the
same points.

2.3.1. Six Implications of Our Being Commanded by God

From the premiss that God gives us commands, we can learn, first, that we and
God are different; we are not, that is to say, part of God. This is because
commands are not addressed to oneself, except in an extended sense in which
one is treating oneself as another. Second, commands are given to responders
of a certain kind: those who can obey. This is explained in the four points
that follow, called subsequently ‘the four Barthian constraints’. First, the

59
Scotus, Ord. I, dist. 44.
60
CD III/4. 4–8 (328–32). These points are not exclusively Barth’s. Thus al-Maturidi says that
‘there is nothing that God would speak to man unless He also caused the way through which he
can understand it’ (Kitab al-Tawhid, ed. Fathalla Kholeif (Beirut: Dar al-Mashriq, 1970), 137).
Maimonides says that divine commandments ‘are not given to beasts and beings devoid of
intellect’ (Guide, ii. 11).
56 God’s Command
commands are given to centres of agency, to responders whose obedience
consists in acting and living a certain way.61 These centres of agency are
individuals, although we can talk in an extended sense about the agency of
collectives. The commands, as Barth emphasizes repeatedly, are primarily to
particular people in particular circumstances. Second, commands are to centres
of agency whose obedience consists in changing how things are, or in resisting
change. These centres of agency who receive the command are therefore (since
time is, as Aristotle says, either change itself or the measure of change) in
time.62 They have to persist, in order to be obedient, through the hearing of the
command to the obeying of it. Third, commands are given to free beings, in
the sense of beings who are not under external causation in their obedience.
Commands are not given, except in an extended sense, to machines, which can
be operated causally by the pressing of the right keys or buttons or levers. The
nature of this freedom requires a great deal more discussion, and we return to
it in Section 5.2. But the notion of ‘command’ itself requires that in some sense
or other the response is ‘up to’ the responder, and is not simply caused by the
giving of the command. Fourth, the responder has to be part of a language
community. Commands are standardly addressed to the responder in lan-
guage, and language is a communal enterprise. Wittgenstein’s Private Lan-
guage Argument may or may not be successful in showing that an individual
in isolation cannot have a language. But, in the case of commands, which have
to be between at least two people, there has to be at least the language common
to the two of them. If commands can be given to groups as well as to
individuals, which seems right, then the language has to be common to the
commander and the group.
So far, these points have been about human beings. The sixth point is about
God. If God gives us commands, and the function of commanding as a speech

61
This point about the nature of the responders is one Ockham relies upon in his discussion
of the question whether God can command us not to love God. His view is that the command not
to love God, though its content is possible in itself, is pragmatically incoherent (a practical
contradiction) because it cannot be disobeyed; this is because to disobey it is already to love God.
Section 1.3 argued that loving God entails obedience. See Ockham, Quodlibetal Questions III. 14.
A content can be non-contradictory in itself, but contradictory as commanded. A content can
also be non-contradictory as commanded, but contradictory as commanded by God. See Lucan
Freppert, The Basis of Morality according to William Ockham (Chicago: Franciscan Herald Press,
1988), 93, who argues that this view is different from that of Scotus discussed in Chapter 1.
62
Metaphysics XII. 6. 1071b10. Gary Watson lays out what he calls ‘constraints of moral
address’, in ‘Responsibility and the Limits of Evil: Variations on a Strawsonian Theme’, in
F. D. Schoeman (ed.), Responsibility, Character, and the Emotions (Cambridge: Cambridge
University Press, 1987), 263–4. Many of these are taken up by Darwall, The Second-Person
Standpoint, who makes reference back to Johann Gottlieb Fichte, Foundations of Natural Right,
ed. Frederick Neuhouser, trans. Michael Bauer (Cambridge: Cambridge University Press, 2000),
and Pufendorf, who held, according to Darwall, that ‘obligations can arise from God’s com-
mands only if God addresses us in a way that presupposes our second-personal competence’
(Darwall, The Second-Person Standpoint, 23).
 What is a Divine Command? 57
act is to change the world through the agency of the responder to whom the
command is addressed, and if the command is an expression of the desire that
the world change in this way, then we can attribute something like desires (in
the broad sense) to God. More usually, theologians would say God has a will.
This point takes us back to the discussion of commands at the beginning of
this chapter. That we have a God who commands is distinctive of the Abra-
hamic faiths, and distinguishes them from, for example, Aristotle’s religion.
Aristotle tells us that ‘the god is a governor not in a prescriptive [or com-
manding] fashion (in Greek epitaktike), but [the god] is that for [the sake of]
which practical wisdom prescribes’.63 Since we saw earlier that God’s creation
is also by command, it is reasonable to say that command is the characteristic
fashion by which, in the Abrahamic faiths, God relates to us, either by creating
or by telling us how to live inside creation. We can infer from God’s com-
mands that God, like us, expresses a will, and, through that expression in
language that we can at least partly understand, intends to accomplish changes
in the world through our response. This is different from Aristotle’s picture,
according to which God moves us as a kind of magnet, not by telling us what
to do. And behind this difference is an even more significant one. God is not,
for Aristotle, in a personal relation with us, but the Abrahamic faiths make our
relation to God personal, and mediate that relation by God’s command to us.
It is true that God’s will and God’s command can diverge, as in the famous
case of Abraham and his son. When they do, are we bound (according to
divine command theory) by God’s will or by God’s command?64 We should
hold ourselves bound by the command. We should take the command to be an
expression of the will, because we do not have other access to the will (except
through other divine prescriptions, including directly effective commands).
But the assumption that a command is an expression of the divine will can, in
certain cases, be overridden by another command.

2.3.2. Three Puzzles

We end with three puzzles about the four Barthian constraints, that we are
individual centres of agency, in time, free, and language-users.65 The first
puzzle is about why beings like us in these four ways could not bind themselves
morally without bringing in God. Stephen Darwall in The Second-Person
Standpoint raises this objection to Pufendorf ’s theory that ‘moral entities’

63
Eudemian Ethics (henceforth EE), VIII. 3. 1249b15–16.
64
Robert M. Adams has a sustained critique of Mark Murphy’s claim that it is God’s will, in
Finite and Infinite Goods, 258–62.
65
Robert M. Adams discusses these difficulties in Finite and Infinite Goods, 267–70, though
not in reference to Barth. We return to some of them when we come back to Barth in Chapter 5.
58 God’s Command
(such as obligations) are produced through God’s ‘imposition’ of his will in
commands.66 Darwall first takes an objection from Ralph Cudworth: ‘It was
never heard of, that any one founded all his Authority of commanding others,
and others’ Obligation or Duty to Obey his Commands, in a Law of his own
making, that men should be Required, Obliged, or Bound to Obey him.’67 This
is a difficulty mentioned already in Chapter 1, which proposed that Scotus has
a solution to it. We can now add that Pufendorf has the same solution, at least
to the extent that both philosophers say that the justification of obedience to
God terminates in something that does not itself need justification. For Scotus,
the principle that God is to be loved is known from its terms, and therefore
does not require any justification from any antecedent principle. We know
that God is to be loved, and so that God is to be obeyed, just by knowing that
God is the supreme good. Pufendorf also takes God’s authority as not needing
justification.68 Like Calvin, he is giving us a double motivation for obedience,
but he is not justifying.69 As several commentators have noted, Pufendorf
could not use gratitude, which he regards as an imperfect duty, as a justifica-
tion for God’s right to demand obedience, which he regards as a perfect duty.70
Rather, he regards God’s authority as axiomatic, in the same way mathematics
has axioms or first principles, which ‘merit belief upon their own evidence’.71

66
Samuel Pufendorf, On the Law of Nature and Nations, trans. C. H. Oldfather and
W. A. Oldfather (Oxford: Clarendon Press, 1934), 5. For a fuller treatment of this section’s
reading of Pufendorf, see Neil Arner, ‘Theological Voluntarism and the Natural Law: The
Integrated Moral Theories of John Duns Scotus, John Calvin, and Samuel Pufendorf ’,
Ph.D. dissertation, Yale University, 2012, ch. 4.
67
Ralph Cudworth, A Treatise Concerning Eternal and Immutable Morality, ed. Sarah Hutton
(Cambridge: Cambridge University Press, 1996), I. ii. 3. See also Gottfried Wilhelm Leibniz, in
his Opinion on the Principles of Pufendorf, in Political Writings, ed. Patrick Riley (Cambridge:
Cambridge University Press, 1989), 70–5.
68
This is obscured by a passage Darwall quotes from Pufendorf (The Second-Person Stand-
point, 110): ‘If it appears that the same person wishes me well, and can take better care for my
future than can I, and [that this person] also claims at the same time a right to direct my acts,
[then] there is no apparent reason why I should wish to question [that one’s sovereignty]. And
this is all the more true if I am indebted to [that person] for my very being (Acts 17: 24 ff.). And
why should not [the one], who gave [humanity] the power of free action, be able from [that
one’s] own imperium to limit some part of [humans’] liberty?’ (On the Law of Nature and
Nations, 1. 6. 12). Darwall interprets Pufendorf as trying to justify the principle that we ought to
obey God by deriving it from an antecedent principle that we owe gratitude to our benefactors.
But in this passage Pufendorf twice takes divine sovereignty as already granted, and is interested
in showing that we have no reason to question it, if hesitations should arise.
69
Calvin insists that our proper motivation for obedience is love of God (Institutes, 2. 8. 11).
But this can be supplemented by considerations of all the good things God has given us
(Institutes, 2. 3. 3). See Benjamin Lipscomb, ‘Power and Authority in Pufendorf ’, History of
Philosophy Quarterly, 22/3 (2005), 217–18, who uses the distinction between justifying and
explaining a requirement.
70
See Terence Irwin, The Development of Ethics (Oxford: Oxford University Press, 2011), 2.
42. 577, n. 44.
71
Pufendorf, On the Law of Nature and Nations, 1. 2. 3.
 What is a Divine Command? 59
Darwall then objects to Pufendorf that, by acknowledging that God’s
command presupposes our competence as free and rational responders, he
has in fact undermined the need for bringing in God at all in understanding
moral obligation. Darwall says: ‘It follows [from this acknowledgement]
that we human beings can enter (individually) into moral community with
God only if we have the authority to form a moral community ourselves as
mutually accountable free and rational persons: a Kantian “realm of ends”.’72
He concludes that, if we can already form such a community, appeal to a
divine sovereign is unnecessary. But Darwall does not see that the relevant
competence here is the competence to bring about what Kant calls the highest
good. We need to return to Kant in the Groundwork, where he introduces the
notion of a kingdom of ends, because Kant has an important truth here.
According to the argument from providence in Chapter 1, Kant thinks it
important to say that the kingdom of ends has a king:
A rational being belongs as a member to the kingdom of ends when he gives
universal laws in it but is also himself subject to these laws. He belongs to it as
sovereign when, as lawgiving, he is not subject to the will of any other. A rational
being must always regard himself as lawgiving in a kingdom of ends possible
through freedom of the will, whether as a member or as sovereign. He cannot,
however, hold the position of sovereign merely by the maxims of his will but only
in case he is a completely independent being, without needs and with unlimited
resources adequate to his will.73
This sovereign can only be God, because only God is without needs and with
unlimited resources adequate to the divine will. Christine Korsgaard prefers to
talk about a ‘republic of all rational beings’, but J. L. Mackie is more accurate
when he says: ‘But for the need to give God a special place in it, [the kingdom
of ends] would have been better called a commonwealth of ends.’74 If
Chapter 1 was right, the kingdom needs a sovereign who can bring about
the highest good, the union of happiness and virtue, which is the end given us
by morality itself.
Kant agrees that we belong together with God in a kingdom of ends, but he
also holds that God is superior to us because God runs this kingdom, and
judges us according to whether we live by the laws of this kingdom. There is an
important passage about conscience in the Metaphysics of Morals that can help
us see how Kant envisages the role God plays in our moral lives.75 When a

72
Darwall, The Second-Person Standpoint, 115. David Novak makes the same argument in
Natural Law in Judaism (Cambridge: Cambridge University Press, 1998), 39–40, and we return
to this in Section 7.2.1.
73
Gl. iv. 433–4 (emphasis added). I have discussed this passage in detail in God’s Call, ch. 3.
74
Korsgaard et al., Sources of Normativity, 99, and see 127. Mackie, E. 45.
75
MdS vi. 438–40 (emphases original). I have discussed this passage in detail in ‘Conscience
and the Moral Epistemology of Divine Command Theory’, in Michael Bergmann and Patrick
60 God’s Command
person is reflecting about the rightness or wrongness of what he has done,
Kant suggests he will think in terms of a court, and the ‘consciousness of an
internal court in the human being (before which his thoughts accuse or excuse
one another) is conscience’. The judge here has to be considered as a figure
outside the individual, because he has to be pictured as having qualities that
are inconsistent with being human at all. Kant says pointedly that this pictured
judge ‘may be an actual person or a merely ideal person that reason creates for
itself ’, but this person has to be conceived as ‘as a scrutinizer of hearts’ (a role
that humans cannot play) who ‘imposes all obligation’ (this is God’s role as
legislator) and who has ‘all power to give effect to his laws’ (this is God’s role
as executive). Then Kant says (emphasis added): ‘When the proceedings are
concluded the internal judge, as a person having power, pronounces the
sentence of happiness or misery, as the moral results of the deed. Our reason
cannot pursue further his power (as ruler of the world) in this function; we can
only revere his unconditional jubeo [I order] or veto [I forbid].’ Kant is
conceding here that, even though I can entertain within myself the thought
of what an actual God would prescribe and the verdict that an actual God
would reach, and I can repeat in my own will the legislating and the verdict,
I cannot repeat within my own will the omnipotent supervision of the world.
The importance of this passage is that it shows what kind of equality we do
and do not have with God. There is no need to move, like Darwall, from seeing
morality in terms of accountability to seeing morality as precluding a superior
will. We humans do have, on this picture, equal membership in the kingdom
of ends with God. We, like God, make the moral law; in our case we make it by
making the law a law for us. This is what Kant means by ‘autonomy’. He does
not mean that we create the law.76 In the terms I quoted in Chapter 1 from
Scotus, we will in our wills what God wills for our willing. The answer to the
first puzzle, in Kantian terms, is that the realm of ends needs a sovereign.
The second puzzle about the Barthian constraints is whether there are
constraints from the limits of our understanding on what God can command.
Robert M. Adams takes a robust, common-sense approach. ‘Under normal
circumstances, one could neither command nor intend to command a sub-
ordinate to load a truck by saying, in English, “Bring me a cup of coffee.”’77
This is because no one (in normal circumstances) could understand ‘Bring me
a cup of coffee’ as a command to load a truck. But, in communications
between God and us, we have to be careful about assuming that God does
not say to us what goes beyond our current understanding. Barth defends an

Kain (eds), Challenges to Moral and Religious Belief: Disagreement and Evolution (Oxford:
Oxford University Press, 2014), 101–7.
76
See my God’s Call, 92–7, and Patrick Kain, ‘Self-Legislation in Kant’s Moral Philosophy’,
Archiv für Geschichte der Philosophie, 86/3 (2004), 257–306.
77
Adams, Finite and Infinite Goods, 269.
 What is a Divine Command? 61
account of language according to which the same words do not mean the
same things when used by us and by God, but God enables us nonetheless to
understand God’s language by the gracious sending of the Holy Spirit. I think
that he goes too far in the direction of radical homonymy, or equivocity, but
I will not try to defend that claim now.78 Some distance, we might argue,
between what God means and what we understand is at least congruent with,
if not required by, our freedom and our being in time. Consider Jesus’s saying
to the disciples (Mark 1: 17) that if they follow him they will become fishers of
human beings, or (Mark 8: 34) that they will have to deny themselves and take
up their cross. There is no indication in the text of any context that would
have enabled the disciples to understand with clarity what Jesus was talking
about. The same is true about the destination, which is unique to each of us,
and which God already knows, but which we catch only in glimpses. God
reveals enough to keep us going, but does not reveal the whole thing, because
God is respecting our need to work out how to live. Kant puts this point by
saying that God did not intend us to be marionettes that could be manipu-
lated by pulling their strings.79 Here is an analogy from inter-human rela-
tions. I may have promised in my wedding vows to ‘love and cherish’ my
spouse. But I may not understand until many years later what cherishing
involves. We are given the word in advance, but what it means we discover
only by our own choices, bit by bit.
A third puzzle is that the model of a human command suggests that the
recipient has to be able to recognize that she has been commanded by the
commander. Is this necessary for accountability? I think it is not necessary
even in the case of human commands. There can be institutions (like Yale
University) in which there are rules that everyone is accountable to follow,
because the institution penalizes non-compliance, even though no one knows
who authorized them or signed off on them. Sometimes they have not been
explicitly articulated by any authoritative voice. In English law, a city can have
a right to the name ‘city’ by ‘ancient prescriptive usage’, even though most
cities may have a right to the name because of explicit royal charter.80 In any
case, it is true about divine commands that their audience may not know their
origin. God, in the divine legislative authority, promulgates the commands,
but those to whom they are promulgated do not necessarily know that it is God
who has promulgated them. This is a version of the doctrine of general
revelation. Even if people do not know God is the author, they can still be
accountable to what has been revealed, and they can still have obligations.
They may even have a sense of being commanded without knowing who it is
that is commanding them. Section 2.1.2 suggested that God sometimes speaks

78
A contrary view is given by Hector, Theology without Metaphysics, 126–46.
79 80
KpV v. 147. I thank Andrew Forsyth for this example.
62 God’s Command
to us through other people.81 We can receive divine commands in the words
presented to us by people we know, or even people we do not know. My wife,
for example, can sometimes show me what I have done wrong, though it
would be dangerous if she or I thought she was reliably God’s mouthpiece.
Nathan the prophet played this sort of role for King David.82 But, when we
have ‘deputized’ communication of this sort, there can be all sorts of errors
that creep in on the part of the deputy. Again, the reception of the divine
command in this sort of case requires discretion and discernment, and it will
often be possible to get it wrong. The answer to the third puzzle is that the
people who receive divine command do not always know that it is divine
command they are receiving.
In summary, this chapter has tried to do three things. First, it gave a general
account of prescription, and then distinguished five different kinds of divine
prescription, giving examples of the main kinds. Then it isolated one kind,
divine command in the narrow sense that generates obligation, and that is tied
to the authority of the commander, and it discussed what ‘authority’ means
here. Finally, it mentioned some features of our own agency and God’s that
follow from God’s commanding us, and it discussed in a preliminary way
some puzzles about these features.

81
See Nicholas Wolterstorff, Divine Discourse: Philosophical Reflections on the Claim that
God Speaks (Cambridge: Cambridge University Press, 1995). But he intends this as an account of
Scripture, and I am not proposing that here. Divine communication through other people
includes the seventh and eighth items on Evans’s list of ways that God communicates prescrip-
tions to us (see n. 24).
82
2 Samuel 12: 1–14.
 3

Eudaemonism

I N T R O D U C TI O N

Chapter 1 described three ways in which we can establish different kinds

of dependence relation of morality upon God. The third of these ways was the
argument from justification. If we ask the normative question ‘Why should I be
moral?’ or ‘Why should I take the moral demand as a valid demand upon me?’,
one answer we might propose is ‘because it will make me happy’. Another
answer is ‘because it will fulfil my nature as a human being’. These two answers
are related because one theory about happiness is that it resides in the fulfilment
of our nature. Nonetheless the issues raised by the two answers are to some
extent distinct, and can be treated separately despite the overlap. If either of
these answers is sufficient, a divine command theory will not be needed for
answering the normative question. The present chapter is about the first of these
answers, and the next chapter is about the proposed derivation of the moral law
from human nature. My claim will be that the first answer, ‘because it will make
me happy’, fails for two reasons. First, it is not strictly true, at least in this life.
Second, even if it is true, it gives the wrong kind of motivation. Questions of
justification and questions of motivation are different but linked on a Kantian
account of morality. This is because, if the terminus of a person’s motivation is
her own happiness, she is not following the moral law for its own sake, and
therefore ‘because it will make me happy’ fails as a justification of this kind of
moral obedience. The claim that such a pattern of motivation is ‘unacceptably
self-regarding’ is the central topic of this chapter.

3.1. DOES MORALITY MAKE YOU HAPPY?

It is not strictly true to say that morality makes you happy in this life, or that, if
you act well, things will go well. There are two reasons for this. The first is that
you can be morally good and still be miserable, because moral virtue does not
have the right kind of leverage to secure happiness. Consider, for example,
64 God’s Command
people who are doing whatever they can to follow the moral law, but who are
clinically depressed. Or consider that a great deal of our happiness is
dependent on our relations with other people. If all the people we know and
love are virtuous, perhaps our virtue will bring about our happiness. This is
what Kant calls ‘the Idea of Self-Rewarding Morality’.1 But, as things are, we
live in a world in which many of the people we know and love are not doing
whatever they can to follow the moral law, and in this way we become sources
of unhappiness for each other. Humans can and do make life for each other so
bad that even virtue is not enough to produce happiness. This is what Aristotle
says about the life of Priam of Troy, which could not reasonably be called
happy even though he was conspicuously virtuous.2
The second reason it is not strictly true to say that morality makes you
happy in this life is that morality not merely fails to secure happiness; it can
actually decrease it. Kant makes the point that, if you are unhappy, but believe
(because you are doing whatever you can to follow the moral law) that you
deserve to be happy, the virtue makes the unhappiness worse.3 The emphasis
on desert is a bit offputting here, but we can make the point without it. Because
the best kind of life includes activities in accordance with virtue, people who
systematically deny you the opportunities for these activities deny you the best
kind of life. The more capacity you have for excellence, including moral
excellence, the more frustrating this will be. Moreover, the more virtuous
you are, the more acutely you will feel the sufferings of those around you and
hate the injustice that causes it. But there are more straightforward ways to make
the point. Morality may require a genuine sacrifice of this-worldly happiness.
Chapter 1 gave the example of the elderly parent that a person thinks she should
take to live with her in her home, even though she foresees that this will not go
well for her; but the claim is general. Most people are not very happy, and they
are also not very good. Most of us seem to have both of these goals (happiness
and being a good person), and seem to end up compromising both of them on a
regular basis, in small ways and large. In this situation, the key question is one of
motivation and ranking. How do we negotiate the continual dilemmas in which
one or the other seems to have to give way?
To say that there is regularly tension between the two goals is consistent
with saying, with RMH, that, if we were bringing up a child purely in his own
interest, we should try to inculcate in him some reasonably demanding moral
principles, with the attendant moral feelings.4 His view was that virtue was

1
KrV A809 = B837: ‘Freedom, partly inspired and partly restricted by moral laws, would itself
be the cause of general happiness, since rational beings, under the guidance of such principles,
would themselves be the authors of both their own enduring well-being and that of others.’
2
NE I. 9. 1100a5–9.
3
MC xxvii. 302: ‘The virtuous disposition may increase the pain of this life still further.’
4
RMH, Moral Thinking (Oxford: Clarendon Press, 1981), 203–5. See OVE 178–9. Hursthouse
does not like calling the judgement here ‘empirical’ because she thinks that it is already informed by
ethical commitment.
 Eudaemonism 65
necessary but not sufficient for a good human life. RMH was not here talking
about the afterlife; ‘but, confining ourselves to this world, we have, if morality
is to be a viable enterprise, to believe that if we adopt moral purposes and
principles, we stand a reasonable chance of carrying them out and not
perishing uselessly in the process. And I myself would bring up any children
that I had charge of accordingly.’ It is fun to quote this as the beneficiary of this
declared intention. He went on to say that parents who have the interests of
their children at heart will teach principles that go somewhat beyond those
common to all and those common to particular roles, ‘for it does look as if
people who set themselves higher moral standards which are within their
capacity, or not too far outside it, are in general happier than those who do not
set their sights so high’. But this is, he said, an empirical judgement. He
conceded that occasions will arise in which the saints’ or heroes’ principles
will require them to make very great sacrifices. ‘In that case, if the cause is a
good one, we shall admire them, and perhaps, significantly, even wish that we
were more like them; but we are not required to say . . . that they were acting
on that occasion in their own interest. . . . It may be that the best sort of person
to try to become is one who on such rare occasions will act contrary to his own
interest in real truth and not, as he often will, merely seem on the surface to be
doing that.’ So, if parents educate their children to admire and practise virtue,
they may be bringing it about that in some unlikely contingencies their
children have to pay a very high price. Even so, he thought, the parents should
bring up their children that way, because it is usually the case that people who
are so brought up are happier. But this judgement depends on empirical
assumptions about the way the world goes. How rare the tension is between
happiness and the demands of morality depends on how high the demand is,
and on what resources are available for meeting it. The demand of Kantian
morality is very high. Two examples may help with seeing this. The moral
demand includes the demand to love our enemies with what Kant calls
‘practical love’; we have to share their ends, as long as these ends are them-
selves morally permissible (so we do not have to share their end of harming
us). Also, we have to share the ends of the poor in the rest of the world who
could be helped by our lowering our standard of living and sending out the
proceeds. This does not mean we have to reduce ourselves to abject poverty;
but, even though it is a complex question just how much to reduce, it is very
likely that most of us in the developed world live too richly.5
Even if it were strictly true, however, that morality leads to happiness,
perhaps mediated (as Kant’s argument from providence proposes) by the
supersensible author of nature, this answer to the normative question would
not give us the right kind of motivation for a justification. ‘Eudaemonist’ is

5
I have tried to spell out some of the complexities in John E. Hare and Carey B. Joynt, Ethics
and International Affairs (London: Macmillan, 1982), 163–83.
66 God’s Command
Kant’s term for someone who says that ‘happiness is really his motive for
acting virtuously’.6 Eudaemonism is thus a single-source view of motivation;
all our motivation derives finally and properly from our own happiness. The
claim of the present chapter is that eudaemonism gives us a view of motivation
that is unacceptably self-regarding. We will look at four proposed defences of
eudaemonism against this claim, and reply to each of them. The first three will
end up compromising on the moral demand, and the fourth will compromise
on the aspiration for happiness.
The argument from providence comes from Kant, and endorsing it suggests
we have to adopt Kant’s account of what happiness is and what morality is.
His argument is, first, that morality becomes rationally unstable if we do
not have a way to assure ourselves that morality and happiness are consistent
(so that we do not have to do what is morally wrong in order to be happy),
and, second, that believing in God provides such assurance. But the terms
‘happiness’ and ‘morality’ are disputed. One typical response to Kant is that he
has too narrow a view of both; he has restricted happiness to the sum of
pleasures (as the utilitarians were initially to do), and he has restricted
morality to the sense of duty, where this requires the elimination of all singular
reference. Of course, the anti-Kantian says, if you set up the terms that way, it
will seem that there is a tension between morality and happiness, and a
problem that requires a deus ex machina for its solution. This objection has
weight, and Section 3.2.3 acknowledges that Kant’s account is defective in two
ways. But the argument from providence does not in fact need the defective
features of Kant’s account. Much depends here on how we set up these terms
‘happiness’ and ‘morality’. If we set them up so that ‘happiness’ is simply a
morally good life, we will not be able to get the argument from providence
started at all. All this argument needs, however, is that happiness is essentially
self-indexed, and that morality is essentially not self-indexed. Sections 3.2.1
and 3.2.2 are designed to clarify what ‘self-indexed’ means, but a rough picture
is as follows. If we take an ‘objective list’ account of the ends of life, such as Jim
Griffin’s list that includes ‘accomplishment, agency, understanding, enjoy-
ment, and deep personal relations’, and if we add to it spiritual ends such as
a relationship with God, and communal ends such as membership in a
community, we can then see that all these components require essential
reference to our own participation.7 If we want to hold that we are also
properly motivated by what is good in itself, independently of its relation to
us, this requires a double-source view of motivation: We are motivated both by
our own happiness and by what is good in itself independently of our

6
MdS vi. 377.
7
James Griffin, Well-Being: Its Meaning, Measurement, and Moral Importance (Oxford:
Clarendon Press, 1986), 67.
 Eudaemonism 67
happiness.8 Once we concede that point, we will see that there is no necessary
coincidence between morality and happiness, and that assurance of consist-
ency between the two requires a view about the governance of the universe
as a whole.
The substance of this chapter is an examination of four defences of eu-
daemonism. The first defence is from the Epicureans.9 This defence starts
from the pleasure we get in the pleasure of others, what Sidgwick calls
‘sympathetic pleasures’, and argues that there is a good sense of ‘for its own
sake’, where what is meant is ‘for the sake of the agent’s pleasure internal to
it’.10 The second defence is from the Stoics.11 The idea is that the notion of
reason brings impartiality with it, and so our good as rational beings requires
that we follow the moral law. The third defence is from a character I will call
Aquinas–Porter.12 Following the practice described in Chapter 1, the view will
not be attributed directly to Aquinas, but to a representation of his position by
Jean Porter. On this representation, Aquinas has a way to reconcile Aristotle’s
eudaemonism with the view that the distinctive mark of charity is loving God
for God’s own sake and promoting the good of the neighbour for the neigh-
bour’s sake and not our own. The key to the reconciliation is to postulate a
nested series of interests that is necessarily harmonious and includes the
agent’s own happiness within it. The fourth defence revises the third by
dropping the nested series, and it proposes instead that an agent perfects
herself by union with God, who is self-transcending. These four proposed
defences are not an exhaustive list. Philosophers and theologians have been
reflecting on this (in the tradition that starts from source-points in Greek
philosophy and the Jewish, Christian, and Muslim Scriptures) for about two
and a half millennia, and there are all sorts of permutations and combinations
that have entered the conversation. These four are among the most important.

3.2. THE S OURCES OF MOTI VATION

To explain the term ‘self-indexed’ we need to go back in the history of the

topic; first to Aristotle, and then to Duns Scotus. Aristotle gives us a single-
source account of motivation and Scotus a double-source account.

8
I am using the terms ‘single-source’ and ‘double-source’ to indicate the distinction between
views that derive all motivation finally from the desire for our own happiness, and views that do
not. There is a different though overlapping distinction, drawn in the same terms by Julia Annas,
where she attributes to the Stoics a ‘double-source’ view, meaning that we are motivated both by
self-concern and concern for others, The Morality of Happiness (Oxford: Oxford University
Press, 1993) (henceforth MH), 275.
9 10 11
MH 236–44. Sidgwick, Methods of Ethics, 499–500. MH 262–76.
12
Jean Porter, ‘The Desire for Happiness and the Virtues of the Will’, Journal of Moral
Theology, 3/1 (2014), 18–38.
68 God’s Command

3.2.1. A Single-Source View: Aristotle

The first sentence of Aristotle’s Nicomachean Ethics reads: ‘Every art and every
discipline, and similarly [every] action and rational choice, is thought to aim at
some good; and so the good has been aptly described as that at which
everything aims.’13 Aristotle identifies in this first chapter ‘the good’, which
he subsequently names as ‘happiness’ following ordinary usage, as an activity
of the agent. What does this mean? Action and rational choice are related in
the same way as art and discipline: in other words, rational choice controls
action. So there is a similarity between these two pairs of terms, but there is
also a contrast that Aristotle emphasizes in the next sentence: ‘But there seems
to be a difference in the ends; for some are activities, and others are products
apart from these.’ This distinction is repeated at the beginning of the next two
major subdivisions of book I, in chapter four and chapter seven, and it
structures the whole discussion. Both art and action have ends, Aristotle is
saying, but art has an end that is a product (in Greek ergon, which also means
‘characteristic work’ or ‘function’) apart from the activity of the art itself,
whereas the end of action is not separate from the activity in this way. He goes
on to give examples of the products of various arts (shipbuilding and bridle-
making and generalship) and he makes the point that the arts come in a
hierarchy corresponding to the hierarchy of their products. The bridle-maker,
if he wants to make a good bridle, has to consult the rider (who is the expert
about bridles because he is the user of them). The rider, if he wants to ride well,
has to consult the general (who is the expert about riding because he is the user
of the cavalry in battle). The general, if he wants to command well, has to
consult the statesman (in Greek, politikos); the statesman is the expert about
warfare because he is the user of the armed forces in the service of the city
(polis). This is how Aristotle gets to the conclusion that political science, the
knowledge-base of the statesman, is the most controlling and highest science,

13
I have taken the translation from Roger Crisp: Aristotle, Nicomachean Ethics, ed. and trans.
Roger Crisp (Cambridge: Cambridge University Press, 2000), except that I have replaced his
term ‘inquiry’ with ‘discipline’. I will use in this section exclusively masculine pronouns, because
Aristotle does so. I have given an account of some of the enormous secondary literature in my
God and Morality, ch. 1. The beginning of the first sentence contains two pairs of terms, ‘art and
discipline’ and ‘action and rational choice’. Aristotle uses ‘art’ (in Greek techne) broadly to cover
all kinds of know-how, not narrowly to cover art as opposed to science as we tend to use it.
‘Discipline’ is probably the knowledge base that controls an art, as the architect (a Greek word in
origin, which literally means ‘ruler-builder’) controls the stonemason. Aristotle characteristically
starts with the opinions of the many and the wise, or from common sense and the opinions of his
predecessors. So, in this first sentence, he starts with what ‘is thought [in Greek, dokei; literally,
“seems”] to aim at some good’, and he starts with a pair of terms that probably come from Plato.
See the commentary on this sentence by René Antoine Gauthier and Jean Yves Jolif, L’Éthique à
Nicomaque (Louvain: Publications Universitaires, 1970).
 Eudaemonism 69
since it uses the other sciences and dictates what must be done and avoided
in the city.
Action, on the other hand, has as its end happiness, and happiness is
activity in accordance with characteristic virtue (or excellence) and therefore
perfects the agent. Aristotle is going to distinguish different kinds of virtue,
but all of them are dispositions to act or feel or think as reason prescribes; so
the end of action is itself doing something or being active in a way that
manifests these dispositions. We can describe the status of the action or
activity in terms of how noble it is or how close it is to the divine.14 Suppose
in a battle I charge at the enemy when the general gives the command. I am
aiming at various ends, perhaps immediately the rout of the enemy. But
more broadly (Aristotle would say) I am aiming at what is noble (in Greek,
to kalon). And what is noble and divine is to be active in just this kind of
way for the sake of the polis. The good for the polis is more noble and divine
than the good for an individual. My action is thus an expression of its own
end; it is doing what a noble person does. The end is not something like a
ship or a bridle, which exists on its own when the artist who produces it is
finished with it.
Aristotle acknowledges that, even though the good is agreed (by both the
many and the wise) to have the name ‘happiness’, there are different accounts
of what happiness consists in. He himself mentions three, and he is probably
reflecting the Pythagorean picture of the three lives, which came to him
through Plato. Pythagoras distinguished three kinds of people who go to the
Olympic Games: the athletes, who go to compete; the businessmen, who go to
make money; and the spectators, who go to watch. By analogy, in Aristotle’s
account there is the life of somatic pleasure, the life of politics, and the life of
contemplation.15 All three of these candidates are activities of the agent.
Pleasure, on his account, is an activity or at least accompanies activity, political
life is constant activity (an activity that crowds out the nobler activity of
contemplation), and contemplation (which has the noblest and most divine
objects) is the activity of our highest part.16
Two reasons might be proposed for disputing this reading of Aristotle.
Someone might say that we act virtuously in order to attain what is noble, and

14
It is significant to note that the terms ‘god’ and ‘divine’ in their Greek forms occur in the
NE roughly twice as often as the terms ‘happy’ and ‘happiness’. See my God and Morality, 66.
15
NE I. 5. See Iamblichus, De Vita Pythagorica 58, and Plato, Republic 581c.
16
Nicholas Wolterstorff has two objections to eudaemonism, in Justice: Rights and Wrongs,
176–9. The first is that life-goods can be passivities as well as activities. His example is that ‘if
your reputation is ruined behind your back but everyone continues to treat you as they always
did, you have been deprived of a good to which you have a right even though your virtuous
activity has in no way been impaired’. But a eudaemonist can reply that your activity is impaired,
and is now less noble than it was; the value of the activity includes its good report or its honour.
For the second objection, see n. 21.
70 God’s Command
the term ‘noble’ involves no essential reference to the agent.17 And someone
might say that Aristotle ends up stretching the identity of the agent to include
his family and friends and (since his notion of friendship is more elastic than
ours) his fellow-citizens. These two reasons converge, because Aristotle thinks
it is noble to care for one’s friends for their own sake, and not for one’s own.
But, to test whether we genuinely escape essential reference to the agent in the
good pursued by the agent, we need to look at cases where there is tension
between the good of others and the agent’s own good, and see how Aristotle
adjudicates these cases. He gives us two paradigmatic examples of this. The
first is where to be noble requires death in battle for the sake of the polis. Here
Aristotle feels he has to give a justification in terms of the brave man’s reward
either by posthumous honour or by the brief moment of exaltation before
being killed.18 (He might also have made the point, though he does not, that, if
the brave man is not willing to die, he has to live the rest of his life with
dishonour or remorse.) Neither of these justifications takes us beyond essential
reference to the agent.
The case of friendship is even clearer. It is noble to enlarge our conception
of happiness to include the well-being of our households and friends for its
own sake. Aristotle uses the language of ‘a different himself ’ to talk first about
a father’s relation to his son, and then a virtuous friend’s relation to his
friend.19 The father loves the son as ‘a different himself ’ because the son
came from him, and the virtuous friend loves his friend as ‘another himself ’
because he relates to the friend and to himself in the same way (for example,
he likes to spend time with himself and with his friend and there is a basic
unity in the griefs and joys internal to himself and common to him and to his
friend). So the happiness of a good person will require the happiness of his
family and friends (broadly construed). But he will aim at their happiness only
to the extent that they have these special relations with him. Aristotle is not
proposing here that we value every human being as an end-in-itself or that our
own happiness counts morally no more and no less than anyone else’s. If we
are noble, we will have concern for the other for the sake of the other, but this
concern is conditional on the maintenance of the special relation. This limi-
tation is made vivid when Aristotle considers the question whether we wish
our friends to become gods.20 The reason we do not want this for our friends is

17
This is the claim of Troels Engberg-Pedersen, Aristotle’s Theory of Moral Insight (Oxford:
Clarendon Press, 1983), ch. 2.
18
NE IX. 8. 1169a18–b2: ‘He will choose a short period of intense pleasure rather than a long
period of mild pleasure, and choose to live nobly for a year rather than for many years in an
ordinary way, and choose one action which is noble and great rather than many small ones.’
Compare Scotus’s discussion of this case in Ord. IV, suppl. dist. 49, 9, 2.
19
NE VIII. 12 and IX, 4.
20
NE VIII. 7. 1159a5–12. See the commentary by Michael Pakaluk: Aristotle: Nicomachean
Ethics, Books VIII and IX, ed. and trans. Michael Pakaluk (Oxford: Clarendon Press, 1998).
 Eudaemonism 71
that, if they become gods, they will be too superior to us to remain our friends.
So we want the greatest good for them ‘as human beings’. Aristotle adds the
additional qualification that we do not strictly want the greatest goods for our
friends, ‘since it is for himself most of all that each person wishes what is good’.
He insists (though later peripatetics were to drop this insistence) that virtue
does not leave the sphere of self-love.21
Two distinctions need to be made here. The first is often associated with
Bishop Butler.22 There are two senses in which every good aimed at by an
agent might be a good for the agent, and the first does not imply the second.
The first sense is that the good aimed at is good for the agent just because the
agent aims at it. In this sense, the good aimed at might not itself contain any
relation to the agent beyond that of being aimed at by means available to the
agent. I could aim, in Scotus’s example, that God have everything good. In the
second sense, the good for the agent is an object whose definition (even if not
explicitly articulated as such by the agent) includes internal reference to the
agent, as in Aristotle’s example ‘the good of my friend’. Sometimes people
make the mistake of arguing from the premiss that we always move towards
what we perceive as good for us in the first sense to the conclusion that we
always move towards what we perceive as good in the second.
The other distinction is implicit in the description of the first. We should
not insist that the good for the agent (in the second sense) be articulated as
such by the agent. Aristotle does not settle this question.23 He gives the
analogy of a racetrack, assuming that racetracks end on the same line where
they begin and go round a turning post in the middle.24 We start off with
certain principles we inherit from our parents and society, and, because we are
human, and humans by nature desire to know, we ask questions about these as
we get older.25 We reach the turning point when we acquire a vision of the
good human life, knowing what virtue is and what happiness is. Then we can
race back to the original line (where the judges sit), but now with the principles
not merely secondhand but fully appropriated as our own. This analogy does
not require, however, that the vision of the good be explicit or articulated. For
most of us it is probably not. Even without this articulation, it can shape the
lives we try to lead.26 Aristotle’s example of my friend becoming a god shows
that I do not want the best thing for my friend (because we will not be friends

21
NE IX. 8. 1168b25–1169a37. See MH 257. Wolterstorff ’s second objection to eudaemonism,
Justice: Rights and Wrongs, 177, is that ‘if you have a right against me . . . whether or not
performing that action would make for greater happiness on my part is simply irrelevant to
what I should do’. ‘Simply irrelevant’ is too strong, because I have to treat myself also as an end,
but the objection is correct that my happiness does not determine what I should do.
22
Butler, Fifteen Sermons, 126–7.
23
See Sarah Broadie, Ethics with Aristotle (Oxford: Oxford University Press, 1991), 198–202.
24
NE I. 4. 1095a32–b13. 25
Metaphysics I. 1. 980a1.
26
We return to this in discussing John Haidt’s views on ‘dumbfounding’ in Section 8.2.2.
72 God’s Command
any longer), but I may not express to myself that my desire for my friend’s
good is conditional on the maintenance of the special relation, even though it
is in fact conditional in this way.
Summing up, we can say that Aristotle gives us a single-source view of the
motivation of an agent; the source is the agent’s happiness, understood as a
perfecting activity of the agent. This is good ‘for the agent’ in the second of the
two senses distinguished by Butler: the object pursued has essential reference
to the agent, not merely because it is what she is pursuing, but in its own
definition. The self-indexed good does not, however, require that the agent
articulate it as self-indexed. The claim of the present chapter is that a single-
source view of the motivation of an agent is a mistake, but that Aristotle is
right nonetheless to say that we start from self-preference. This is not because
we are human, however, but because of a disorder of our wills. It is not
necessary for humans to prefer themselves in this way.

3.2.2. A Double-Source View: Scotus

We can go on now to a contrasting view, a double-source view of motivation.

Duns Scotus accepts from Anselm that there are two basic affections of the
will, what Anselm calls ‘the affection for advantage’ (affectio commodi) and
‘the affection for justice’ (affectio justitiae). The affection for advantage is an
inclination towards our own happiness and perfection. The affection for
justice is directed towards what is good in itself, regardless of its relation to
us. Aristotle’s account of motivation has nothing corresponding to the affec-
tion for justice; we do everything that we do for the sake of our own happiness,
even if we do not represent this to ourselves as such. Since, for Scotus, we have
both affections, we face the question of how to rank them. He is not proposing
that there is anything wrong with the affection for advantage.27 Even in
heaven, we will have both affections. The affection for advantage becomes
wrong only when it is ranked improperly. He presents the affection for justice
as required by the divine will to be a ‘checkrein’ or moderator of the affection
for advantage:
A free appetite is right in virtue of the fact that it wills what God wills it to will.
Hence, those two affections, the affection for advantage and the affection for
justice, are regulated by a superior rule, which is the divine will, and neither of
them is [in itself] the rule for the other. And because the affection for advantage
on its own is perhaps immoderate, the affection for justice is bound to moderate

27
On my reading, Scotus’s view is different from that of Timothy Jackson, Love Disconsoled
(New York: Cambridge University Press, 1999), 6, that Christians are to ‘move from reciprocity
to the Golden Rule to the cross of Christ’. Self-sacrifice (the cross), if made into a higher goal, is
(paradoxically) a self-indexed good.
 Eudaemonism 73
it, because it is bound to be under a superior rule, and that rule . . . wills that the
affection for advantage be moderated by the affection for justice.28
Aristotle does not have the idea of a divine will requiring us to live a certain
way. As quoted in Chapter 2, he denies that God gives us commands: ‘For the
god is a governor not in a prescriptive [or commanding] fashion, but it is that
for [the sake of] which practical wisdom prescribes.’29
Chapter 1 gave the example of the depressing parent to illustrate how these
two affections might come apart. Another way to show the distance between
them is counterfactually. If God were to require us, which fortunately God
does not, to sacrifice even our own salvation for the sake of God’s glory, then
we should be willing to do so. This thought requires a certain view about God’s
election: God is not required by necessity to elect all human beings for
salvation. It is a view common to Aquinas and Calvin that God not only can
but does elect some for union with God (predestination), and some for
separation (reprobation). But all that is needed for the thought experiment
is that God can, not that God does. Samuel Hopkins gives expression to the
thought: ‘But to him who loves God supremely, and desires his glory above all
things, it is so far from being impossible to be willing to be damned, on
supposition this is most for God’s glory, that he could not will or choose any
thing else.’30 He is repeating a pattern of thought that can be found in Moses,
who says that he is willing to be blotted out of the book of life, and Paul, who
says that he is willing to become a curse.31 Jesus too accepts separation from
his Father as the price for saving his people, which was the declared motiv-
ation of both Moses and Paul.32 In the case of Hopkins’s thought experiment,
the point is not that God actually requires us to choose between the divine
glory and our salvation, but that God could do so; and the thought about what
we would choose, if God did require this, is an indication of how we rank the

28
Scotus, Reportatio, II, d. 6, q. 2, n. 10, trans. Thomas Williams, in ‘From Metaethics to
Action Theory’, in Thomas Williams (ed.), The Cambridge Companion to Duns Scotus (Cam-
bridge: Cambridge University Press, 2003), 346. This passage was given the wrong reference in
my God and Morality, 97. Other translations of Scotus are taken from Duns Scotus on the Will
and Morality, ed. William A. Frank, trans. Allan B. Wolter (Washington: Catholic University of
America Press, 1986). If we ask, ‘Why is the affection for justice higher than the affection for
advantage?’, the answer is that the highest object of the first (namely, God) is higher than the
highest object of the second (namely, our union with God). This is consistent with saying that the
first is not in itself higher than the second, because the second could have an object higher than
some object of the first.
29
EE VIII. 3. 1249b14–15.
30
Samuel Hopkins, The Nature of True Holiness (1770), in The Works of Samuel Hopkins, iii
(Boston: Doctrinal Tract and Book Society, 1852), 147. Jonathan Edwards refers to people who
say this, but does not himself say it (Religious Affections, ed. John E. Smith (New Haven: Yale
University Press, 1959), ch. X).
31
Exodus 32: 32; Romans 9: 3. There is a story also about Abu Bakr, who said he was willing
to be put in the mouth of the cave that leads to hell, so as to stop others from getting in.
32
Matthew 27: 45: ‘My God, My God, Why have you forsaken me?’, quoting Psalm 22: 1.
74 God’s Command
two affections of the will. There is a partial analogy with purely inter-human
relations. Suppose my love for my son requires me to act in a way that destroys
my relation with him. I might do this for his good, but in a way that shows
I am putting him above my relation to him.
Introducing a double-source view of motivation allows us to understand a
bit better the fall of Lucifer, though this and our own fall finally remain, in
Kant’s term, ‘inscrutable’. Scotus distinguishes three kinds of love we can have
for God: love for God independently of any relation to us, love of union with
God, and love of the satisfaction we get from that union. Lucifer started from
the second of these, which is indeed something good in itself though self-
indexed, and came to love it inordinately as his own advantage. According to
the argument from grace outlined in Chapter 1 we humans are now born with
this wrongful ranking of the affection for advantage, and it can be reversed
only by God’s assistance. The final chapter considers the relation of this claim
to contemporary evolutionary psychology.
Scotus draws a connection between the two affections and freedom of
the will. It is interesting that Aristotle, as well as having no doctrine of the
affection for justice in Scotus’s sense, also has no doctrine of freedom of
the will.33 Scotus reports the thought experiment of Anselm about an angel
who has the affection for advantage but not the affection for justice.34 He says
that such an angel would be unable not to will what is advantageous, and
unable not to will it above all. But then this would not be imputed to the angel,
because the pursuit of the advantageous would be triggered automatically by
apprehending it in the intellect, just as bodily appetites are triggered by sense
perception. For the angel to be held accountable, it would have to have both
affections, since the affection for justice is what opens up the possibility of
not pursuing what is most to our advantage, or at least the possibility of not
pursuing it above all. Scotus concludes that the affection for justice ‘is the
liberty innate to the will, since it represents the first checkrein on this affection
for the advantageous’.35
Scotus says, as quoted earlier, that neither of the affections is (in itself) the
rule for the other, but it is the divine will that is the superior rule that binds the
affection for justice to moderate the affection for advantage. On the other
hand, he says, ‘the moral goodness of the act consists mainly in its conformity
with right reason, which dictates fully just how all the circumstances should be
that surround the act’.36 By ‘right reason’ he means to include our right reason,
and it is tempting to conclude from this and similar passages that Scotus is

33
This claim is open to doubt. Anthony Kenny has a book entitled Aristotle’s Theory of the
Will (London: Duckworth, 1979), in which he disputes it.
34
Anselm includes the reverse possibility, that the angel might have the affection for justice
but not the affection for advantage, but Scotus does not mention this because he is assuming in
this discussion that humans and angels have the affection for advantage.
35 36
Scotus, Ord. II, dist. 6, q. 2. Scotus, Ord. I, dist. 17, q. 62.
 Eudaemonism 75
saying divine command is not necessary for the moral goodness of an act
(since our right reason is sufficient), and that therefore Scotus is not a divine
command theorist at all.37 But there are some distinctions to be made here,
developed more fully in the next chapter. The first is between a theory of value
and a theory of obligation. Goodness is possessed by anything that takes us to
our end, but God has discretion over which route to this end, and so which
good things to require. Only what God commands has the authority of
obligation. A second distinction is between our knowledge of moral goodness
or obligation and our knowledge of what makes them good or obligatory. It is
possible that what makes something good or obligatory is some relation to
God (different in the two cases), but that we can know by right reason that the
thing is good or obligatory without knowing this relation. On some versions of
the doctrine of general revelation, God can reveal that some route to our end is
required of us without our knowing that it is God who requires it. A third
distinction is between harmony or fittingness with nature and implication
from nature. If God does command what fits our end, we can expect to see a
harmony between this route and our end (or our nature in the sense of our
end). We can expect to be able to tell a story about how, for example, we tend
to flourish when we honour our parents, and refrain from murder, adultery,
theft, lying, and coveting. But Scotus insists that what we see here is a
harmony, or a beauty, or a fittingness, and not an implication from our nature.
When we put these three distinctions together, it is plausible to say that he
thinks it is God’s command that makes something morally obligatory.

3.2.3. Two Errors of Kant

We have seen two sketches so far, one of a single-source view of motivation

(Aristotle), and one of a double-source view (Scotus). Eudaemonism is a
single-source view. Before moving on to discuss and counter four defences
of eudaemonism, we need to face two difficulties with Kant’s account of
morality and happiness. Fortunately we can modify Kant’s own account in
order to overcome these difficulties without losing the argument from provi-
dence with which we began in Chapter 1, and the modification will remove
some distractions. The first difficulty is with Kant’s account of happiness, and
the second with his account of morality. Both problems come from Kant’s over-
strict dichotomies. Happiness, he says, is ‘a rational being’s consciousness of

37
This is the view of Richard Cross, Duns Scotus (Oxford: Oxford University Press, 1990), 90.
The first two distinctions in the present paragraph are not drawn in just these terms by Scotus,
but he does have a distinction between goodness and law, and he does have a notion of general
revelation. It is plausible that he has something similar to the present paragraph in mind when he
says both that God’s will and that our right reason determine what we should do.
76 God’s Command
the agreeableness of life uninterruptedly accompanying his whole existence’,
and he goes on to say that ‘to make this the supreme ground for the deter-
mination of choice constitutes the principle of self-love’.38 Kant here ties
happiness to pleasure, pleasure that derives from the satisfaction of one’s
inclinations as a sum. In turn, inclinations are defined in terms of the lower
faculty of desire, and their satisfaction is something empirical, and cannot
therefore determine practical (necessary) laws. To try to make happiness the
ground of morality would therefore be to lose morality altogether. Despite first
appearances, however, Kant is against what he calls a ‘morose’ ethics, which
sets morality in opposition to all pleasure and renounces all concern of moral
persons for their own happiness.39 He distinguishes, it is true, between what he
calls ‘practical love’ (the will’s obedience to the moral law) and ‘pathological
love’ (which is a feeling such as sympathy and compassion), and denies that
the latter can be commanded, whereas the former is ‘the kernel of all laws’.40
This might lead one to think that the state that the moral law commands is one
in which inclinations do not appear at all.41 But in fact Kant says that to love
God with practical love (which can be commanded) means to do God’s
commandments gladly, and that to love one’s neighbours means to practise
all duties toward them gladly.42 He has in mind a translation of the theological
doctrine of sanctification, in which our inclinations become over time more
and more in line with duty. In the resultant state, our wills will be in
conformity to duty for its own sake, and this deserves what Kant calls ‘esteem’,
and our inclinations will also conform to what duty requires, and this deserves
what he calls ‘praise and encouragement’.43 But merely including inclinations
in this way is not enough. There are two revisions we need to Kant’s account of
happiness. The first is that Kant needs to acknowledge a kind of ‘gladness’ that
is not merely the satisfaction of sensuous inclination. He needs an account of
not-purely-sensuous moral pleasures, such as the awe we feel in the presence
of the moral law within, or a delight in goodness that is like the astonishment
at the wisdom displayed in the order of nature, an affect ‘stimulated only by
reason’.44 But this kind of ‘higher’ pleasure is never properly integrated into
Kant’s account. The second revision is that it is better not to insist on tying
happiness to pleasure at all, even if we continue to index the content of
happiness to the agent. There are many self-indexed goods, such as accom-
plishment in Griffin’s list mentioned earlier, which are only derivatively
pleasures. That is to say, we get pleasure from them but only because we
antecedently think of them as good.

38 39 40
KpV v. 22. MC xxvii. 302–4. KpV v. 83.
41 42
Gl. iv. 399. KpV v. 83.
43
Gl. iv. 398–9. See Barbara Herman, The Practice of Moral Judgment (Cambridge, MA:
Harvard University Press, 1993), 2–40.
44
KpV v. 161, Anth. vii. 261.
 Eudaemonism 77
The difficulty with Kant’s account of morality is that he holds that motiv-
ation is either by self-indexed inclination or by universal moral principle. This
dichotomy is a mistake, and it is interesting that Scotus makes very much the
same mistake.45 After elaborating the distinction between the two affections,
Scotus proceeds to argue that every motivation that is for justice rather than
advantage is for God, and so the choice is always: God or self. But surely I can
be motivated to achieve something for Peter (perhaps something simple, like
his getting safely across the street), without this being self-indexed by my
caring essentially either that Peter is in some special relation to me or that the
result be achieved by me. My motivation here is indeed indexed to a particular
(to Peter) and I may not be motivated to pursue similar good things for other
similar people. On Scotus’s dichotomy, this motivation does not belong under
either the affection for justice or the affection for advantage. In the same way
Kant holds that the first formula of the Categorical Imperative, the formula of
universal law, requires the eliminability not only of self-reference, but refer-
ence to any particular person. But we need an intermediate category, of
inclinations that are not universal (since they contain ineliminable reference
to an individual), but that are indexed not to the self but to some other
individual. This is not only a terminological question, whether to call a
principle ‘moral’ if it contains ineliminable reference to, for example, Peter.
There is a substantive question about whether to have the highest kind of
admiration (what Kant calls ‘esteem’) for a person who acts on such a
principle. The fifth chapter, in the course of defending Barth’s particularism,
defends the view that we should. One way to put this point is that the two
formulations of the Categorical Imperative discussed in Chapter 1 can come
apart on one plausible interpretation of the second (though not on Kant’s
own interpretation of it). It is possible to care for another person as an end in
herself but not be willing to eliminate reference to her from the maxim of
one’s action.

3.3. FOUR ATTEMPTED D EFENCES

OF EUDAEMONISM

The rest of this chapter considers four defences of eudaemonism, and rejects
them all. The first is an Epicurean defence, the second a Stoic defence, the
third a Thomist defence, and the last a defence through the notion of self-
transcendence.

45
Scotus, Ord. III, suppl. dist. 46.
78 God’s Command

3.3.1. The First Defence: Epicurean

The first defence of eudaemonism against the charge that it is unacceptably

self-regarding derives from the Epicurean tradition, which identifies the good
with pleasure. There is an important division within the hedonist tradition
between what Sidgwick calls ‘egoistic hedonism’ and ‘universalistic hedon-
ism’.46 The egoistic hedonist proposes that the agent should think about her
own pleasures, and the universalistic hedonist proposes that she should think
about the pleasures of all those affected by her decision, and count those
people as worth the same as herself in the calculation about what to do.
(Sidgwick thought she should also count the pleasures of all sentient beings
affected, but that question goes beyond the scope of this book.) Both kinds of
hedonist have in common that it is pleasures the agent should think about, and
that these pleasures constitute happiness. Sidgwick proposed that one way to
try to bring the two kinds of hedonism together is to consider the sympathetic
pleasures we get from the pleasures of others. The hope, which he finally
rejected, was that, if we include these sympathetic pleasures, even the egoistic
utilitarian will be led to care about the well-being of others. This is the first
defence that a eudaemonist might make, and there is a version of it already in
Epicurean sources:
We cannot maintain a stable and lasting enjoyment of life without friendship; nor
can we maintain friendship itself unless we love our friends no less than we do
ourselves. Thus it is within friendship that this attitude is created, while at the
same time friendship is connected to pleasure. We delight in our friends’ happi-
ness, and suffer at their sorrow, as much as we do our own.
Hence the wise will feel the same way about their friends as they do about
themselves. They would undertake the same effort to secure their friends’ pleas-
ure as to secure their own.47
Cicero’s Epicurean spokesman in this passage, Torquatus, adds a couple of
other points. The friendship marked by equal regard for the friend may
develop gradually, as frequency of association leads from a merely instrumen-
tal friendship to real intimacy, and it may sometimes involve something like a
pact or agreement to love one’s friends as much as oneself. But the basic point
is the first one, that pleasure as our chief good should be expanded to include
the pleasure we get from the pleasure of our friends.48 With some kinds of

46
Sidgwick, Methods of Ethics, 497.
47
Cicero, De Finibus I. 66–8, and II. 82–4. The translation is taken from Cicero, On Moral
Ends, ed. Julia Annas, trans. Raphael Woolf (Cambridge: Cambridge University Press, 2001).
48
There is a question in the dialogue whether these views are consistent with Epicurus’
foundational principle that ‘friendship is sought on the basis of utility’, and whether, if we add in
this equality of regard, and these pacts, we have in fact conceded (though not in words) the point
that not everything is pursued for the sake of pleasure. But Cicero is being unfair here, and on
 Eudaemonism 79
pleasure such as pleasure in friendship, pursuing something for its own sake
(such as my friend’s good) and for its particular kind of pleasure amount to the
same thing, and with these pleasures we can give a plausible account of a good
life that includes concern for the good of others for its own sake.
Note that this first eudaemonist defence does not have to be put in terms of
pleasure. Some utilitarians moved from an emphasis on pleasure to an em-
phasis on happiness, because there seemed to be ingredients of happiness that
are not in any obvious way pleasures, and they thought we should be maxi-
mizing those ingredients as well. And some utilitarians moved beyond an
emphasis on happiness, because that term also seemed too value-laden, to an
account in terms of the maximization of the satisfaction of preferences. Thus
RMH pointed out in Freedom and Reason that the term ‘happiness’ does not
merely describe a state as satisfying most of the happy person’s preferences,
but evaluates it in terms of whether the satisfaction of those preferences is
enough for a good life.49 ‘Few of us’, he thought, ‘would say that an opium
addict was happy (really happy) if he always got enough opium’. The essential
form of the first eudaemonist defence is that an agent’s own good, whether this
is defined in terms of pleasure or happiness or preference-satisfaction, can be
structured in a complex way, so that it contains both merely one-at-a-time
goods and life-as-a-whole-affirming goods (what Torquatus calls ‘a stable and
lasting enjoyment of life’); the former can be evaluated by their contribution to
the latter. Some goods (such as friendship) have leverage over one’s life,
making it worthwhile as a whole.50 The point of the first eudaemonist defence
is that objectors to eudaemonism focus on the merely one-at-a-time goods,
and fail to see the resources of the life-as-a-whole-affirming goods for
addressing the objection that eudaemonism is unacceptably self-regarding.
Not all of these latter have to be pleasures. We can put the point in terms of
Scotus’s distinction between the three different kinds of thing we want in
loving God: wanting God to have everything good, wanting union with God,
and wanting the satisfaction that comes from union with God. The first
eudaemonist defence does not need to rely only on the third kind, but can

some interpretations of the text he knows he is being unfair. See C. Brittain, ‘Cicero’s Sceptical
Methods: The Example of the De Finibus’, presented at the 12th Symposium Hellenisticum,
‘Cicero’s De Finibus: A New Appraisal’, September 2010.
49
RMH, Freedom and Reason (Oxford: Oxford University Press, 1977), 125–9. In Moral
Thinking this point was made in terms of ideals as opposed to basic preferences, and RMH
conceded that he did not know how to integrate ideals into the structure of his defence of
utilitarianism. I discuss this in Section 4.3.3.
50
Philosophers disagree about what these goods are, and there is considerable variety
between different people in what ‘matters’ to them in this way. See Griffin, Well-Being, ch. 4,
for a list of ‘objective goods’. See also Derek Parfit, On What Matters, 2 vols (Oxford: Oxford
University Press, 2011), for a sustained defence of the claim that there is such a list.
80 God’s Command
work with the second kind as well. If, however, it moves to the first kind, it will
no longer be eudaemonist.
The strategy in this first eudaemonist defence is to distinguish two different
ways in which we can enjoy something ‘for its own sake’. In one way, if
something is loved for its own sake, there cannot be anything at all for the
sake of which it is loved. The analogy with music is helpful here. Suppose I say,
‘I love Mendelssohn’s Octet for its own sake’, but someone objects, ‘Don’t you
get pleasure from the piece?’; and suppose that, if I say I do, he objects that this
shows I do not love the Octet for its own sake, but for the sake of the pleasure.
There is something wrong about this objection. One way to see this is to see
that the counterfactual question, ‘Would you still love it if you did not get the
pleasure?’, does not make good sense. The proper and standard way of loving
music is to get pleasure of a certain kind (‘aesthetic’ pleasure) from it. If I did
not, I would not be loving it, except perhaps for some alien good such as its
capacity to annoy someone I dislike.51 When I get the proper kind of pleasure
from it, the pleasure is not something else, or something external, for the sake
of which I love the music. For the counterfactual question to be well formed,
the pleasure for the sake of which I love something has to be a pleasure shared
with other activities, and external to them, so that I can love all of them for the
sake of it. Then I can think about the question whether I would still love them
if I did not get this pleasure from them.
The second way we can enjoy something ‘for its own sake’, then, is when there
is nothing external to it for the sake of which it is loved. This point can be
generalized. Pleasures come in two different kinds, as do ingredients of happi-
ness and preference-satisfactions. There are what was called earlier ‘one-at-a-
time’ goods and ‘life-as-a-whole-affirming’ goods. When I love my friend for
her own sake, this is like loving music for itself; both of these loves are perfectly
compatible with, indeed they require, getting a certain kind of pleasure or
satisfaction, and loving that satisfaction. The two loves also have the capacity
to have leverage over one’s life as a whole. Life-as-a-whole-affirming goods
characteristically have instances (for example, hearing a performance of the
Octet) such that the instance is loved both for its own sake and for the sake of
the life-as-a-whole-affirming good, which belongs with it internally. The first
eudaemonist defence argues that in loving my friend for her own sake I am also
loving her for the sake of my happiness, and there is nothing paradoxical in this
because her well-being is internal to, or an ingredient in, my own.
We can appeal here to two different levels at which we do practical
thinking.52 We operate most of the time at an intuitive level with principles

51
Aristotle discusses the relation between proper and alien pleasures at NE X. 5. 1175a29–b13.
Also, see the point I attributed to Butler earlier.
52
Annas (MH 242 n. 51) says that RMH’s later moral philosophy developed the two-level
strategy. I have, indeed, drawn the terms ‘intuitive’ and ‘critical’ from him. See Moral Thinking, 34.
 Eudaemonism 81
that we do not think out from scratch. But, when we have leisure, we can try to
work out critically what principles or intuitions we should live by. We could
call this higher level ‘the critical level’. This strategy works for the first
eudaemonist defence because we can say that the self-referential eudaemon-
ism comes into play only at the critical level. Most of the time we live at the
intuitive level, and at this level we can think entirely about the well-being
of our friends or of other people. In this way we can accommodate both
self-regard and other-regard into the good human life, without leaving a
eudaemonist framework behind.
Julia Annas objects that this kind of two-level theory is ‘schizophrenic’:
‘While helping my friend I would bear in mind only my friend’s needs, while
with another part of my mind remaining aware that the point of all this
activity in the first place was simply to obtain pleasure for me.’53 She is here
repeating an attack by Bernard Williams, who says that a parent who reasoned
that everyone is better off if parents give preference to their own children
(a critical-level reflection) would have ‘one thought too many’.54 But this
objection is unfair. The two-level account is not supposed to be an account
of two simultaneous pieces of reflection. Consider a good marriage, in which
one spouse cares about the other for her own sake. There is nothing schizo-
phrenic about his sometimes reflecting on the benefits to himself of the
relation between them. There is also nothing schizophrenic about parents
thinking sometimes about the benefits to the world as a whole of parents
feeling special obligations to their own children.
There is, however, something troublesome about the application of the two-
level picture to a defence of eudaemonism, rather than to an analysis of
morality more broadly. Here we return to Sidgwick. The question is how
much concern for others the higher or critical level will let through. Suppose
we concede that it will endorse principles at the intuitive level that call for
loving family and friends for their own sake.55 The problem is that the critical
level is still by hypothesis eudaemonist, and, when I consider the interests of
others beyond the family and friends, it will not make all that much room for
them. Even if we grant Aristotle’s claim that fellow-citizens are friends, a claim
much more plausible for him than for us in view of the size of the Greek polis,
what about those in need in the rest of the world, whom we could help if we
cared about them? The situation is worse than this question suggests. We have
a limited capacity for caring. Even if the eudaemonist critical level endorses

But his use of a two-level theory was not in the service of defending a eudaemonist theory of
motivation. On this point, he was Kantian.
53
MH 241.
54
Bernard Williams, ‘Persons, Character, and Morality’, in Moral Luck (Cambridge: Cam-
bridge University Press, 1981), 18.
55
Hursthouse discusses this point at OVE 131–2.
82 God’s Command
principles of non-self-indexed concern for family and friends at the intuitive
level, this will itself diminish the caring we can do for those outside these
limits. Sidgwick puts the matter this way:
Or again, a man may find that he can best promote the general happiness by
working in comparative solitude for ends that he never hopes to see realized, or
by working chiefly among and for persons for whom he cannot feel much
affection, or by doing what must alienate or grieve those whom he loves best,
or must make it necessary for him to dispense with the most intimate of human
ties. In short, there seem to be numberless ways in which the dictates of that
Rational Benevolence, which as a Utilitarian he is bound absolutely to obey, may
conflict with that indulgence of kind affections which Shaftesbury and his
followers so persuasively exhibit as its own reward.56
Suppose we grant that the satisfaction we get from seeing the happiness of
those we love is capable of exerting significant leverage on whether our lives
seem to us worthwhile on the whole. Nonetheless there are needs of people
and indeed (for Sidgwick) of all sentient beings beyond these limits that we
both can and should try to meet. The point is not that morality requires us to
equalize the care we give to all those in need. Promoting the general happi-
ness allows giving some additional weight to those with whom we have
special relations, though it is a very difficult task for normative ethics to
say just how much.57 Also, there are institutions, like the church, that can
send members of the congregation to places in the world where there is need,
and those villages can become, as it were, our villages according to what
Chapter 2 called ‘the principle of providential proximity’. Nonetheless, we
seem to reach natural limits of human caring, and we almost certainly have
moral obligations that go beyond those limits. Sidgwick’s points still apply.
Benevolence is different from music in this way. Pieces of music do not have
interests. When I spend resources of time and energy on Mendelssohn’s
Octet, I am not thereby depriving other pieces of their interests. But there are
people I should be helping whether this helping gives me a sense that my life
as a whole is more worthwhile or not. If the eudaemonist responds that my
sense of a meaningful life is not what counts, but rather the degree of my
perfection, we have gone beyond the limits of the Epicurean defence. We
will discuss what we can call ‘self-transcendent eudaemonism’ at the end of
this chapter.

56
Sidgwick, Methods of Ethics, 503.
57
I have considered what I call ‘the obligation of radical sacrifice’, the thesis that the wealth
of a rich individual or a rich nation should be given away until the point where personal or
domestic need is as great as that of the people to whom the money or food might be sent, and
I have raised eleven objections to it. See Hare and Joynt, Ethics and International Affairs,
163–83.
 Eudaemonism 83

3.3.2. The Second Defence: Stoic

We turn next to a different defence that the eudaemonist might make, one that
derives from the Stoics.58 The idea is that the notion of reason brings impar-
tiality with it, and so our good as rational beings requires that we follow the
moral law. Zeus, says Epictetus, ‘has set up a rational creature’s nature in such
a way that it cannot achieve any of its own goods without contributing to the
common benefit’.59 Julia Annas glosses this as follows: achieving the good of
all, as you do if you act impartially, just is achieving your own good properly
understood.
The key is ‘properly understood’. The good or ‘benefit’ has to be distin-
guished from the merely ‘convenient’, and the reason goods, when properly
understood, are shared or common is given in the following statement by
Arius Didymus:
Goods are all common among the excellent, and evils among the bad. Hence the
person who benefits someone is himself benefitted, and the person who harms
someone harms himself. All excellent people benefit one another, even when they
are not totally friends to one another, nor even well-disposed nor in good repute
nor accepted as such, on account of not being apprehended as such and not living
in the same place. However, they are so disposed as to feel goodwill, friendship,
approval and receptiveness to one another.60
We have from the Stoics, in a sense, a ‘double-source’ view of motivation. They
distinguish between motivation by self-concern and motivation by concern for
others. Both of these develop (though the second itself comes later in our
development than the first). They have a technical term oikeiosis for this
development, which we might translate ‘appropriation’, since oikeion means
‘one’s own’ like the Latin proprium (which is the root of ‘appropriation’).61 We
can distinguish accordingly between two kinds of appropriation, ‘personal
appropriation’ and ‘social appropriation’. The first is the development that
takes us from self-preservation to valuing reasoning as a way to get the things
that fit our nature, such as health and wealth, and then to valuing reason in its
own right.62 The second is the development through progressively wider social
groupings, until one cares for every human being; caring about parents and
immediate family develops into caring about extended family, fellow-demesmen,

58
MH 266.
59
Epictetus, Discourses II. 19. 13–14, cited in MH 274, along with the following gloss.
60
Arius Didymus, in Stobaeus Eclogae II. 101.21–102.2, cited in MH 266. We need to add the
distinction between ‘benefit’ and ‘convenience’, from, e.g., Cicero De Finibus III, 69.
61
Annas (MH 262) prefers ‘familiarization’, following a suggestion of Jonathan Barnes, but
the term has other uses that get in the way of its intelligibility in this context.
62
Cicero puts a description of this development into the mouth of his spokesman for the
Stoics, Cato, De Finibus III. 20–1.
84 God’s Command
fellow-tribesmen, fellow-citizens, citizens of ‘cities’ nearby, members of the same
ethnic group, and finally the whole human race.63 One image for this is a series of
concentric circles, like those in a pool when a stone is thrown into it. One
suggested psychological technique for moving through these circles is to call
people in a farther circle by the name or term appropriate to a closer circle; one
calls an aunt ‘mother’. The idea is to associate a feeling with the name, so that one
starts to care about one’s aunt in the way one cares about one’s mother. But the
movement through the circles is a matter not primarily of psychological tech-
nique, but of reason.
This is a double-source theory in a sense, but not in the sense introduced at
the start of this chapter. This is because the Stoics were eudaemonists, and held
that both forms of appropriation enable one to see better a single thing, one’s
own chief good, which remains the source of all motivation. The argument
that holds the personal good and the social good together, so that happiness
remains a single thing, is the argument given in the quotations from Epictetus
and Arius Didymus. This is the argument, therefore, that we have to evaluate,
and it is not very strong. It is true that, if I already see my aunt, for example, as
worth the same as my mother, then I will see that a benefit to my aunt is worth
the same as a benefit to my mother. But, if I do not already see this, it is unclear
how reason can require me to see it. What the argument needs to show is that
it is irrational to prefer benefiting those near to benefiting those far. Arius
Didymus claimed that all excellent people benefit one another, even if they do
not know each other, but this sense of mutual benefit requires that the interest
of the other be already my interest. This is indeed what a Stoic or Kantian
morality requires. But the question is why reason should move me to it. This
objection goes all the way back to the question why reason should prevent me
preferring benefiting myself to benefiting someone else. Sidgwick again makes
the point elegantly:
‘I’ am concerned with the quality of my existence as an individual in a sense,
fundamentally important, in which I am not concerned with the quality of the
existence of other individuals: and this being so, I do not see how it can be proved
that this distinction is not to be taken as fundamental in determining the ultimate
end of rational actions for an individual.64
A contemporary philosopher who makes an argument rather similar to the
Stoic one I have just described is Peter Singer. He uses the same picture of an
expanding circle of care (extending to all sentient beings), and in the service of
this picture he proposes the analogy of an ‘escalator of reason’.65 On this

63
This is the sequence described by Hierocles, in Stobaeus, Florilegium 84. 23.
64
Sidgwick, Methods of Ethics, 321–2.
65
Peter Singer, The Expanding Circle: Ethics, Evolution, and Moral Progress (New York:
Farrar, Straus and Giroux, 1981), 88–124. I have discussed Singer’s views both in God and
 Eudaemonism 85
analogy, I start from believing that I have a reason to promote my own
interests. But then I shall, as a social being, wish to justify my conduct to
others. This requires me to step, as it were, onto the escalator. But, as I do so,
I will find the fact irrelevant that it is I who benefit from some distribution, for
example, and you who lose by it. I will find myself moving towards the
standpoint of the impartial spectator or ideal observer. From this standpoint,
which Sidgwick calls the point of view of the universe, it makes no difference
to the moral decision whether the interest being served is mine or someone
else’s. Having said this, however, Singer also accepts from Sidgwick that reason
allows me to get off the escalator at any point. He denies the argument he
attributes to Kant that reason by its nature requires universalizing, or willing
the maxims of our actions as universal laws.66 The point is that Singer
acknowledges that we cannot get a convincing argument from the nature of
reason itself for moving to the moral point of view.67 There is nothing
irrational in itself, he says, about preferring oneself.
A number of other contemporary philosophers have tried a similar strategy.
One is Christine Korsgaard, whose conception of ‘the normative question’ we
have been using all along. She thinks that Kant intended the formulation of the
Categorical Imperative in terms of universal law to give us, all by itself, the
moral law.68 But she thinks Kant failed to see that prescribing universal law
does not yet settle the question of the domain over which the law of the free
will must range.69 The domain could be the desires of the moment, for
example, or of the agent’s whole life. She says: ‘It is only if the law ranges
over every rational being that the resulting laws will be the moral law.’ But
then we need an argument for extending the domain this way. She thinks she
can provide one, from the nature of reflection. She starts from the observation
that what separates humans from other animals is their ability to act on the
basis of a self-conception, what she calls ‘a practical identity’, conceptions like
‘sister’ or ‘philosopher’.70 She thinks practical identities create unconditional

Morality, 243–8, and in ‘Morality, Happiness, and Peter Singer’, in Nigel Biggar (ed.), Peter
Singer and Christian Ethics (Oxford: Clarendon Press, 2014), 93–103.
66
Actually Kant in the second Critique simply starts from what he calls ‘the fact of reason’—
namely, that we are under the moral law, and he does not attempt to derive this claim from
anything more fundamental. Kant’s project in the second Critique is less ambitious in this
respect than it was in the third section of the Groundwork. He seems to have concluded that
the earlier attempt to provide a foundation for morality in freedom was viciously circular.
67
Singer has changed his mind about the objectivity of ethics, after reading Parfit, On What
Matters; but he has not yet changed his mind about this.
68
Actually all the formulations of the Categorical Imperative are needed, and Kant does not
intend the formula of universal law to be self-sufficient. See Wood, Kant’s Ethical Thought, 82,
who traces the ‘empty formalism’ objection in, for example, Hegel and MacIntyre to a failure to
see Kant’s intentions in relating the formulations of the Categorical Imperative together.
69
Korsgaard et al., Sources of Normativity, 99.
70
Frans de Waal has shown the empirical claim about non-human animals to be doubtful, by
narrating all sorts of stories about ‘precursors’ of practical identity in, for example, chimpanzees.
86 God’s Command
obligations, and this is one place where her view can be questioned.71
G. A. Cohen raises the case of the idealized Mafioso: ‘When he has to do
some hideous thing that goes against his inclinations, and he is tempted to fly,
he steels himself and we can say of him as much as of us, with the same
exaggeration or lack of it, that he steels himself on pain of risking a loss of
identity.’72 Surely we are better off not having to say that the Mafioso has an
obligation to do the horrible thing. Korsgaard qualifies her claim by adding
that the Mafioso has a deeper obligation to give up his immoral role. This is
because ‘the activity of reflection has rules of its own; and one of them, perhaps
the most essential, is the rule that we should never stop reflecting until we have
reached a satisfactory answer, one that admits of no further questioning. . . .
Following that rule would have led the Mafioso to morality.’ But here the
argument breaks down. There is nothing in the nature of reflection as such
that requires giving priority to the way humans are the same (namely, that
they have practical identities) over the ways they are different. Even if it is true
that valuing your practical identity implies valuing the human capacity to set a
practical identity, you can still value your differences from other human beings
more. It is, alas, possible to be a fully reflective Nazi. To be sure, we can always
define ‘reflection’ or ‘reason’ in such a way that it brings morality with it.73 But
there does not seem to be a morally neutral account of reason or reflection that
allows us to deduce morality from it. The history of attempts to give such an
argument is not encouraging. If, however, we accept that the moral law is
God’s command, then we can see a way to argue for it from the premiss that
reason tells us that, if there is a God, God is to be loved and obeyed. Chapter 1
argued from Scotus that we can know this proposition ‘from its terms’.

3.3.3. The Third Defence: Thomist

The third defence of eudaemonism is drawn from the picture of Aquinas’s

account of the relation between morality and happiness described by Jean
Porter. She starts with a difficulty about understanding Aquinas, which is that
he seems both to affirm eudaemonism and to assert that charity loves God for
the sake of God and the neighbour for the sake of the neighbour, and not for

See, e.g., ‘The Animal Roots of Human Morality’, New Scientist, 192 (2006), 60–1. But he does
not claim that non-human animals aspire to ‘the point of view of the universe’.
71
See my God and Morality, 176–83.
72
Korsgaard et al., Sources of Normativity, 183.
73
Thomas Nagel does this in The Possibility of Altruism (Oxford: Clarendon Press, 1970), and
again in his reply to Korsgaard in Korsgaard et al., Sources of Normativity, 203. Susan Wolf uses a
similar strategy: that Reason (she gives it a capital letter) is by definition receptivity to the values
there are in the order of their ranking, in Freedom within Reason (Oxford: Oxford University
Press, 1990), ch. 4. But, then, the question is why we should do this kind of reflection.
 Eudaemonism 87
our own sakes. Porter proposes a way that we can reconcile these apparently
contradictory claims. Aquinas expresses his eudaemonism when he states that
every human being acts for some one end, towards which all his actions are
directed; and, formally considered, this end is the same for all—namely, the
individual’s own perfection, or, which Porter regards as equivalent, her beati-
tude, or happiness. Porter says: ‘Aquinas begins the prima secundae with the
claim that everyone necessarily acts for one final end, in accordance with some
reasoned grasp of her overall good, in such a way as to direct all her actions in
some way towards this end.’74 All of this sounds like a clear denial of the
picture I attributed to Scotus and Kant at the beginning of this chapter, that
there are two fundamental motivations behind human action. But Aquinas
also says that the love of God for God’s own sake is the distinctive mark of
charity, and that charity towards the neighbour requires us to promote the
neighbour’s good for the neighbour’s sake and not our own.75 There is a
paradox here, and Aquinas–Porter proposes a way to resolve it in two steps.
After describing these steps, this chapter replies to them and then suggests and
repudiates in Section 3.3.4 a revision of the proposal.
The first part of the proposed Thomist solution comes in the following key
paragraph:
Whatever overarching good the agent loves as her final end, she must regard it as
in some way a meaningful goal for her own actions. The inclination of the will
towards any final end whatever implies a commitment to judge, choose, and act
in accordance with that end (ST I. 83. 4; I–II. 8. 2, I–II. 9. 3). The agent need not
know exactly what she ought to do out of love for whatever she identifies as her
end, but unless she is committed to do something, and has a general idea of what
that practically means, her protestations of love are empty. At this point, we recall
that every agent perfects itself in and through appropriate activities, bringing
actuality to the potentialities of its form. This implies that any choice to act is
ipso facto directed towards one’s own perfection, and the commitment to an
overarching final end, seen as a commitment to a consistent pattern of activity,
implies a commitment to seek one’s own perfection through appropriate
activities, which in some way place the agent in a right relation with whatever
she most loves.76
This point is not yet a solution to the paradox all by itself, but it provides an
account of the fundamental motivation behind the eudaemonism. What is a
Kantian going to say about this point? That the premiss of the whole argument
begs the question. What we have here are two fundamentally different accounts
of human agency. This point can be made in language from Scotus, but it could

74
Porter, ‘The Desire for Happiness and the Virtues of the Will’, 34. She is commenting on ST
I–II. 1. 5, 6, 8.
75
ST II–II. 23. 5, ad 2, and 31.1.
76
Porter, ‘The Desire for Happiness and the Virtues of the Will’, 34.
88 God’s Command
just as well be put in language from Kant. There are two basic affections or
drives of the will, the affection for justice and the affection for advantage. For
Aristotle and for Aquinas–Porter there is just one; the final end, the only one,
the ‘overarching’ one, is the individual’s own perfection or happiness, and
therefore ‘any choice to act is ipso facto directed towards one’s own perfec-
tion’. Scotus and Kant do not deny that, in Kant’s language, ‘to be happy is
necessarily the demand of every rational but finite being’.77 But this incentive
is contrasted with an independent incentive, which is also operative in every
rational but finite being. On this view, the key question, because humans are
free, is how they will rank these two incentives. Kant thinks humans are in
this way unlike other animals, though the final chapter will look at some
evidence that there are strong precursors of human morality in non-human
animals. The Aquinas–Porter account of action depends on the premiss that
action is natural activity, and therefore like all natural activities it is the
actualization of an individual’s final end. But the Scotus–Kant account is
that human action is not in this sense natural, but free. Scotus, following
Anselm, says that, if an angel had the affection for advantage but not the
affection for justice, its actions would not be free. It begs the question to assert
that we know, just on the basis of understanding human agency, that there is
just one final end.
It is tempting to say: ‘If the action is to be my action, it is intelligible only if it
is directed towards my good, and so the inclination has to be towards a self-
indexed object.’ But here we have to return to the two distinctions we drew
earlier in connection with the single-source view of Aristotle. The first dis-
tinction, from Bishop Butler, was that there are two senses in which every
good aimed at by an agent might be a good for the agent, and the first does not
imply the second. The first sense is that the good aimed at is good for the agent
just because it is aimed at. In this sense, the good aimed at might not itself (in
its definition) contain any relation to the agent beyond that of being aimed at.
In the second sense, the good for the agent is an object whose definition (even
if not articulated as such by the agent) includes internal reference to the agent.
Kant’s picture makes use of this distinction; the moral law occasions in the
agent a feeling of respect. Respect is my feeling (even, he says, a sensuous
feeling), but it is not occasioned by a self-indexed object. Rather, it is occa-
sioned by the moral law, which has no reference to me at all. The second
distinction was between cases where the explicit description under which the
object is loved is self-indexed and cases where it is not. Here the Aquinas–Porter
position and the Scotus–Kant position have common ground in denying that
the self-indexed description has to be present to the mind of the agent. Porter
distinguishes between ‘the reasons that inform a given set of dispositions or

77
KpV v. 25.
 Eudaemonism 89
commitments, and the reasons out of which an agent chooses or acts at any
particular time’. But what is at issue between the two positions is the former
question: must the object of an action be self-indexed if it is to be intelligible to
an agent? Aquinas–Porter says yes (though the object may not be articulated as
such), and Scotus and Kant say no.
Aquinas–Porter gives a second step in the proposed solution to the appar-
ent contradiction of asserting both eudaemonism and the thesis that we
should love God and the neighbour for their own sakes, and not for our
own. The solution is to point to the fact that the individual belongs in a nested
series of comprehensive general goods: the political community, the natural
world, and God’s friends.
She sees that she herself is part of the community, playing a role in its common
good, that she is one active entity in a dynamic cosmic order, and finally—given
the transforming effects of grace—that she has a place in the fellowship of God’s
friends. What is more, she naturally grasps that her own good as an individual
depends on her standing in right relation to these more comprehensive goods.
Even though the claims of the commonwealth, or the wider demands of natural
reverence or grace, may go contrary to her immediate interests, they cannot
ultimately come into conflict with her true good, her perfection and (therefore)
her happiness. The integral link between the good of the individual and more
general goods opens up a conceptual space within which the agent’s practical
reason and will can operate in a unified, coherent way.78
What should we say about this nested series, starting with the political
community, and then the natural world, and then the community of the
redeemed? The answer goes somewhat beyond Kant in the details of the
responses to all three cases, but Kant himself would reject the overall claim,
that there cannot be a conflict of true interest in any of these cases. One
difficulty here is that Kant and Aquinas do not mean the same thing by
‘happiness’. For Kant, my happiness is the satisfaction of my inclinations as
a sum. For Aquinas, it is my chief good. A direct comparison of texts,
therefore, on the question whether it is possible for me to will to sacrifice
my happiness is going to be misleading. But we can take our cue from Porter
and ask whether it is possible for me to will to sacrifice my own ultimate (not
just my immediate) interests. It is important here to bear in mind the modal
nature of the claim that is in dispute, that there cannot be a conflict. It is not
enough in any of the three cases just to say that there is no conflict. The claim
is one of necessity, a necessity that applies even to the divine will.
Both sides agree that God brings virtue and happiness together, and that, in the
words of the Psalmist, ‘justice and peace embrace’.79 But the anti-eudaemonist

78
Porter, ‘The Desire for Happiness and the Virtues of the Will’, 35.
79
Psalm 85: 10.
90 God’s Command
insists that this is not a necessary connection, but one due to God’s providential
care. That makes a difference to the kind of gratitude we have to God. It also
makes a difference to how we see our attachment to our own happiness. We see
that, since there is no necessary connection, there could be conflict, and this leads
to the thought of a ranking. Which do we put first and which second? In biblical
terms (Matthew 6: 33 (KJV)), the command is: ‘Seek ye first the kingdom of God
and his righteousness and all these things shall be added unto you.’
If we turn to the nested series of goods, is it right to say that the interest of
the individual necessarily coincides with the ‘true interests’ of the polis? In
book I of the Nicomachean Ethics Aristotle’s position is that to be completely
noble (kalon) requires ruling: ‘For while the good of an individual is a
desirable thing, what is good for a people or for cities is a nobler and more
godlike thing.’80 The knowledge of the statesman is the knowledge of this
good, under which all goods produced by action are ranked. The highest good
is a life of action by which one realizes this good, and this means it is the life of
the statesman. The best life, on this view, is the life of the leader. As Meno says,
when Socrates asks what human excellence is in Plato’s dialogue named after
him, ‘it is simply the capacity to govern human beings, if you are looking for
one quality to cover all the instances’.81 This generates a problem. What if it is
best for the city if I do not rule, but hand over the rule to someone else, and
retire from public life? The situation will then be like what Aristotle says about
friendship, and this is not a coincidence, because Aristotle thinks of fellow-
citizens as the broadest class of friends. I want good for my friends, Aristotle
says, for their sake and not for my own. But I do not want the highest good for
my friend, which is to become a god.82 For that will put my friend too far
above me, and I will no longer be able to maintain the special relation. So the
good I want for the friend is contingent on the maintenance of the special
relation, and the same is true for the good of the polis; the highest good of the
polis is only coincident with my highest good contingently upon my being one
of the leaders who rule it. It is not enough merely to be a citizen if the polis
prevents my exercising the noblest kind of activity.
Even if we leave the Aristotelian framework, the possibility of tension
between the true interests of the individual and of the state arises in the so-
called problem of dirty hands.83 The state needs in its leadership people who
are willing to compromise moral standards in a way that is inappropriate for
private individuals. This leads, it can be argued, to a corruption of individual

80
NE I. 2, 1094b9–10. There is a large literature here about whether books I and X are
consistent. I have given a summary in God and Morality, 7–72.
81 82
Meno 72d1–2. NE VIII. 7. 1159a5–12.
83
There is a characteristically nuanced and historically informed account in Michael Walzer,
‘Political Action: The Problem of Dirty Hands’, in Marshall Cohen, Thomas Nagel, and Thomas
Scanlon (eds), War and Moral Responsibility (Princeton: Princeton University Press, 1974),
62–82.
 Eudaemonism 91
moral character that is to the advantage of the state but to the disadvantage
(morally speaking) of the individual. To argue that it is advantageous to the
state for our leaders to have to some degree dirty hands may seem a cynical
view of politics. But the claim is not that morality should simply be aban-
doned, but that moral compromises have to be made, and that this is a cost to
moral character that our leaders have to bear.84 Consider the analogous case
of surgeons who have to cut into living tissue on a regular basis, and who
have to develop a certain kind of hardness of heart if they are going to do
their jobs. This, too, is a kind of sacrifice of moral sensitivity they make. Or
consider the case of soldiers who, if one believes in a just war theory, have to
be willing to become hardened to some degree about killing other people.
For a Kantian, the sacrifice of some degree of individual moral character is a
sacrifice of one’s highest interest. Kant himself might take the absolutist
position that such a sacrifice is never justified under any circumstances, just
as he sometimes says that lying is never justified. But it seems highly
plausible that there is in fact a conflict between the interests of the state
and of the individual in this sort of case. And remember again the modal
claim that is in dispute, that there cannot be a conflict. It seems highly
plausible that the state, even if there is no current conflict, could have an
interest in this kind of sacrifice by an individual. Kant’s own position on the
union of individuals into a state is that coincidence of interest is not thereby
achieved without divine assistance; there is no natural harmony, in Kant’s
sense of ‘natural’, because we are all, even after political union, under radical
evil. We can hope for this harmony only through the establishment and
spread of a society constituted in accordance with, and for the sake of, the
laws of virtue, and ‘this will require the presupposition of another idea,
namely, that of a higher moral Being through whose universal dispensation
the forces of separate individuals, insufficient in themselves, are united for a
common end’.85
The second level in the nested series is the level of the natural world, or
cosmos. Is it true that the interests of the individual necessarily coincide with
the interests of the natural world? Here again we need to go beyond Kant, but
this time because Kant did not know about the theory of evolution. Kant
thought that God has the overriding purpose in the entire natural order of
creating human moral character.86 But, even if we accept that, and suppose

84
I should probably divulge that I spent some time working on the staff of the House Foreign
Affairs Committee in Washington, though I do not mean to imply that it is only American
politicians who compromise in this way. Someone might reply that most forms of employment
have tendencies towards moral compromise of one kind or another, and so there is no real
possibility of avoiding at least the danger of this. But this is just my point. The way the world is
puts a tension between an agent’s good and her social roles.
85 86
Rel. vi. 97. KU v. 436.
92 God’s Command
that God accomplishes this end through evolution, can we say that the
interests of all the individuals who suffer and die in the course of evolution
are somehow realized in the development of human beings?87 Whole species
have had to be terminated. In any case, it seems doubtful that the whole
natural order exists for our sake. If we grant that species other than ours have
interests independent of ours, it seems that we have had an impact on the
biosphere that is to a significant extent negative. Hundreds of species are dying
out because of our activities. Can we say that the interests of the human
inhabitants of Washington, DC, for example, are necessarily coincident with
the interests of the individual birds and insects and mammals that occupy the
same territory? Or, if we think that individual members of non-human species
do not have interests, but species do, then we can ask whether the interests of
human beings necessarily coincide with the interests of all other species, or
would many species be better off if human beings had never existed?
Kant’s own position is again one that sees no natural coincidence of
interest, though he agrees that humans cannot survive or flourish without
the laws of nature in place. In the Critique of Judgement he mentions Spinoza,
a righteous man who does not (on Kant’s view) believe in God.
For while he can expect that nature will now and then cooperate contingently
with the purpose of his that he feels so obligated and impelled to achieve, he can
never expect nature to harmonize with it in a way governed by laws and
permanent rules (such as his inner maxims are and must be). Deceit, violence,
and envy will always be rife around him, even though he himself is honest,
peaceable, and benevolent. Moreover, as concerns the other righteous people he
meets: no matter how worthy of happiness they may be, nature, which pays no
attention to that, will still subject them to all the evils of deprivation, disease, and
untimely death, just like all those other animals on the earth. And they will stay
subjected to these evils always, until one vast tomb engulfs them one and all
(honest or not, that makes no difference here) and hurls them, who managed to
believe they were the final purpose of creation, back into the abyss of the
purposeless chaos of matter from which they were taken.88
Aquinas–Porter does not seem to have looked into this abyss (or, if she has,
has not been terrified by it). There is no disagreement between Aquinas–
Porter and Kant that God maintains the natural order, but Kant has a view of
‘nature’ by itself that Aquinas–Porter does not. Kant follows Scotus here too,

87
Note that this would not require ‘guided evolution’ of the kind that seems not to have
occurred—namely, where mutations occur selectively for the benefit of the species in which
they occur.
88
KU v. 452. Kant is sensitive to the different ways we use the term ‘nature’. In Rel. vi. 20–2 he
says that in one sense we are good by nature (the predisposition to good is essential to us), in one
sense we are evil by nature (the propensity to evil is innate), and in one sense we are neither by
nature (where nature is distinguished from freedom).
 Eudaemonism 93
in denying final causes to nature in itself, though preserving the purposiveness
of nature in God’s intention for it.89
The third level in the nested series is the community of the redeemed, the
‘friends of God’. Here we need to go back to the counterfactual thought
experiment mentioned in the first part of the chapter in connection with
Scotus. Again we go beyond Kant, because he has grave doubts about the
doctrine of double election.90 But suppose we accept, with Aquinas, that God
can predestine some for salvation and some for reprobation. Here we return to
the thought experiment I mentioned in connection with Scotus, the affirm-
ation by Samuel Hopkins that he would be willing to be damned for the sake of
the glory of God. Is it possible that the loss of salvation by some persons might
be to the glory of God, and that charity might therefore require them to accept
damnation? Note that this is not the same as choosing ‘to act in a damnable
way, out of the love of God’. Porter is right to deny that this is coherent.91 But
this is not what is at issue. In previous generations, Calvinist congregations
would regularly contain people who thought they were among the reprobate,
but nonetheless attended divine worship with devotion, and no doubt tried to
do their duties recognizing them as God’s commands. Is such a frame of mind
incoherent? If not, then surely there is not a necessary coincidence between my
greatest good, in this case my salvation, and the glory of God. Again, because
of the modal nature of the claim under dispute, it is not necessary to argue that
God in fact predestines some to reprobation, but that God could; that this is
within the scope of God’s sovereignty.92
The point is just that God is not constrained here by necessity any more
than we have necessity at the other two levels of the nested series. About all
three levels we can say that, if there is a harmony, it is a contingent harmony
established because God is both, in Leibniz’s terms that Kant also uses,
sovereign of ‘the kingdom of nature’ and sovereign of ‘the kingdom of grace’.93

89
Compare Aquinas in the fifth way, ‘But things lacking cognition tend toward an end only if
they are directed by something that has cognition and intelligence, in the way that an arrow is
directed by an archer’ (ST I. 2. 3). The account in the Summa Contra Gentiles does not have this
same implication.
90
e.g. Rel. vi. 143–4: ‘the salto mortale of human reason’.
91
Porter, ‘The Desire for Happiness and the Virtues of the Will’, 37.
92
Even though Kant does not endorse the doctrine of double predestination, he does have a
place in his system for God’s discretion. In KpV v. 123, he argues that reason in its practical
employment requires the postulation of the immortality of the soul. Even though the moral agent
does not see holiness of the will at the time of death, and will never in fact reach it, ‘nevertheless
in this progress which, though it has to do with a goal endlessly postponed, yet holds for God
as possession, he can have a prospect of a future of beatitude’. Kant sees this ‘holding for God as
possession’ as God’s grace. There are many obscurities here that I am not going to go into. See
my God and Morality, 160–1. My interpretation is controversial, but it is not controversial that
Kant says this is God’s grace.
93
KrV A812 = B840.
94 God’s Command
The argument about the three nested levels in Aquinas-Porter can be
summarized as follows. At the first level, our interests as individuals may or
may not coincide with the ‘true interests’ of the political units to which we
belong. So the ‘right relation’ to these units may or may not produce the
fulfilment of our own interests. The only way we can be sure that there is
harmony of this sort is if we have achieved what Kant calls an ‘ethical
commonwealth’, in which all its members live their lives not just in accordance
with, but for the sake of, the laws of virtue. But, while we do already belong to a
kingdom of ends, the ethical commonwealth is at this point still only an object
of aspiration. We can now add the key point that, even if there were not a
coincidence of interest, there could still be an obligation to obey the moral law.
Duty does not have a necessary coincidence with self-interest, but one medi-
ated by the free choice of the sovereign of the kingdom of ends. At the second
level, the ‘true interests’ of the natural world may or may not coincide with the
interests of the individual humans who live in it, and so the ‘right relation’ to it
may or may not produce the fulfilment of our own interests. Consider the
choices whether to put up a cell-phone tower or a string of colossal windmills
to produce energy. The right picture here seems to be one in which the
interests of humans are in tension with the interests of the rest of the natural
world, and the choices we should make depend on how important the interests
of humans are within this tension. There is not a necessary harmony. But
again we can add the key point that the existence of a duty towards the natural
world does not depend on its relation to our self-interest. We perhaps should
care about the survival and flourishing of an endangered species of plant, for
example, even if this does not have any impact by way of future medical
discoveries or other such advantages for human beings, and even if it requires
sacrifice of significant benefit to human beings. Finally, at the third level, even
in our relation to God through salvation there seems to be a possible tension,
unless we believe that God is required by some kind of necessity to save us all.
It is possible that God allows some individuals to reject eternal happiness. It is
possible, in other words, that it is not necessary for God to save everybody, and
that even the existence of some who are not saved promotes the glory of God.

3.3.4. The Fourth Defence: Agent-Transcendent Eudaemonism

Finally, the Aquinas–Porter argument can be revised in a way that is not liable
to the objections to the first step and the second. Suppose we say that the
agent’s encompassing good is not some more general good, like the good of the
polis or of the natural world or of God’s friends, that necessarily includes her
individual good, but the divine itself, which is by its own nature self-
transcending. For Scotus, the end for human beings is to enter into the love
that the persons of the Trinity have for each other, or to become co-lovers
 Eudaemonism 95
(condiligentes). This means that the highest activity is one of will, prepared by
intellect, since Scotus accepts the principle that nothing is willed except what is
previously cognized (nihil volitum quin praecognitum). This view should be
distinguished from a view that happiness involves both the intellect and the
will, but the activity of the will (the loving) is consequent upon the highest
activity, which is the beatific vision in the intellect. On Scotus’s view we can
say (with Paul in 1 Corinthians 13) that, of faith, hope, and love, the greatest
is love both here and in heaven. On the other view (which some attribute to
Aquinas on the basis of his commentary on this passage), love may be the
greatest of the three down here on earth, but in heaven the greatest will be a
state of the intellect.94 The proposal we are now considering is that the
agent identifies her happiness as entering into a kind of loving (God’s
loving) that is itself self-transcending. Following Barth, we can see this
divine self-transcendence within the Trinity in the relations (the perichor-
esis) between the three persons, in the creation (in which God comes in
loving relation to what is not God), and in the election and then redemption
of human beings (in which it is revealed that God is for us).
The proposal we are considering would appear to overcome the difficulties
with the two steps of the Porter–Aquinas argument. First, it does not beg the
question about motivation. It allows that we have both a self-perfecting and a
self-transcending love. But it holds that the second comes out of the first,
because we identify in perfecting ourselves with a being that is itself self-
transcending. Second, it does not hold that there is a necessary harmony
between the self-indexed interests of an agent and the wider groups in
which she is included.
The proposal has been stated in theistic terms, but can also be framed in
non-theistic terms. Consider Shelley’s couplet: ‘True love in this differs from
gold and clay | That to divide is not to take away.’95 We could expand upon
Shelley’s differentiation of love from gold. Love is the kind of thing that by
itself induces you to give it away, and, if you give it away, you end up having
more of it. Gold, by contrast, does not have this internal dynamic of self-
diffusion. It can indeed produce more gold, if you invest it; but it can also
induce hoarding and the urge to defend it from other people. It might seem,
then, that, if you identify the best (the most perfect) state of yourself as loving,
then there will no longer be a tension between perfecting yourself and spend-
ing yourself for others. We could put this point, as is sometimes done in the
moral education of children, in terms of a ‘big me’ and a ‘little me’. The big me
cares equally about others, and so there is a natural progression from caring

94
Aquinas’s view is hard to discern here, because he also says, ST I. 82. 3: ‘The love of God is
better than the cognition of God.’
95
Shelley, ‘Epipsychidion’, ll. 160–1, cited and discussed by Annas in MH 257.
96 God’s Command
about myself first (the little me), to seeing the best part of myself as not self-
preferential, to caring about others. A biblical story that can be interpreted in
this way appears in Matthew 15, which describes the encounter between Jesus
and a Canaanite woman. The bread that Jesus offers, the food of the covenant,
is not the sort of thing that has to be taken away from the children and given to
the dogs. It is not, so to speak, a zero sum. Rather, there is enough for Israel
and for Gentiles. This bread is always enough, and, in the retelling of this story
in Section 5.2.2, this feature is emphasized. Once we grasp this, we can see that
there is even a Kantian translation of this picture. If you identify your chief
interest (what you care about the most) as becoming a better person, and if
this is best put in terms of aiming at practical love, ‘the kernel of all laws’, then
you will achieve what you care about most by becoming the sort of person who
cares most not about yourself, but about others (and, as Scotus would put it,
about yourself as a quasi-other).96
There is a difficulty here, however, which discloses itself in the following
dilemma: either this is a single-source theory (deriving all motivation from my
own happiness and perfection) or it is not. Kant objected to what he took to be
the view of the Stoics that reduced happiness to virtue. At issue here is not the
term ‘happiness’, but the substantive point that we do have interests that do
not reduce to virtue, or to conforming our lives to the Categorical Imperative
for its own sake, and that it is completely appropriate for beings like us to have
these interests. The point was made earlier that Kant should have allowed
that self-indexed motivation includes more than the satisfaction of sensuous
inclination. Scotus, for example, took such motivation to include one’s own
salvation. Now a single-source theory can to some extent accommodate such
motivation. Aristotle, for instance, can insist that self-love of the right kind
(not the vulgar and greedy self-love of the many) is consistent with various
forms of self-sacrifice (giving away money, or the opportunity for action, or
even one’s life). But he also insists (in NE IX. 8) that these are forms of self-
love (philautia). His picture of motivation is that, if the agent were to ask
herself (and she does not have to do so) ‘Why am I doing this?’, the funda-
mental answer would be ‘because I am assigning myself the best thing’. Scotus
and Kant would say, correctly, that this answer is unacceptably self-regarding.
But there is a dilemma here. Someone who is impressed with the line of
argument in this fourth eudaemonist defence may disagree with Aristotle at
this point. This is what in fact happened in the development of the Aristotelian
school. The Magna Moralia repeats the doctrine of NE IX. 8 in some detail,
but drops the insistence on self-love, explicitly stating that self-love is for the
many, but that the virtuous person should be called ‘good-loving’ (philagathos)

96
See KpV v. 83, and Scotus, Ord. IV, dist. 46, on God’s justice to God as a quasi-other.
 Eudaemonism 97
instead.97 This is unobjectionable, but now we no longer have a single-source
theory. In Scotus’s terms, God will be loved both for God’s own sake and for the
sake of the union. The dilemma is that either we stick with the single-source
theory and say that this single source is self-indexed, in which case we are left
(like Aristotle) with something unacceptably self-regarding, even if we can
accommodate some forms of self-sacrifice; or we drop the ubiquitous (even if
not represented) self-indexing, and then we will have another choice. Either we
say that the self-indexed goods need to drop out, in which case we will have a
different kind of single-source theory, that all motivation derives from the
desire for the good-in-itself, and Kant rightly thinks this inhumane. Or we will
allow that self-indexed motivation properly remains, and then we will have a
double-source theory again, like that of Kant and Scotus.
The point is that the self-indexing of some goods needs to remain. We can
see this by comparing a view in which it does not remain, and asking whether
we can endorse the view. One such view is that of Maimonides, at least on one
reading of him. We will discuss his views in much greater detail in Chapter 7.
Maimonides discusses what is meant by saying that humans are in the image
(tzelem) and likeness (demut) of God, and he does this by looking at other uses
of these terms, such as Isaiah 40: 18, ‘To whom then will ye liken God? Or
what likeness [demut] will ye compare unto Him?’98 The problem is that we
seem to have to say also that we are not in the likeness of God: ‘One must
likewise deny, with reference to Him, His being similar to any existing thing.’99
The historical background to Maimonides’ discussion is Aristotle’s doctrine of
the agent intellect in De Anima 3. 5, and the medieval Muslim commentary
upon it, much of which Maimonides knew.100 There are ten degrees of celestial
intellect (nous), from the highest level nearest to God to the lowest level, which
is that with which humans make conjunction. The purified human intellect so
to speak breaks out of its bodily housing and is united with the agent intellect.
Maimonides says that it is ‘because of the divine intellect conjoined with man’
that man is said to be in the likeness of God. We do not need to say here (or in
Chapter 7) what Maimonides in fact means by this doctrine, since the purpose
is just to illustrate one extreme kind of self-transcendence. It is provided by the
interpretation of Maimonides by Marvin Fox:
Some reflection will reveal that this is the same as saying that man is similar to
God only at the point where he stops being fully man and is, instead, conjoined

97
Magna Moralia II. 13–4; MH 262. I do not need to settle whether this is by Aristotle, since
it does not matter for my argument if it is from some later peripatetic source. Interestingly, the
work says much less about God than either EE or NE.
98 99
Maimonides, Guide, i. 1. Ibid. i., 55, 128.
100
For a discussion of the similarities and differences between Averroes and Aquinas on this
question, see Stephen Ogden, ‘Receiving and Making Aristotle’s Intellect: A New Assessment of
Averroes and Aquinas’, Ph.D. dissertation, Yale University, 2015. See also Herbert Davidson,
Alfarabi, Avicenna, and Averroes, on Intellect (Oxford: Oxford University Press, 1992).
98 God’s Command
with the divine being. At that stage, we are no longer comparing man to God or
God to man, but only God to Himself. . . . When we are told that man is created in
the image and likeness of God, we are being informed of the remarkable potential
that man has for self-transcendence. He can, so to speak, leave his human form
and elevate himself to the point where he is absorbed into the divine being.101
Here we reconcile Genesis and Isaiah. But the question is whether we can
endorse this view. Surely there is something important lost in such an account
of our destiny. If we are absorbed into God, there is a sense in which we lose
ourselves. We lose, in Scotus’s term, our haecceity. One way to put this is that
the fourth defence of eudaemonism paradoxically ends up compromising the
aspiration to happiness.
So there is a dilemma for this kind of ‘agent-perfective’ account of eu-
daemonism.102 It is the best form of eudaemonism, one free from many of the
objections raised in the present chapter. But it still faces the present dilemma.
Suppose we think that an agent should be motivated by the desire to perfect
herself, and suppose that being perfected is becoming the kind of person who
is not always motivated by self-indexed goods. Do we now have a form of
eudaemonism that is not ‘unacceptably self-regarding’? The problem is that
we need to know whether this is a single-source account of motivation. If it is,
and this account does not retain motivation towards goods that are self-
indexed and necessarily so (such as the particular way of loving God that is
unique to an individual), goods that could be (counterfactually) in tension
with God’s own good, then the account is, we might, say ‘unacceptably self-
neglecting’. But if it keeps these goods, then it is no longer a single-source
account. By the definition at the beginning of the chapter, this means it is no
longer a form of eudaemonism. But what matters is the substantive theory, not
the label: we need an account that gives us both kinds of goods.

101
Marvin Fox, Interpreting Maimonides: Studies in Methodology, Metaphysics, and Moral
Philosophy (Chicago: University of Chicago Press, 1990) (henceforth IM), 171.
102
I take the term from Jennifer Herdt, and her Warfield lectures at Princeton Theological
Seminary in 2013.
 4

Can We Deduce Morality from

Human Nature?

I N T R O D U C TI O N

One important divide in theist moral theory comes in the answer to the
question whether we should accept with Aquinas that ‘every one naturally
wills happiness; and from this natural willing are caused all other willings,
since whatever people will, they will on account of the end’.1 This is the
question of eudaemonism, and it was the topic of the previous chapter,
which argued that eudaemonism (in the four forms discussed) does not have
a proper place for what Scotus calls, following Anselm, ‘the affection for
justice’. The present chapter is about a different dividing question in moral
theory: can morality be deduced from ‘natural’ facts, or from statements about
the ‘natural’ properties possessed by people and actions?2
The term ‘natural’ here is problematic. The first section of the chapter
discusses Scotus’s view that the moral law cannot be deduced from human
nature, but is exceedingly fitting to it. In the second and fourth sections the
denial of deducibility from ‘natural facts’ will extend to a dispute with con-
temporary theorists. The phrase ‘natural facts’ comes from RMH, whose
position on the topic is the subject of the third section. He distinguished
descriptivist from prescriptivist moral theories. He then identified ‘naturalism’
as a type of descriptivism, where ‘naturalism’ claims to derive the truth
conditions of moral statements from statements about the natural (that is,
descriptive) properties of human beings and their actions. His usage des-
cended from G. E. Moore’s condemnation of the ‘naturalistic fallacy’.3 Moore’s
target was any view that attempts to reduce or completely analyse a value
judgement in terms of a non-evaluative judgement. He gave as two examples

1
ST I. 60. 2. But the interpretation is controversial. See Scott MacDonald, ‘Egoistic Ration-
alism: Aquinas’s Basis for Christian Morality’, in Michael Beaty (ed.), Christian Theism and the
Problems of Philosophy (Notre Dame, IN: University of Notre Dame Press, 1990).
2
See RMH, Sorting Out Ethics (Oxford: Clarendon Press, 1997), 65–8.
3
Moore, Principia Ethica, 89–110.
100 God’s Command
of the fallacy the attempts to define right action in terms of what produces the
most pleasure for the affected parties or in terms of what is commanded by
God. But this usage does not fit well our present non-philosophical use of the
term ‘natural’, according to which social facts about consensus and supernat-
ural facts about God’s command are distinguished from natural facts. When
Philippa Foot and Rosalind Hursthouse use the term ‘naturalism’, they are
departing from Moore’s usage. They also deny that ‘natural facts’ are purely
descriptive or evaluatively neutral.4 Because of this confusion, it is better to
avoid the term ‘naturalism’ completely in this context.
The second section looks at the attempt by Robert M. Adams to deduce a
way to fix the reference of ‘good’ from facts about what most humans most of
the time think is good. We can call this ‘consensus deductivism’. The fourth
section looks at the attempt by Foot and Hursthouse to deduce conclusions
about moral goodness from facts about the characteristic human form of life.
We can call this ‘form-of-life deductivism’. No neat separation between these
two kinds of view is possible, since one of the relevant facts about human
nature is that we humans sometimes form consensus views about how best
to live.

4.1. SCOTUS

4.1.1. The Non-Deducibility of the Law from our Nature

Scotus rightly denies that the moral law can be deduced from human nature.
There are three items necessarily contained in this denial: the relation that is
denied and the two terms between which the relation is denied. First, it is
deducibility that is denied, rather than some weaker relation such as fitting-
ness. Scotus accepts that the moral law, such as the second table of the Ten
Commandments, fits human nature exceedingly well, but he insists nonethe-
less that this is natural law only in an extended sense, not in the strict sense,
because God is free to command creatures with human nature otherwise.
Natural law in the strict sense is ‘known from its terms’ or deducible from
what is known in this way. Second, it is moral law that is denied to be
deducible from human nature, and not goodness. His examples of moral law
are from the second table of the Ten Commandments, though we should not

4
Philippa Foot in NG. Rosalind Hursthouse (OVE 192) defines ethical naturalism as ‘the
enterprise of basing ethics in some way on considerations of human nature’, and adds that this
nature is not construed in terms of a relation to some supernatural entity, and that the
consideration of it is already within the framework of ethics. In the same work, p. 97, she
describes a change in Foot’s use of the term ‘naturalism’ from an acceptance of the label
‘descriptivist’ to an Aristotelianism that denies it.
 Can We Deduce Morality from Human Nature? 101
assume that he intends this list to be exhaustive. Third, the term from which
deducibility is denied is human nature. But this has to be understood in a way
that is not already conceptually subsumed under moral law. This is a delicate
point. If we define ‘human being’ as ‘a being under moral law’, or if we make
being under moral law essential to the kind ‘human being’ (as Kant does), then
there will be a trivial deduction from the premiss that a being is human to the
conclusion that it is under moral law.5 On the other hand, if we deny that
humans are by their nature under the moral law, then we have simply denied
deductivism from the beginning, and surely begged the question. Here is a
third alternative. We can grant that human beings are by nature such that they
are fulfilled, or they reach their end, by loving God. Scotus says our end is to be
condiligentes, co-lovers, which is to say that we enter into the love that is
between the three persons of the Trinity. This human love of God, however, is
distinct from the way angels love God or God loves Godself; humans love God
as rational animals. We can also hold, though this does not follow from the
above and it goes beyond the texts of Scotus, that the end for each of us is some
particular way of loving God. Scotus does say that the particular is more
perfect than the universal, and the particular is the object of will (and so of
love), whereas the universal is the object of the intellect.6 Sticking with the
universal for the moment, we can ask whether we can deduce the moral law
from our end specified as being condiligentes.
One prominent case Scotus gives is the commandment ‘You shall not steal’.
Private property, he points out, is presupposed by the commandment but is
not essential to human beings.7 He says we were not created with it, the church
gave it up at Pentecost temporarily, and we will not have it in heaven. He could
have said, but did not, that there are human communities now such as the
Franciscans who do not have it, but who do not thereby cease to be human.
Moreover, even at times that we did have it, God could and did ‘dispense’ from
it, as with the despoiling of the Egyptians in Exodus under the threat of the
plagues.8 There is no deduction of the proscription of theft just from human
nature, and the divine command is to that extent contingent; there is no
necessity here binding the divine will. It is tempting to reply (leaving aside
the question of the dispensation, and Pentecost, and Franciscan poverty) that,
when we do have the institution of private property, it must be wrong to steal.

5
Kant’s use of different senses of the term ‘nature’ (Rel. vi. 21–2) is described in Ch. 3, n. 87.
6
See Scotus on haecceity and intellect, and particular happiness, Ord. IV, suppl. dist. 49, q. 9,
art. 1.
7
See Bonaventure, De Perfectione Evangelica 2. 1: ‘Nature itself, whether as originally
constituted or as lapsed, provides this way [the counsel of poverty] in a distinctive fashion.
For the human person was made naked, and if he had remained in that state [that is, unfallen], he
would not have appropriated anything at all to himself ’ (cited in Jean Porter, Nature as Reason
(Grand Rapids, MI: Eerdmans, 2005), 354).
8
Scotus, Ord. III, suppl. dist. 37. See Exodus 12: 36.
102 God’s Command
But this begs the question. If ‘property’ is defined so that it is wrong for one
person to take another person’s property, then indeed we have a necessity in
the proscription of theft, but it is a necessity produced by the definition.
Stealing is not, however, the most difficult case for someone who wants
to maintain the contingency of the second table. I have tried to make the
same claim for the commandment ‘You shall not kill’.9 I have argued that,
without changing human nature, God could command us to kill 18-year-olds,
at which point God would bring them back to life. I was reflecting on the story
of Abraham and his son (though he was older), and the view of Scotus (and of
the Epistle to the Hebrews) that Abraham believed God was going to bring his
son back to life.10 But this move has seemed to many to be cheating.11 The
question of whether this commandment is contingent tout court is not settled
by the question of whether God could command otherwise given merely
human nature (though that was the question I was addressing). We can ask:
‘Could God command the killing of the innocent if not just human nature, but
our circumstances stayed the same; in particular, if we stayed dead once
killed?’12 Mark Murphy puts the point as follows. There is a dilemma. Let us
call facts about moral obligation ‘moral facts’ and facts about God ‘divine
facts’. We can then ask whether every moral fact is contingent (including the
fact that killing the innocent is wrong), given the maximal set of non-moral
and non-divine facts. If we say yes (the first horn), this seems to make the non-
moral, non-divine facts morally inert, as though the moral necessitating or
obliging comes out of God’s will or command just by ‘fiat’. If we say no (the
second horn), this seems to mean that, even if, as divine command theorists,
we deny that we are obliged directly by the non-moral, non-divine facts, God
is so constrained; and this seems very odd. In Section 4.1.3 we return to this
dilemma, and make the argument that both horns draw an illegitimate
inference.

4.1.2. The Fittingness of the Law to our Nature

We can start by returning to Adams’s answer to the familiar arbitrariness

objection to divine command theory discussed in Chapter 1. We can then go
back to Scotus on the way the law fits our nature. Adams’s answer is that God’s

9
See my God’s Call, 68–9.
10
Hebrews 11: 19. See Scotus, Ord. III, suppl. dist. 38, art. 2.
11
Mark Murphy raises this difficulty and the accompanying dilemma in God and Moral Law:
On the Theistic Explanation of Morality (Oxford: Oxford University Press, 2012), 121–4.
12
I am using ‘circumstances’ here in the contemporary sense, and not in the different,
technical sense in which Scotus uses it. Thus Scotus does not treat Abraham’s belief that God
would raise Isaac from the dead as one of the circumstances. We do not know what Scotus would
say about circumstances in the modern sense.
 Can We Deduce Morality from Human Nature? 103
command is not arbitrary in the contemporary pejorative sense, because it
commands what is good, but it has discretion (arbitrium) over which good
things to require. To make this picture work we need an account of goodness.
Chapter 1 suggested that a person who says a thing is good expresses that she
is drawn or attracted by it and says that it deserves to draw or attract her in
that way. But now we need an account of the criteria for this deserving.13
Adams proposes a single criterion for when ‘good’ means ‘excellent’: a thing is
good to the extent that it resembles God. But, while this fits many kinds of
excellent things, it does not fit all. In particular it does not fit the goodness of
natural kinds. Consider the greater lobelia and the spiked lobelia, two different
species of plant (though both lobelias) that differ because the first is a deeper
blue and taller and thicker in stem than the second. There is a different kind of
flourishing or good for each (the habitats are slightly different, for example).14
It seems implausible that these differences are differences in resemblance to
God. It may be that there are different ideas in God’s mind of the two species,
but then what needs to be explained is the difference in these ideas; they are
not themselves differences in likeness to God. The two species of lobelia, it is
true, both have a form of life that takes energy into a continuing subject of
change, and reproduces itself in its offspring; in this way they resemble what
God does in creation, and there is a long tradition of thinking of life as being
‘godlike’ in this way.15 But there is also goodness in the way these species
differ. Indeed, there is a different goodness in the way a particular plant within
a species differs from another. This is what follows from Scotus’s picture that
we get more perfection as we move from genus to species, and so within the
genus lobeliaceae to the spiked lobelia, and then more perfection as we move
from the species to each individual member, to the haecceity. Resemblance to
God does not seem well equipped to explain all this difference in the good. The
spiked lobelia is beautiful in a different way: fragile, pale, tapering in its spike.
The great lobelia is strong in its stem, vivid in colour, dominating its niche.
Here are four plausible ways of thinking of the relation of this difference to
God. First, God created it. This suggests not simply that God is the source, but
that God delights in the variety; perhaps this is one of the reasons for the ‘very
good’ in Genesis 1. But we are still talking here about something that is the

13
The distinction between meaning and criteria is drawn by RMH in Language of Morals,
94–110. We return to the distinction in Section 4.3.1.
14
We might try saying that they resemble God in the same way—i.e. by flourishing. But then
all goodness of life will resemble God in the same way. See Section 4.4.1. We should be looking
for a way in which they resemble God in their differences.
15
See Aristotle, Metaphysics XII. 7. 1072b3–5, and Aquinas, ST I–II. 1. 8. The exegete has to
choose how to translate ho theos in Aristotle, e.g. at 1072b25. ‘Therefore if ho theos is always in a
good state in this way, that we are sometimes, [that is] wonderful.’ By translating as ‘God’ I do
not mean to assume that it is the God of the Abrahamic faiths. The extent of the difference is a
large and interesting question.
104 God’s Command
same, the amazing variousness, even though what is various is different. This is
a subtle point, but we are imagining that God is delighting in the same
property in everything created: its difference from everything else. The second
way of thinking about the relation of the variety to God is that these various
beauties draw us to God. Chapter 1 claimed that what finally draws us and
deserves to draw us is God and what draws us to God. This makes the account
dynamic rather than static, and that is an improvement, because it connects
better with the account of the meaning of ‘good’, which is based on the idea of
being drawn. Gerard Manley Hopkins in ‘Pied Beauty’ writes:
Glory be to God for dappled things—
· · · · ·
Whatever is fickle, freckled (who knows how?)
With swift, slow; sweet, sour; adazzle, dim;
He fathers-forth whose beauty is past change:
Praise him.16
But the goodness of the variety does not reside only in its drawing us to God. It
also does not reside only in its having been created by God, which would make
the value too transparent to God and insufficiently intrinsic, like Wolter-
storff ’s suggestion about the conferred value of Nathan’s stuffed rabbit dis-
cussed in Chapter 1. The third way of thinking is metaphysically the most
ambitious. Perhaps we learn about God from the creation in the same sort of
way we learn about an artist from his or her work. It is not that the straight
lines and bright blocks of colour in late Mondrian, for example, resemble him.
But the work is, nonetheless, utterly characteristic. It manifests his aesthetic
preferences and aesthetic personality, which is in a sense present in the work.
Perhaps the two kinds of lobelia each manifest something different about the
goodness of God, even though that goodness is not divided into different parts.
This does not mean simply that God thought of each of them, though that is
true, but that each manifests God in a slightly different way. That is close to
Adams’s account, but not quite the same; manifestation and resemblance are
not the same because of this sense of God’s presence in the manifestation.
A fourth picture might be that each kind of lobelia, indeed each plant, is a part
of the biotic whole, which God loves.17 This is different than the previous
picture because it makes the value of the parts derive from the value of the
whole. We can add that God intends a whole new heaven and earth; not just

16
Hopkins, ‘Pied Beauty’, in Gerard Manley Hopkins, 30.
17
See Aldo Leopold, Round River (New York: Galaxy Books, 1953), quoted in Ann Gilliam
(ed.), Voices for the Earth (San Francisco: Sierra Club Books, 1979), 186: ‘The last word in
ignorance is the man who says of an animal or plant: “What good is it?” If the land mechanism as
a whole is good, then every part is good, whether we understand it or not.’
 Can We Deduce Morality from Human Nature? 105
new humans as co-lovers, but a new creation.18 This doctrine implies that
humans and angels are not the only part of creation to have an eternal destiny.
In this case perhaps non-human species have goodness in their destination
just as individual humans do, though there are large difficulties in this view,
such as dinosaurs. In any case, this chapter focuses on goodness in the natural
kind ‘human’ and what is accordingly good for humans, and connects this
goodness with God, not only by resemblance, but by the different ways in
which rational animals enter into the love of God by loving God’s creation.
The moral law is good for humans because it fits human nature. Scotus says
that the precepts in the second table of the Ten Commandments are ‘exceed-
ingly in harmony’ with the first practical principles that are ‘known from their
terms’, even though they do not follow from them necessarily.19 The first
practical principles give us our end—that we become co-lovers of God (con-
diligentes)—and Scotus is saying that the second table fits our end, but is not
deducible from it. He standardly describes this relation of ‘fit’ in aesthetic
terms. Thus he describes moral goodness as ‘a beauty [decor] including a
combination of due proportion to all to which it should be proportioned’.20
Perhaps at least this means that we can see an intelligible connection between
the principles of the second table and our end, so that, if a prescription for
action is proposed to us that violates the principles, we should probably reject
it. This account of goodness allows us to give some content to the way in which
the good, at least the good for humans as humans, overridably constrains what
we should take to be obligatory. We will then have a way to reply to Murphy’s
dilemma.
This suggestion is partly the same and partly different from the one Peter
Geach made in his famous article about the good.21 He distinguished ‘pre-
dicative’ adjectives like ‘red’ from ‘attributive’ adjectives like ‘good’, on the
grounds that the meaning of attributive adjectives cannot be detached from
the meaning of the nouns to which they are applied. He held that ‘good’ in the
phrases ‘good human’ and ‘good human act’ is a purely descriptive term, and
we know its meaning just by knowing the meaning of ‘human’ and ‘human
act’ and so by knowing what humans and human acts are for. I am going to
concede that we do get a constraint on what we should take to be good
for humans from knowing what humans are, and that this in turn constrains
what we should take to be commanded by God. But I want to make two points
against Geach. First, the term ‘good’ in ‘good human’ and ‘good human

18 19
2 Peter 3: 13, and Revelation 21: 1. Scotus, Ord. III, suppl. dist. 37.
20
Scotus, Ord. I, dist. 17, nn. 62–7.
21
Peter Geach, ‘Good and Evil’, Analysis, 17 (1957), 33–42. See RMH’s reply in ‘Geach: Good
and Evil’, Analysis, 17 (1957), republished in Essays on the Moral Concepts (London: Macmillan,
1972), 29–38. There is a substantive but acrimonious correspondence between them in the RMH
archive in the library of Balliol College, Oxford.
106 God’s Command
act’ has, in our ordinary language, more than a purely attributive use. For a
Kantian a good human act is one that displays a good will, and this is defined
in terms of the whole procedure of the Categorical Imperative. We need to be
pluralist about this, and concede that ‘good’ in ‘good person’ is sometimes not
used attributively, and that in ‘good human being’ it may be used both ways.
Second, even the attributive use is more than merely descriptive. To ascribe a
function to a chisel is normally to say how it is to be used, and to say that
a chisel is to be used for cutting fine grooves in wood and is not to be used as
a screwdriver is to make a prescription.22
Adams’s divine command theory needs a constraint on what we should take
to be commanded by God in order to overcome the objection from arbitrari-
ness. If God’s command makes something obligatory, should we think that we
might be commanded, and so be obliged, to fly an aeroplane into a skyscraper?
No, says Adams; for this could not be the command of a loving God. He tells
approvingly the story of the Hurka tribe, who were faced with a command by
one of their deities for child sacrifice. They decided to throw the sacred whisk
representing that deity into the river, and follow the commands of a less
demanding deity instead. In Adams’s theory, what counts as ‘loving’ is settled
by ordinary valuation. This seems wrong, for reasons given in the next section
of this chapter. But there is a related idea that seems right: We should probably
not take something as commanded by God if it does not fit the characteristic
kind of loving of God done by a rational animal. This would not by itself give a
presumption against supposing God is commanding us to take another
person’s property, since property is not required for the kind of loving of
God done by a rational animal, though there are other reasons for thinking
God prohibits theft. But this criterion would imply a presumption against
taking an innocent human life (except, perhaps, in self-defence, but that is a
topic for another occasion) in our current circumstances (in our contempor-
ary sense of ‘circumstances’, where this does not include our forthwith coming
back to this life), unless given an indication by God to the contrary. This
criterion would not rule out God commanding the taking of an innocent
human life and at the same time changing the circumstances, or even giving
some dispensation in the same circumstances. Not only is this presumption
weak, because overridable, but it is also redundant. We already have such a
presumption from God’s already revealed commands. One function of the
story of Abraham on Mount Moriah, for example, is to teach us that we are not
to think that we should demonstrate our devotion to God by sacrificing our
children. And our situation is different from Abraham’s also in that we now
have the Ten Commandments, the prophets, and (for Christians) especially
the life of Jesus.

22
Geach and the early Foot accepted the label ‘descriptivist’, but the later Foot rejected it.
 Can We Deduce Morality from Human Nature? 107
Why should we think that a presumption against taking an innocent human
life is implied by this criterion? The idea is that rational animals come to love
God by being what Scotus calls ‘pilgrims’ (viatores). There is for each one of us
a particular way of loving God that is our end. We do not yet see this, except in
glimpses, but we are being prepared for it.23 What God creates us to be fits us
for our doing particular good things. The idea is that the particular destination
of each one’s pilgrimage is internally related to, or is partly given its form by,
the journey towards it. For example, there are particular dangers to be faced
and particular temptations to be resisted and particular lessons to be learned
in particular relationships. The story of Jesus’s dealings with the woman with
the issue of blood and Jairus, the ruler of the synagogue, is instructive here.24
Jesus asks each to go through a time of trial that is opposite in character. He
requires the woman to become public with her condition, which she has been
hiding as she visited doctors fruitlessly for twelve years. He calls her ‘daughter’
(the only time he uses this form of address in any of the Gospels). On the other
hand, he requires Jairus to wait, while he heals the woman and listens to her
whole story, when Jairus’s 12-year-old daughter is dying. Then he narrows the
group to just three disciples, and he leaves the mourners behind, outside the
house, as though Jairus is step-by-step leaving his public role. We are sup-
posed to see the contrast between these two stories, which are deliberately
interlinked by the details of their telling. There are two opposite temptations
through which Jesus is respectively leading each of these two people. Our
particular end will reflect the character of this formation, and this means that
we can get a partial sense of our particular destination by looking at our
particular trajectories. It is inconsistent with this feature of our rational-
animal agency to command that innocent life be terminated before the
pilgrimage is complete.
This point comes helpfully into connection with the four Barthian con-
straints mentioned at the end of Chapter 2. We can learn from the premiss
that God gives us commands that we are responders of a certain kind; for
example, we are individual centres of agency, we are in time, we are free, and
we are language-users. In the present discussion, all four of these constraints
are in play, but particularly the first three. We have to be in time, in order to be
responders of the right kind to a command. This is because a command is to a
centre of agency, whose obedience consists in changing how things are or in
resisting change. Also, we have to persist, in order to be obedient, through the
hearing of the command to the obeying of it. But we can say more than this. In
order to be centres of free agency, we have to have a history. We cannot choose
moment by moment, without comparing the situation in which we choose

23
See Ephesians 2: 10 (KJV): ‘For we are God’s workmanship, created in Christ Jesus unto
good works, which God hath before ordained that we should walk in them.’
24
Luke 8: 43–56, and see Mark 5: 22–43.
108 God’s Command
with situations we have known before. Sometimes this point is put in terms of
a narrative that is required for the intelligibility of individual choices.25
We do not know in advance what the good works are that God has prepared
for us, and God is free with respect to the route selected for each of us, and the
duration of that route. There is, according to the preceding account, a con-
straint on God’s arbitrium, God’s discretion: God chooses in accordance with
our good as pilgrims. But the question of what access we have to the nature of
this good is a different one, though also an unavoidable one.26 To complete the
epistemological part of the theory would require arguing for an assumption,
because we do not know what route God has chosen for each of us to our final
end, that the pilgrimage is not complete before a full lifespan, and therefore an
assumption that killing would be a premature termination, unless God were to
indicate to the contrary. This presumption applies to all taking of innocent
human life. Scotus holds that the particular substance is a perfection of the
species, just as the species is a perfection of the genus, and one would not,
absent such indication, be justified in violating the universal norms that fit the
species in the name and for the sake of the particular perfection.
This gives us one example of a constraint from our nature on what we count
as good, if we are to be able to respond to the divine call to be co-lovers. There
is a form of argument here that can be extended to other examples. The form
of argument is transcendental in the Kantian sense; it argues from the condi-
tions of possibility of some fact that is taken as basic. Thus the argument from
providence is a transcendental argument from the ‘fact of reason’, that we
humans (creatures of sense and creatures of need) are under the moral law.
The present argument argues backwards from the fact that we are recipients of
God’s call to conclusions about what we (and God) have to be like to make this
possible. But then it reverses direction, and asks what constraints are placed on
what we can take to be divine commands by the fact that we are this kind of
people. Now consider the proscription on bearing false witness. There is a
plausible argument, this time from the last of the four Barthian constraints,
that we are language-users. Our language is a system of external signs, used to
communicate internal thoughts.27 But then we have to be able to assume that
these signs are being used most of the time to communicate thoughts correctly.
What is at issue here is not whether the thoughts communicated are true, but

25
Charles Taylor, Sources of the Self (Cambridge, MA: Harvard University Press, 1989), 47:
‘But this is to state another basic condition of making sense of ourselves, that we grasp our lives
in a narrative.’ See David Velleman, ‘Narrative Explanation’, Philosophical Review, 112/1 (2003),
1–25.
26
The idea that our nature puts a constraint on God’s discretion is well defended by Evans,
God and Moral Obligation, 62–8.
27
I say our language because each of the angels, according to Pseudo-Dionysius, knows what
the others are thinking immediately. Such angels do not need external signs for this sharing of
thought. It does not follow, however, that they are not thinking in language.
 Can We Deduce Morality from Human Nature? 109
whether what is communicated corresponds to what is thought true. Consider
a society in which the people communicated only one-third of the time their
actual beliefs and intentions, and two-thirds of the time what they did not
believe or intend, but suppose there was no way for the recipients of the
communication to tell which was which. There would not be much point in
trying to communicate in such a society. This argument resembles Kant’s
argument in the Groundwork that we have a perfect duty not to make false
promises.28 But it repeats only the most plausible part of Kant’s argument, not
deriving an absolute obligation not to lie. The plausible part of the argument is
that the institution of language-using requires that we be able to trust that
most of our fellow language-users most of the time are communicating what
they believe to be the truth, or what they really want. So there is a presumption
against telling a lie, because if anyone, anytime, with any degree of frequency,
may lie, this undercuts the very institution of language, which is necessary for
lying as well as for telling the truth.
In the case of lying, as in the case of killing the innocent, we are left with an
overridable constraint. Consider by contrast the exceptionless kind of con-
straint offered by the ‘new’ natural law theory, which proposes eight basic
goods, as Jean Porter puts it, ‘elemental enough to be regarded plausibly as
self-evident to all and yet provided with enough content to provide an
immediate basis for practical reflection’.29 The dilemma that Porter poses
for this kind of theory is that either the list is sufficiently general to be self-
evident, but then it does not have enough content significantly to guide action
in the exceptionless ways the theory proposes, or it is specific enough to guide
action in this way, but then it is not self-evident to all. To pose this dilemma,
however, is not to deny that goods such as life and truth pose some constraint
on moral obligation.

4.1.3. The Social Character of Obligation

Have we, by bringing in human nature in this way, abandoned the distinctive
mark of divine command theory, and simply turned it into a species of natural

28
Gl. iv. 422. But note that deceptions and lies are not the same thing, if we include
withholding truth under deception.
29
Porter, Nature as Reason, 128. The ‘new’ natural law theory is that of John Finnis, Germain
Grisez and Joseph Boyle. The eight basic goods are life (including health and procreation),
knowledge and aesthetic appreciation, skilled performance of all kinds, self-integration, authen-
ticity or practical reasonableness, justice and friendship, marriage, and religion or holiness.
According to this theory there are principles of practical reasonableness that are not self-
evident to all, but that mediate between the basic goods and action. But it is doubtful that any
plausible principles can in fact get us to exceptionless norms that tell us, for example, that lying is
never permitted.
110 God’s Command
law theory? These terms are not completely determinate. The argument has
not been (to put this in Scotist language) that the moral law is natural law
strictly speaking. Rather, the content of at least two of the Ten Command-
ments has been turned into presumptions against taking God to be command-
ing us to act in a certain way, and these presumptions are taken from what fits
human nature. This manœuvre has also been denied in the case of another of
the Ten. For Scotus, the first table is natural law strictly speaking (with the
exception of the ‘seventh day’ prescription, where Scotus is not sure about the
periodicity). The command to love the neighbour would also be natural law
strictly speaking, since we are necessarily commanded to love God, and to love
the love of God, and therefore to love the neighbour’s love of God.30 But
Scotus also believes in the possibility of reprobation, so that there is a
restriction needed: we are commanded to love the love of God in the neigh-
bour ‘at least by anyone whose friendship [God] is pleased to have’.31 None-
theless we can and should have a defeasible presumption that we and the
neighbour are not among the reprobate, because the judgement is God’s and
not ours. Moreover, we can say a bit more at the more general level. Since we
are necessarily commanded to love God, and since human nature is specified
in terms of this end, we can say that God necessarily commands what fits
human nature. But Scotus does not think that any of the specific commands in
the second table can be deduced from this. There are two different possible
kinds of deduction from what fits us. There is a deduction of a presumption in
two cases, but in no case is there a deduction of an absolute prohibition.
A second point is more important. The argument given so far does not
imply that the moral law or moral obligation is deducible from human nature
even in the case of the prohibitions on killing the innocent or on lying. This is
because what makes something obligatory is that God commands it. There can
be a presumption against doing something and still not an obligation not to do
it. Here we return to Adams and Darwall and their notion of the social
character of obligation, which we can accept with one qualification.32 The
social character is that we are obligated to someone, or by someone.33 The
opposite of ‘obligatory’ is ‘forbidden’. It is not at all an easy matter to delineate
this social character, but the general point seems right. The qualification is that

30
Scotus, Ord. III, suppl. dist. 37. The argument was spelled out in Section 1.3. There is the
same argument in Kierkegaard, Works of Love, 107.
31
Scotus, Ord. III, suppl. dist. 28, art. 1.
32
It is also the view of Aquinas, at ST I–II. 90. 3, that law, including natural law, has to be laid
down by a lawmaker, since law is an ordinance of reason made by one who has care of a
community.
33
Murphy objects that, whereas tort law always has a tortfeasor and a victim, and so has a
‘bipolar’ structure, this is not true of criminal law, which can have a ‘monadic’ structure in which
there may be no victims at all (God and Moral Law, 126). If this is right, we should not say that
obligation as such is bipolar. But there is good reason, explored by Darwall in The Second-Person
Standpoint, ch. 5, to think that moral obligation is more like tort law in this respect.
 Can We Deduce Morality from Human Nature? 111
we should not derive the agent’s obligation from the goodness to the agent of
the relation that would be damaged by violating the obligation.34 This would
just be another form of eudaemonism. But, leaving that aside, suppose we start
with the way Adams puts the basic idea, that, where there is a violation of an
obligation, one ‘may appropriately have an adverse reaction to it’.35 The
question is: who is it whose appropriate reaction is here in question? Adams
argues that the best candidate is God, since other humans are in various ways
disqualified.36 Human beings have limited information, and limited sympa-
thies. We do not, for example, see into each other’s hearts, and so we often
have to guess at each other’s preferences. Moreover, even if we did know those
preferences, we would tend to prefer the preferences of some people to the
preferences of others in a way not countenanced by the moral law.
We might ask: ‘Why should we assume that the person to whom we are
accountable in an obligation is the same as the person who generated the
obligation in the first place?’ There is a tradition of argument, in Kant, for
example, and also in Suarez, that God is legislator, executive ruler, and judge
(by analogy with the three functions of human sovereignty), and that moral
law assumes that it is the same person who carries out all three functions.37
This tradition lay behind the discussion of God’s authority in Chapter 2. In
Kant’s terms, the author of the law (which we repeat in our own wills) has to
have a holy will, the administration of the law has to be by the ‘supersensible
author of nature’ (this is the argument in Kant’s second Critique for the moral
postulate), and the judge has to be able to see into our hearts; and there is one
person, and it is the same person, who does these three things.38 We might ask:
‘Why could it not be three different persons?’ After all, in human societies it
can be an advantage to have these functions divided.39 Here we can afford to

34
I am grateful to Jennifer Herdt for this point. Scotus distinguishes three forms of love of
God, at Ord. III, suppl. dist. 27. There is loving God for God’s own sake (or wanting God to have
everything good, which might not include union with me), and loving God for my relation with
God, and loving God for some satisfaction in the relation. Someone who derives obligation from
the good of the relation with God will love God in the second or third way, but not the first.
35
Adams, Finite and Infinite Goods, 233.
36
Thomas Carson has an argument to similar effect, for what he calls ‘the divine preference
theory of rationality’ (Value and the Good Life (Notre Dame, IN: Notre Dame Press, 2000),
239–48). Note that it does not follow from there being an appropriate reactor that there is a best
reactor. We humans might be appropriate reactors to violations by other humans. But Adams
and Carson are making the point that we are disqualified in various ways from doing this well.
37
One place to see this is Kant’s M. viii. 257–8. He argues that the moral concept of God
requires God’s holiness as lawgiver first, then God’s benevolence as ruler, and then God’s justice
as judge. This order is, he thinks, required, so that the lawgiver does not accommodate himself to
benevolence (because he knows he cares first for holiness), but also the judge can presuppose
(because they are his own) the benevolent intentions of the laws to which he holds us
accountable.
38
See my God and Morality, ch. 3.
39
Murphy, in God and Moral Law, 130–1, argues that Hart (The Concept of Law (Oxford:
Oxford University Press, 1994), 26–49, 51–61) has refuted the view of Austin and Hobbes that all
112 God’s Command
be modest in our claims for necessity. If there is only one God (and perhaps
there can be only one, but that needs to be shown), that one God is the most
appropriate person for these three roles.
If this argument works, or something like it works, we can say that moral
obligation requires not just a presumption against doing something but an
obligator, and that deducibility from human nature and non-divine facts alone
therefore has to be denied. Now we can return at last to Murphy’s dilemma.
The first horn proposed that, if God is free to command what God wants, the
non-moral and non-divine facts are inert. But if, as I have argued, God is
constrained though not determined by facts about our nature, these facts will
not be inert. The second horn of the dilemma proposed that it is odd to say
that we are not obligated by the maximal set of non-moral, non-divine facts
(where these include facts about our nature), but God is so constrained.40 The
response is that this is not odd at all. We and God are different. Both God and
we are constrained by the non-moral and non-divine facts, and neither God
nor we are obligated by those facts. But we are obligated by God’s commands.
God does not require an obligator at all, but is the obligator. Even in those
cases of moral law (if there are any) in which God’s command is constrained
by the non-moral, non-divine facts, we are obligated not by those facts but by
God’s command.

4 . 2. CONS EN SUS D ED UCTIVI S M

Now what about Adams’s claim that we can fix the reference of ‘good’ by the
evaluations of most of the people most of the time? His analogy is to ‘water’,
where we start from what most people refer to, what is in rivers, and oceans,
and falls from the sky in rain. What constitutes water is something else, being

legal authority originates in a single, legally unlimited sovereign. But it is not clear why Hart’s
account of human law should bind what Murphy calls an ‘explanans-driven explanation’ like his
own that starts with God’s role in explanation, as it were from the top down, rather than arguing
to the need for God in an explanation of something else. Sometimes Adams seems to be arguing
from obligation to the divine obligator. But his view is better seen as assuming God as the loving
model of all good, and then working out what our relation to God as obligating some good must
be like.
40
Note that tying the commandment to our end, and our end to a unique way of loving God,
means that the fact about our good that was, on Murphy’s view, a non-divine fact is now a divine
fact (about God’s relation to us). So non-deducibility from non-divine facts is preserved even
without the point about the necessity of an obligator. Murphy does not himself accept the
argument from the social character of obligation, or Aquinas’s position on law, and he will not be
convinced by my reply. Strictly, his dilemma is about moral necessitation rather than moral
obligation. But God’s command necessitates and the reply here applies to that necessitation. This
whole section of the chapter has been much helped by private communication from Murphy.
 Can We Deduce Morality from Human Nature? 113
H2O, and on an imagined twin earth (which has rivers, lakes, and oceans, and
rain like ours, but made of XYZ) they do not have water at all. The structure of
Adams’s account is that the meaning of ‘good’ does not give us the nature of
‘good’; what is given by the meaning is, instead, a role that the nature is to
play.41 He limits this claim to the meaning of ‘good’ in certain contexts, those
in which ‘goodness’ is naturally interpreted as meaning excellence. He admits
that, if we did not make this limitation, and tried to accommodate all uses of
‘good’, we would end up with some kind of expressivist theory.42
Adams says that the role assigned by ordinary understanding to the good is
that it is an object of pursuit. He quotes Plato, who says in the Republic that the
Good is ‘what every soul pursues, doing everything for the sake of it’, and who
in the Symposium defines ‘the Good or Excellent or Beautiful’ as the object of
love (in Greek, eros).43 Adams agrees with this, though he thinks eros here
should be taken primarily as appreciative contemplation rather than pursuit.
This enables him to say without embarrassment that God has this kind of
love.44 He then considers an objection, that ‘in our folly, we often love, with
admiration and desire, things that are not good, and fail to love things that
are’.45 But he denies that this is a problem for his theory, because ‘it is not one
of those theories that analyze the nature of the good as consisting in some fact
about our desires. The role of our desires, on my view, is only to help fix the
signification of our value terminology to a property or object that has its own
nature independent of our desires.’ My question is whether the failure of our
actual desires is a problem for Adams’s theory also, even though Adams
proposes that those desires only fix the reference of the good, rather than
determining its nature. He is forced by his account to say that ‘we cannot
always or even usually be totally mistaken about goodness’.46 This is the claim
we need to consider. It is a salient claim for a divine command theory, if we are
constrained in what we take to be divine command by our conception of the
good. Surely it is true that we can be, and very often are, deeply wrong about
the good? This is not quite the same as saying we are ‘usually totally mistaken’,
but we are mistaken enough that we should be hesitant about our ability to fix
the reference of ‘good’. We will look at some texts in Aristotle, and compare
them with claims Jesus makes in the New Testament. Jesus overturns or
‘transvalues’ (in Nietzsche’s term) our conception of the good, and Aristotle
gives a more accurate picture of what we usually think of as ‘common sense’. We
can then return to divine command theory and ask how we can, nonetheless,

41
Adams, Finite and Infinite Goods, 16–20.
42
When the Mets fan says ‘Good!’ when she hears that the Mets have beaten the Dodgers, she
is expressing a favourable attitude towards the event; ‘but she probably supposes that Dodgers
fans who say “That’s too bad” at the same news are not disagreeing with her about a matter of
objective fact, but simply expressing contrary feelings’ (ibid. 17).
43
Plato, Republic 505d–e, and Symposium 210d–212c.
44 45 46
Adams, Finite and Infinite Goods, 132–6. Ibid. 19. Ibid. 20.
114 God’s Command
be constrained in what we take to be a divine command by a conception
of the good.
The chief contrast between Aristotle and Jesus has to do with what we might
call ‘competitive goods’. A competitive good is one where, in order for one
person to have it, another person has not to have it, or to have less of it. Wealth
and power are two relatively clear examples, and honour is another, in a
particular sense that needs to be articulated. For one person to be rich requires
that another person or persons be not-rich. This is not true if wealth is defined
as the ability to produce goods, but it is true of wealth in the ordinary sense.
Power is the same in this respect. For one person to be powerful requires that
another person or persons be not-powerful. Power is power over. ‘When
everyone is somebody, then no one’s anybody,’ as the Grand Inquisitor sings
in The Gondoliers. The term ‘honour’ has many senses. As Aristotle uses it, it
means centrally being thought well of by other people. Aristotle makes the
point that it cannot therefore be the chief good, since this requires being
honoured by the best people, and this shows that the chief good is actually
activity in accordance with virtue.47 But again, being honoured in this sense is
a competitive good. One person can have it only if others do not. One
melancholy aspect of the academic life, for example, is that it seems to have
space only for a limited number of thinkers and views to be fashionable at any
one time. This means that a person who was famous in his forties often has to
suffer the indignity and distress of becoming less famous, and being replaced
in the spotlight by newer stars.48
When we look at Aristotle’s account of the chief good for human beings, we
find that at least most of the time he seems to think that it requires wealth,
power, and honour.49 In the Eudemian Ethics, for example, when introducing
the virtue of ‘generosity’ (though in this context this translation is misleading),
he says that it is possible that neither the virtue nor the vices between which
the virtue is a mean apply to some person, because he is behaving appropri-
ately, but on too small a scale.50 But even the large scale of the expense is not
sufficient; it has to fit a person’s life-status. The virtue of magnificence is where
this fit obtains. The embassy of Themistocles to Olympia was elaborate but not
magnificent, because it did not fit his former low station. It would have fitted
Cimon, and thus have been magnificent for him. The case of honour is similar.

47
NE I. 5. 1095b28–31.
48
Perhaps in heaven honour is not like this, a limited resource, but we honour each other
with full mutuality. I am grateful to Terry Irwin for this point.
49
See John E. Hare, ‘Eleutheriotes in Aristotle’s Ethics’, Ancient Philosophy, 8 (1988), 19–32.
But Aristotle puts money in a different category from power and honour. Power and honour are
good in themselves, though not sufficient for happiness; but we choose wealth ‘for the sake of
something else’ (NE I. 7. 1097a25–6).
50
EE III. 4. 1231b28. I put the name of the virtue in quotation marks because Aristotle is
talking about the virtue characteristic of an eleutheros, or free man, as opposed to a slave.
 Can We Deduce Morality from Human Nature? 115
To have the relevant virtue (magnanimity) you have to be worthy of great
honour and deem yourself worthy of it. Fit is not enough. The man who is
worthy only of small honour and deems himself worthy of such is not
excellent, because there is no greatness.51 In terms of power, Aristotle tells
us in the first book of the Nicomachean Ethics that the best life is the life of
action, and is the province of the ruler’s science, ‘for while the good of an
individual is a desirable thing, what is good for a people or for cities is a nobler
and more godlike thing’.52 It is a large question how to reconcile this with the
last book of the same work. However this is adjudicated, these sorts of passages
teach that the chief good for a human being requires power over others. The
claim is not simply that the competitive goods are good, but that they are
necessary for a highly admirable life.53
The Gospels portray Jesus as overturning this sort of view, and the rest of
the New Testament follows suit. To cite one example, Aristotle says that
‘humility’ (tapeinotes) is the state in which persons are so low they should
not even aspire to virtue; in the New Testament we are told in humility
(tapeinophrosune) to consider others better than ourselves.54 For a second
example, Jesus commends the widow who put two small coins, all she had in
the world, into the temple treasury (Mark 12: 42). Aristotle would deem such
an act wasteful; she simply could not afford it. For a third example, Jesus
describes the prevailing opinion that you should hate your enemies, and says:
‘But I say to you, love your enemies’ (Matthew 5: 43–4). This overturning of
the world’s values is a central Christian theme, and is abundantly discussed in
the literature.55 There is an important and difficult question whether the
difference of Christian virtue, as described in these texts, shows more continuity
or discontinuity with pagan (or ‘natural’) virtue. Can we know by human
reason, unaided by special revelation, what is the best human life? Joseph Pieper,
for example, writing out of a Thomist tradition, says both that Christian love is
‘something new and fundamentally different’ and that it is a perfection or
completion of natural love.56 He also speculates that, when we talk of Christian
love as a consuming fire, it is because there are earthly loves that have to be
consumed. There is a rich vein of enquiry here, but it is not necessary for our
present purposes to explore it. It is enough to agree with the observation, ‘and
this is the verdict: Light has come into the world, but men loved darkness instead

51
EE III. 5. 1233a18. 52
NE I. 2. 1094b9–10.
53
See my God and Morality, 45–51.
54
NE IV. 3. 1124b20–3; Philippians 2: 3.
55
See, e.g., Stanley Hauerwas and Charles Pinches, Christians among the Virtues: Theological
Conversations with Ancient and Modern Ethics (Notre Dame, IN: University of Notre Dame
Press, 1997).
56
Joseph Pieper, The Four Cardinal Virtues: Prudence, Justice, Fortitude, Temperance (Notre
Dame, IN: University of Notre Dame Press, 1966), 35–40. See also Gilbert Meilaender, The
Theory and Practice of Virtue (Notre Dame, IN: University of Notre Dame Press, 1984), 27–36.
116 God’s Command
of light because their deeds were evil’ (John 3: 19). In Chapter 1, the second
argument from Kant about the dependence of our morality upon God was from
the need for grace, because we are all born under the evil maxim that prefers our
happiness to our duty. But in the present context all we need is that there is not
enough truth in most people’s desires most of the time for those loves to fix the
reference of the term ‘good’ and its related family.57
This claim is not shown to be true just because Jesus disagreed with
Aristotle. It is possible, after all, that general revelation is progressive. If
consensus deductivism implies a consensus between all people at all times,
these deep disagreements are relevant; but the view might be that there is
enough consensus now, or in the contemporary period (whatever that is), or
among us (whoever that is), to fix the reference. But then it is relevant to
consider the return of ‘post-Christian’ ethical thinkers such as Nietzsche and
Bernard Williams to the Greeks. It is true that the culture of large parts of the
world has been shaped by the Abrahamic faiths, but it is also true that
pluralism has emerged as an increasingly dominant feature of our way of
life. The central point is not about whether most people get most of their
evaluations and preferences right most of the time. Even if they do (which is
doubtful), this is not the right way to fix the reference of ‘good’. As RMH
pointed out years ago, this approach to fixing the reference by consensus is
inherently relativistic.58 We end up saying that the reference of ‘good’ is fixed
by whatever most people say it is. We ought to have a way of being able to say
that most people most of the time are wrong, even if it is not the case that they
are. If we take the consensus model, we lose such a way.
Adams is aware that there is a problem here. He wants to maintain what he
calls the ‘critical stance’. Thus he says that the truth behind Moore’s Open
Question argument
amounts to at least this. For any natural empirically identifiable property or type of
action that we or others may regard as good or bad, right or wrong, we are committed
to leave it always open in principle to raise evaluative or normative questions by
asking whether that property of action-type is really good or right, or to issue an
evaluative or normative challenge by denying that it is really good or right.59
The problem is that this openness extends only to limited questions within
what he takes to be the overall massively reliable field. Adams is not willing to
concede that the framework as a whole might have been largely distorted.

57
See OVE 263–4. If we ask ‘why so many human beings are leading, and have led, such
dreadful lives, we see that occasionally this is sheer bad luck, but characteristically, it is because
either they, and/or their fellow and adjacent human beings, are defective in their possession and
exercise of the virtues on the standard list’.
58
RMH, Sorting Out Ethics, 72–3. He already had a version of the ‘twin-earth’ thought
experiment.
59
Adams, Finite and Infinite Goods, 78.
 Can We Deduce Morality from Human Nature? 117
So how can we be constrained in what we take to be a divine command by
our conception of the good, if not by consensus of actual belief and desire?
Presumably, if we are going to take something as a divine command, we are
going to be within the framework of a belief in a God who gives commands.
Such a framework is a religion, though we do not have any set of necessary and
sufficient conditions for what counts as a ‘religion’. Those within each reli-
gious tradition in which there might be a divine command have to use the
resources of that tradition about what is good. Many of these religions will
have a distinction like the one within Christianity between general revelation
and special revelation. We will discuss a distinction of this kind within Islam
in Chapter 6 and within Judaism in Chapter 7. There will be, accordingly,
some things known outside and some things known only inside the revelation
contained in the sacred texts. A divine command theorist can say that what we
take to be a new divine command has to be screened through both the general
and the special revelation about the good that has already been given. This is
the topic of Section 5.3.3, where other conditions will be mentioned. The
remarks about Aristotle should give us pause regarding how much access
we have merely through general revelation to what is the best life for a
human being.

4.3. PRESCRIPTIVISM

The third and fourth sections of this chapter are about a debate between
RMH’s views about the objectivity of moral judgement and the contrasting
attempt by Philippa Foot in Natural Goodness and Rosalind Hursthouse in On
Virtue Ethics to deduce conclusions about moral goodness by what Foot called
a ‘natural-history story’ from the characteristic form of life of the human
species.60 I recognize the danger that I will seem merely to be summoning up
the ghosts of old battles. Foot and RMH were sparring partners from the
beginning, and, when RMH attacked descriptivism, it was often Foot that he
had first in mind. But the battle I am about to join is not simply the old battle,
though there will be some continuities. For Foot changed her mind repeatedly.
Foot scholars divide up her career, like Plato’s, into three periods: an early
Foot, a middle Foot, and a late Foot.61 Natural Goodness was late Foot.
Hursthouse has, moreover, added significant structure to Foot’s account.
There are some ways in which late Foot was closer to RMH than early or

60
NG, e.g. 51.
61
For example, Gavin Lawrence, in ‘The Rationality of Morality’, in Rosalind Hursthouse,
Gavin Lawrence, and Warren Quinn (eds), Virtues and Reasons (Oxford; Clarendon Press,
1995), 89–91.
118 God’s Command
middle Foot, and Hursthouse acknowledges this.62 Nonetheless, there were
still important differences, and one of them is that Foot affirmed and RMH
denied the deducibility of conclusions about moral goodness from facts about
human nature. I argue that we should accept some of the positions of each side
in this dispute, but that form-of-life deductivism should be rejected.
One theme in this discussion of Foot will be that we need to disentangle her
deductivism from her attack on what she calls ‘subjectivism’. I think she was
mistaken in both cases, but I will try to show that it is possible to be opposed to
subjectivism of certain kinds that she was also opposed to, and also be opposed
to deductivism. So what is subjectivism? People who knew RMH will know
that it used to make him furious to be classified as a subjectivist. He also
steadfastly refused the label ‘non-cognitivist’, though Foot repeatedly used this
label for him in Natural Goodness.63 She acknowledged that RMH refused the
names ‘non-cognitivist’ and ‘subjectivist’, but she thought he was still guilty of
the errors that she referred to under those labels. So what were those errors?
The central error she was concerned with was the error of thinking that value
is desire-based, rather than being (‘objectively’) there whether it is desired or
not. But there are at least three things this might mean, and they can be
distinguished under three headings: ‘motivation’, ‘moral properties’, and
‘ideals’. RMH’s views can be helpfully separated under these headings.

4.3.1. Motivation

RMH held that when we make a moral or evaluative judgement we are

expressing a pro-attitude towards, or an endorsement of, some prescription.64
The position Foot was attacking was what we might call ‘judgement internal-
ism’, the view that motivation is internal to moral and evaluative judgement
(but again we have to take ‘motivation’ very broadly, so that it is not confined

62
Rosalind Hursthouse has told me that Foot told her that she was going to write ‘something
nice about Dick’, but died before she was able to do so. In her own work, Hursthouse talks
repeatedly about ‘Hare and Foot’, or more frequently ‘Foot and Hare’, as a conjunct particular (e.g.
OVE 180–7).
63
For example, RMH said (Objective Prescriptions (Oxford: Clarendon Press, 1999), 4): ‘An
“ethical cognitivist” ought to be someone who thinks that one can know that moral statements
are true. So he must at least think that some moral statements are true, and also therefore,
presumably, that some are false. Since I myself think both that moral statements can be true or
false, and that we can know them to be true or false, I get extremely cross when people classify me
as a non-cognitivist . . . We need to say what it means to call moral statements true or false, and
what it means to say that we can know them to be true.’
64
Foot sometimes called what was supposedly expressed here, ‘something “conative” ’ (NG 8).
But this is too narrow, since a ‘conation’ is an attempt or endeavour, and moral and evaluative
judgements are often not connected with action in the way the term ‘conative’ suggests. There is
no word in English that fits exactly the family of emotional, desiderative, and volitional attitudes
involved here.
 Can We Deduce Morality from Human Nature? 119
to motivation to action).65 Why did RMH care about this? He thought it was a
true analysis of the logic or grammar of evaluative language.66 But we can say
something else. RMH was a prisoner of war of the Japanese, and spent the
Second World War on the Burma–Siam railroad. When he came back to
England, he and his generation found a world that had changed in funda-
mental ways. He was, throughout his life, concerned for the possibility of
communication about moral matters between different cultures and different
generations within the same culture. He thought that his account of the
difference between the meaning of evaluative terms such as ‘good’ and
‘wrong’ and the criteria for the use of such terms in evaluative and moral
judgement was important for the preservation of this possibility. He thought
we were more likely to be capable of genuine dialogue over moral issues if we
shared the meaning of these basic terms, and could then talk together about
what criteria to employ for their use.
What did he think was the difference between meaning and criteria? He
thought that it was given in the meaning of evaluative terms that, when we use
them sincerely in an evaluative judgement, we commit ourselves to an im-
perative. If the judgement is a moral judgement about action, the imperative is
a command to act a certain way.67 For RMH, the criteria for an evaluative
judgement were the descriptive facts about the world that we use in our
evaluations. The strawberry is good, we say. Why? Because the strawberry is
sweet, red, ripe, and fragrant.68 If I use the sharpness of a knife as a criterion
for commending the knife, as most people do, I am endorsing this standard of
the community to which I belong. This endorsement RMH called ‘a decision
of principle’. The principle here is that knives are good when they are sharp,
and my decision is to endorse this principle in commending the knife.
Here is one place the early Foot and RMH disagreed. She held that we
cannot simply decide what criteria to apply; some are internal to the moral
point of view. She thought it was simply nonsense to consider it a moral
judgement that it is wrong to look at hedgehogs in the light of the moon, or

65
In David O. Brink’s elaborate categorization of types of internalism, this is an internalism
about motives, but it is an appraiser internalism. See David Brink, Moral Realism and the
Foundations of Ethics (New York: Cambridge University Press, 1989), 40–2.
66
The uncovering of a ‘descriptive fallacy’ comes in Austin, How to Do Things with Words, 3.
RMH often acknowledged that he was taking up this account from Austin—e.g. Objective
Prescriptions, 19.
67
Expressivists after RMH have revised his account so as to talk about the acceptance of
norms. See Allan Gibbard, Wise Choices, Apt Feelings: A Theory of Normative Judgment
(Cambridge, MA: Harvard University Press, 1990), 71–80. But the distinction between meaning
and criteria is preserved; we can talk about what norms to accept, once we agree that accepting
norms together is what we are after.
68
Many of these ‘descriptive’ terms are already tinged with value. Take the term ‘fragrant’, for
example, and compare it with the term ‘smelly’. What is it for a strawberry to be ‘ripe’, other than
‘appropriate for eating’? RMH acknowledged that it is actually very hard to find purely descrip-
tive language.
120 God’s Command
run around trees right-handed. RMH held that it was extremely odd, but still
intelligible as a moral judgement; that we would know what a person who
made such a judgement meant, even though we would have no inclination to
agree with it. It is significant, however, that RMH agreed to the following
point:
If a man said that somebody was a good man because he clasped and unclasped
his hands, we should at first find ourselves wondering whether we had under-
stood him. . . . We do not think that men who do this are good. The explanation
of our not thinking this is that such choices would hardly contribute to our
survival, growth, procreation etc.; if there have been any races of men or animals
who have made the clasping and unclasping of hands a prime object of their pro-
attitudes, to the exclusion of other more survival-promoting activities, they have
gone under in the struggle for existence. I am, I know, being rather crude; but in
general, to cut the matter short, we have the pro-attitudes that we have, and
therefore call the things good which we do call good, because of their relevance to
certain ends which are sometimes called ‘fundamental human needs’.69
This passage is remarkable because of its similarity to many things in late Foot.
The difference is just that these considerations about the human form of life
and its evolutionary history were located by RMH as constraints upon criteria,
whereas Foot did not admit the meaning/criteria distinction.
There is a second, more significant, place that RMH and Foot disagreed,
and this gives one reason for Foot’s rejection of judgement internalism. We
can take her example of Alec D’Urberville, from Hardy’s Tess of the D’Urber-
villes, who says, on the night he seduces Tess, ‘I have lived bad, and I shall die
bad’, meaning what he said but without the slightest intention to reform.70
Foot called this kind of disposition ‘shamelessness’. She thought it showed that
a person may make a full-fledged moral judgement without endorsing the
norms he is referring to in the judgement. RMH’s response to this was that
Alec D’Urberville is most probably rejecting conventional morality.71 He is
saying that he has lived and will die by standards that most people condemn,
and he defies the standards they uphold. There are many different kinds of
case here. In RMH’s terms there are cases where there is not commending
going on, but some other speech act (as in the case of D’Urberville, defying),
and different cases where there is commending, but not subscription to the
commending. One important example of the second kind of case is the
depressed person who says: ‘It is wrong for me to be lying here in bed at
two o’clock in the afternoon.’72 RMH wanted to emphasize that there is
something non-standard or defective about all these cases.

69
RMH, ‘Descriptivism’, in Essays on the Moral Concepts, 72. 70
NG, 19–20.
71
RMH, Objective Prescriptions, 87–95, esp. 95.
72
In RMH’s terms, the first kind of case is a tropic defect and the second is a neustic defect.
See my ‘Prescriptive Realism’, 83–101, especially 85–7.
 Can We Deduce Morality from Human Nature? 121
We could put this in terms of a natural-history story. The human form of
life needs not only norms—for example, norms of justice—to hold us together,
but also ways to express to each other that we are committed to such norms.
We need a form of expression that conveys, across a huge range of evaluations,
‘if I were you, I would’. We need this function because we cannot carry out our
characteristic human projects without it. We are dependent on each other to a
surprising degree for our decision-making about what to do and how to live.
Even if we do not explicitly ask others to tell us what they would do in our
situation, we are carrying around in our heads all the time implicit externally-
sourced prescriptions of this sort. Being social animals is a feature of our
thought life as much as our action. Moral language is plausibly construed as
having this social function. But, as with all functions, misuse or defective use is
also possible. The depressed person and the shameless person have lost the
ability at the relevant time to use this kind of language with its appropriate
function. It is like not being able to use a chisel except as a screwdriver.
This point about the function of evaluative language is what is essentially
right about judgement internalism. It is true that each side in the dispute can
explain the same phenomena. For Foot, Alec D’Urberville is making a full-
fledged moral judgement, but cannot live by it; and, for RMH, he is not
making a full-fledged, but rather a defective, moral judgement.73 But the
internalist account preserves one central contribution that evaluative language
makes to our form of life. For our present purposes, the key is the implication
of this disagreement for deductivism. RMH thought that his internalism about
judgement meant that no deduction of evaluative judgement from descriptive
facts was legitimate. But surprisingly, even if they were to agree that a full-
fledged evaluative judgement is an expression of some state of desire or
emotion or will, they could still disagree about whether the state of the
world being commended in such a judgement is a state of the world with
natural properties and evaluative properties that have some kind of mutually
implicative relation. This takes us to the second heading under which to locate
the disagreement about whether moral value is desire-based.

4.3.2. Moral Properties

Anti-realism about (moral) value is a thesis not about judgement, but about
the moral or evaluative properties that are picked out in such judgement. The
thesis is that these properties are not metaphysically real. It is rather difficult
to discern Foot’s view here, perhaps because (like Wittgenstein) she was

73
This is similar to the result in the previous dispute, but the other way round. For Foot, the
judgement that it is wrong to look at hedgehogs by the light of the moon could not be a full-
fledged moral judgement; for RMH it was full-fledged but extremely odd.
122 God’s Command
suspicious of this kind of metaphysical question. I think, however, that her
sympathies were with the metaphysical realist. This is certainly true of Hurst-
house, whose view is closest to John McDowell’s account of moral realism.74
What Foot called ‘natural goodness’ is an intrinsic property that things have,
which ‘depends directly on the relation of an individual to the “life form” of its
species’.75
RMH was, as Simon Blackburn puts it, ‘quietist’ on this second thesis,
holding that no real issue can be built around this kind of objectivity or
otherwise of moral value.76 He was agnostic about whether there are ‘real’
evaluative properties, but he was not explicitly anti-realist. He was not a
supporter of an error theory, or of the view (like Hume’s) that we project
our own desires onto the world as independent properties. So the interpreter
of RMH who thinks the metaphysical question about the objective reality of
these properties does make sense is in the same position as the interpreter of
Foot who shares that view. We have to speculate artificially about what our
authors would have said if they had thought this was a good question.
I suggested Foot’s sympathies would have lain with metaphysical realism.
I think RMH’s sympathies would not. He consistently held that the truth
conditions of moral statements are given by the criteria adopted by the
speaker.77 But, if we had asked about the evaluative meaning as opposed to
the criteria (or descriptive meaning), and asked whether there are real prop-
erties picked out, RMH would probably have said he did not know what we
meant. If pressed on just this point, he would have said there are not any,
because the claim is just nonsense. This explains why he was so often and so
consistently misunderstood. It is worth pointing out that, in relation to a
completely different kind of objectivity (for example, the objectivity of a good
umpire) RMH was an unflinching objectivist. The moral agent, like the umpire

74
John McDowell, ‘Virtue and Reason’, Monist, 62 (1979), 331–50.
75
NG 27. Foot’s view may be that there are moral properties, but they are deducible from the
facts of human nature. This combination is, roughly, the view of Richard N. Boyd. See ‘How to
Be a Moral Realist’, in Geoffrey Sayre-McCord (ed.), Essays on Moral Realism (Ithaca, NY:
Cornell University Press, 1988), 181–228, though he would go further and say not merely
‘deducible from’ but ‘reducible to’. Both Foot and Hursthouse think it is important to deny a
basis of ethics in ‘scientific foundations accessible from a neutral point of view’ (OVE 224).
76
Simon Blackburn, ‘Errors and the Phenomenology of Value’, in Essays in Quasi-Realism
(Oxford: Oxford University Press, 1993), 153. Alas, Derek Parfit misconstrues RMH on this
point, because he relies on an early popular piece that RMH subsequently corrected. See On
What Matters, ii. 410–13. There are no footnotes in this section, but Parfit is referring to an early
radio talk Nothing Matters, from 1957. When RMH republished this in Applications of Moral
Philosophy (London: Macmillan, 1972), 32–47, he explained (p. 40) that he was using ‘subject-
ivist’ in a popular way, because ‘I thought it would be more familiar to my listeners’. His
preferred view is not that values are the same thing as feelings or experiences of value, but
that my intuition of the value of an act is the same thing as that act arousing in me an attitude of
disapproval.
77
RMH, Objective Prescriptions, 16–18.
 Can We Deduce Morality from Human Nature? 123
whose daughter is on one of the teams, gives no special weight to his own
preferences.
My own view on these first two topics is what I call ‘prescriptive realism’,
and agrees with RMH about motivation but disagrees with what he would
probably have said about moral realism.78 Judgement internalism is a thesis
about moral or evaluative judgement, and realism is about moral or evaluative
properties, and there is no reason why we should not say that there are indeed
these properties, but that, when we make judgements about them, we not only
claim that they exist, but express an attitude of emotion, desire, or will.79 If we
do say this, we will be both expressivist and realist, expressivist about the
judgements and realist about the properties, in the sense that they are there
whether the relevant attitudes are there in the person making the judgement or
not.80 Why should we want to be realist about the properties? Our evaluative
language strongly suggests an ontological commitment that, to take Nicholas
Sturgeon’s example, any full causal explanation of the events of Hitler’s life
requires reference to his moral depravity.81 It is certainly possible to propose
an error theory here that there is indeed this kind of commitment, but it is
mistaken.82 Error theories, however, are a last resort, and we would need to be
shown that there is some completely persuasive metaphysical principle that
rules out the reality of moral and evaluative properties. For a theist in
particular it is going to be hard to find such a principle. For example, a theist
will find it implausible to say that only material things have causal properties
and that only causal properties are real. The point of prescriptive realism,
however, is that, even if we concede the reality of the moral and evaluative
properties, we do not have to deny the insight of the expressivists about one of
the central functions of moral and evaluative judgement, the function of

78
See my ‘Prescriptive Realism’, and God’s Call, ch. 1.
79
This combination is also suggested by David Copp, in ‘Realist-Expressivism: A Neglected
Option for Moral Realism’, Social Philosophy and Policy, 18/2 (2001), 1–43. See also his ‘Milk,
Honey, and the Good Life on Moral Twin Earth’, Synthese, 124 (2000), 113–37. My difference
from David Copp is that he wants in the end to deny that he is any kind of internalist except what
he calls a ‘discourse internalist’, which is restricted to the kind of expression that is consistent
with the absence (even dispositionally) of the relevant attitude. Another author who suggests a
similar combination is Linda Zagzebski in Divine Motivation Theory (Cambridge; Cambridge
University Press, 2004) 11–15.
80
When I say that the properties are ‘there whether the attitudes are there or not’, I mean that
their existence is independent of the attitudes of the person making the judgement. I have been
persuaded, however, by John McDowell that the properties can be ‘real’ without being inde-
pendent of human faculties in general, and that they are in this way like secondary qualities such
as colour. See McDowell, ‘Values as Secondary Qualities’. See also my God’s Call, 25–33.
81
Nicholas Sturgeon, ‘Moral Explanations’, in Sayre-McCord (ed.), Essays on Moral Realism,
229–55. I have also been influenced by Parfit, On What Matters, 58–82. See also Angus Ritchie,
From Morality to Metaphysics (Oxford: Oxford University Press, 2012) 21–39.
82
We return to Mackie’s error theory in Section 8.3.1.
124 God’s Command
allowing us to coordinate our lives together by expressing in these judgements
our commitment to live a certain way.
Again, what is important for present purposes is the implication of this
disagreement for deductivism. Even if we allow, with the realists, that there are
evaluative properties independent of our judgements about them, the case still
has to be made by a deductivist that there is an implicative relation (inde-
pendent of a decision of principle) between natural facts and moral goodness.
Even if RMH were to agree on the realism, he could still disagree on the claim
about implication. That claim will be the issue of Sections 4.4.2 and 4.4.3.

4.3.3. Ideals

The third topic is a particular kind of value, which we can call ‘ideals’. Foot
gave an excellent example in her discussion of happiness.83 She considered the
case of some brave men who opposed the Nazis, and were captured, and then
wrote letters from prison to the people they loved. We can tell from their
letters that the writers ‘were especially well fitted for the enjoyment of the best
things in life: for great happiness’. It is tempting, therefore, to say that they
knowingly sacrificed this happiness. But Foot said this is not the whole truth.
For there is also a sense in which they recognized that, because of the state of
their homeland, happiness was no longer possible for them; it was, so to speak,
already too late for happiness. Happiness in this sense was an ideal no longer
realizable. This does not mean merely that, if they had compromised, they
would have subsequently felt shame or self-disgust (though that might be
true). For we can imagine they were offered a ‘Lethe-drug’ that removed all
future memory of the compromises they had made (including, presumably,
the taking of the drug). They would not have accepted it. They would not ‘have
felt that happiness lay in acceptance’, even if they then experienced a satisfac-
tion of all their desires for their concurrent experience. Another example is
that we want the people we love not to forsake their virtue when life makes
virtue difficult for them. This kind of wanting is an ‘ideal’ preference. The
preference is ideal in the sense that it does not itself depend upon the
maintenance of our or their desire for the kind of life preferred. Foot claimed
that the virtuous person wants to go on being virtuous (or wants those whom
she loves to go on being virtuous) even if at some future time she no longer
wants that (or those whom she loves no longer want it).
Putting the matter in terms of ‘ideal’ preferences allows a connection with
RMH. He was never satisfied with what he had to say about ideals, or about
fanatics who hold ideals. This matters for the present chapter because we are

83
NG 94–6.
 Can We Deduce Morality from Human Nature? 125
discussing whether we can deduce moral goodness from our nature. If a large
part of the goodness of a human life is specified by our ideals and most ideals
are not deducible from our nature, this will put the deductivist in a hard
position. That will be the conclusion in Section 4.4.3. The goal of the present
subsection is to show that RMH was also in difficulty about ideals. He was
unable to accommodate ideals in his version of utilitarianism, which analysed
morality as universalized prudence, meaning that moral thinking has a two-
step process: we first determine what prudence dictates from each person’s
point of view who is affected by our decision, and we then make a moral
decision by giving equal weight to all those points of view including our own.
In Moral Thinking he defined ‘prudence’ stipulatively by saying that prudence
leads us to satisfy over our lives as a whole our now-for-now and our then-for-
then preferences. If I have a preference now for an experience now (say, a
chocolate ice cream), prudence will tell me to satisfy it. It will also tell me that
in ten years’ time, if I then want chocolate ice cream, I should get it. It does not,
however, tell me that if I now want the ice cream in ten years’ time, that
preference should then be satisfied. Rather, my life will go well as a whole from
the perspective of prudence if I satisfy most of my preferences at the time
I have them for my experience at those times. Note that the preferences here
are all preferences for experience to go a certain way, and we might call such
preferences ‘basic preferences’.
Perhaps an example of a now-for-then preference will help. I remember
that, when I was an adolescent at a boarding school and was miserable,
I observed middle-aged men saying that their boarding-school years were
the best years of their lives. I vowed to myself not to say that, and I also had
the thought that, even if I were to change into the sort of person who did say
that, I did not now want any accompanying preference to send my son to
boarding school to be satisfied. This is a now-for-then preference. In RMH’s
terminology it is a kind of ideal. A fan of a team wants the team to win, but this
desire is not an ideal, because she wants the team to win only so long as she
supports them. But someone who thinks it is degrading for a person to have
sex for money probably has an ideal of human dignity that is offended by such
practices even if every party is getting what he or she wants and no disap-
prover will ever know about it. If I have an ideal of faithfulness in marriage,
I may prefer that my wife not be unfaithful even if I never find about it and
even if there is no change in the quality of my experience of our life together.
In Moral Thinking RMH openly admitted that he could not accommodate
ideals within his theory.84 But he went on to argue that it would not make
much difference in practice, because most idealists do not in fact have strong
enough preferences to prevail in the overall judgement about what to do. If we

84
RMH, Moral Thinking, 104: ‘I am inclined to think that we ought [to accommodate them],
but that I cannot at present do so. There is obviously unfinished business here.’
126 God’s Command
imagine an archangel who knows the preferences of all the affected parties,
and the intensity of those preferences, the archangel who wants to maximize
preference satisfaction will override the relatively weak ideal preference, for
example, of a doctor who wants to save her patient’s life at any cost in suffering
to the patient. The archangel will prescribe the satisfaction of the patient’s
intense basic preference not to suffer. Only if the doctor has an extraordinarily
strong ideal preference will it prevail. Moreover, RMH argued, the doctor can
usually change her preference, or at least its strength, and therefore she should
do so. But there is a problem here, elegantly displayed by Allan Gibbard.85 He
shows that RMH’s proof of utilitarianism does not work if we try to make it
cover ideal preferences. Moreover, if a person can change an ideal preference,
does it follow that he is rationally required to do so? We need some criterion
from the theory to tell us which personal ideals and which universal ideals
should change and which should not, for we do not want the result that pig-
headed people always get what they want because they change their prefer-
ences less easily.86 Ideals can be good or bad, and many of them derive from
our conception of what Section 4.2 called the ‘competitive’ goods. RMH’s
central example was the Nazi ideal of Aryan purity. If our conception
of the good human life cannot be deduced either from maximizing basic-
preference satisfaction, or from our nature, we will need some other standard
for discernment.
For both RMH and Foot, religion provided central cases of ideals. The
letter-writers Foot described were Christians. RMH’s central example in
Freedom and Reason was St Francis. For Hursthouse, piety to the Judeo-
Christian God is a virtue that ‘undoubtedly brings great joy and serenity to
its possessors, [but] no atheist can regard such joy as “characteristic of human
beings”, that is, as something that reason can endorse’.87 But we need to

85
I am following here Allan Gibbard’s argument in ‘Hare’s Analysis of “Ought” and its
Implications’, in Douglas Seanor and N. Fotion (eds), Hare and Critics (Oxford: Clarendon
Press, 1988), 52–72. Gibbard considers the case of Cheops, who has the ideal personal preference
for an expensive funeral. Gibbard shows that, according to RMH’s theory, which explicitly
refuses to consider ideal preferences, archangels will not necessarily agree about whether
Cheops’s preference should be satisfied. This shows that RMH’s proof of utilitarianism does
not work if we try to make it cover ideal preferences.
86
RMH conceded in Seanor and Fotion (eds), Hare and Critics, 233–4, that ‘when I wrote
Moral Thinking I thought that such a longer way [of ruling out fanatical ideals] might be found,
and it still may’. But there is nowhere to be found in his writings after Hare and Critics any
statement of a different and more satisfactory longer way. One of RMH’s students, Peter Singer,
adopted RMH’s universal prescriptivism, but in his contribution to Hare and Critics, 147–59,
departed from RMH’s qualification about ideals. For Singer at that time preference-satisfaction
covered the whole of moral thinking. But later, after being convinced by Derek Parfit’s On What
Matters, he came to reject the Humean view of the relation between value and desire, and to
accept that there are values that obtain whether we desire them to obtain or not. This means that
he is back in the same difficulty as RMH, the difficulty of discerning which ideals are to be
endorsed and which are not. See my ‘Morality, Happiness, and Peter Singer’.
87
OVE 233.
 Can We Deduce Morality from Human Nature? 127
acknowledge how utterly pervasive ideals are even in ordinary non-religious
thought about the good life. In Freedom and Reason RMH considered the
ideals of the ascetic and the bon vivant. But consider also how private some
people are in opening their lives to others, while other people do not seem to
care about privacy at all. Or consider that brave people may not want
themselves later to give in to timid desires even if they later become timid
people. All these ideals have universal as well as personal forms. The most
important point is this. Morality itself is a universal ideal if the moral agent
prefers that she herself and everyone else live morally whether she continues to
have that preference or not. The preference to approximate the archangel’s
thinking is itself a universal ideal, and the morally good life is not dependent
for its value upon anyone desiring it.
RMH introduced the figure of the archangel so as not to have to talk about
God. The archangel has ‘superhuman powers of thought, superhuman know-
ledge and no human weaknesses’.88 The archangel has two sorts of decisions to
make; the first (at the level of ‘critical thinking’) is what should be done,
situation by situation, so as to maximize preference satisfaction impartially,
and the second is what intuitions or principles we mere humans should live by
(at the level of ‘intuitive thinking’) when we are not trying to approximate the
critical thinking of the archangel. Sometimes RMH acknowledged that he was
in fact talking about God. Most Christians, he said, will think ‘that God has,
and we have to a much more limited degree, the means (rational moral
thinking) wherewith to resolve at the level of critical thinking conflicts
which arise at the intuitive’.89 The fact that God was his model of critical
thinking supports the claim that this kind of thinking is an ideal.
RMH agreed with Foot that ‘happiness’ in one sense of that protean word
includes an evaluative component. For Foot, this was the kind of happiness
that the letter-writers do not sacrifice, but rather recognize has already become
impossible for them. For RMH, one example was the opium addict, ‘whom we

88
RMH, Moral Thinking, 44.
89
Ibid. 34. See ibid. 99: ‘Archangels can do [this full representation and identification], and of
course God.’ His conception of God was influenced by the Sermons of Joseph Butler, as mediated
through C. D. Broad’s Five Types of Ethical Theory (London: Kegan Paul, 1930). (I own RMH’s
copy of this book, with his markings in it.) Butler says: ‘It is manifest that the common virtues,
and the common vices of mankind, may be traced up to benevolence, or the want of it. And this
entitles the precept, Thou shalt love they neighbour as thyself, to the pre-eminence given to it; and
is a justification of the Apostle’s assertion, that all other commandments are comprehended in it.’
For humans, he concedes, stating what virtue and right behaviour consist in requires cautions
and restrictions. But ‘We have no clear conception of any positive moral attribute in the supreme
being, but what may be resolved up into goodness.’ For Butler, God’s aim is our happiness (this is
God’s benevolence), but God also reveals to us particular precepts—for example, of fidelity,
honour and strict justice—which we would not be able to work out merely from the desire to
make each other happy. This part of Butler RMH rejected. But this left him with the problem
about ideals I have been discussing.
128 God’s Command
would not say is happy (really happy)’, even if he always has enough opium.90
But he did not want to build a disputed normative conception into the very
terms of his theory of moral thinking. This fact about the word ‘happiness’ led
him away from the formulation of his basic normative principle in terms of
‘the greatest happiness of the greatest number’ to some more neutral formu-
lation in terms of maximizing preference-satisfaction. But much of our moral
thinking, including our commitment to morality itself, is in terms of ideal
preferences and is not reducible to now-for-now and then-for-then prefer-
ences for experience. We need some way to determine which ideals to try to
live by. The archangelic method is insufficient. The next section of this chapter
argues that the way of deduction from facts about human nature is insufficient
also, and for some of the same reasons. We should learn from the references to
religion in RMH and Foot and Hursthouse that faith in God and receptivity to
divine command can give us a way to select the ideals that shape our
conception of the good human life, even when basic-preference satisfaction
and deduction from our nature do not give it to us.

4.4. FOOT AND H URSTH OUSE ON DEDUCTIVISM

Foot gave three striking examples that can organize this section of the chapter:
an example from vegetable life (‘good roots’), from non-human animal life
(‘the good wolf ’), and from human life (‘the promise-keeping photographer’).
We will look at these, and then consider some additions to the theory by
Hursthouse. But, before turning to the examples, it is helpful to say something
more general to introduce the main criticism, and to say how this criticism
relates to the previous part of the chapter. The main criticism is that Foot and
Hursthouse treat our nature too much as a single united package, and they are
too optimistic in their account of practical rationality as sensitivity to the
reasons this package gives us. Our nature does indeed give us reasons, but
some of them are good and some of them are bad. We therefore need some
way to discern which reasons given us by our nature we should follow.91

4.4.1. Too Much and Too Little

Our nature is both too much and too little to allow us to deduce conclusions
about moral goodness. It is too much, because the promptings it gives us are a
mixture of good and bad, and are therefore not a reliable source. Theologically

90
RMH, Freedom and Reason, 127.
91
See the discussion of this by al-Maturidi in Section 6.2.3.
 Can We Deduce Morality from Human Nature? 129
this is a reading of the doctrine of the Fall: that we are now corrupt in every
part (though not totally corrupt in every part).92 But there is also a philosoph-
ical ‘translation’ that Kant has given us of this theology. He says that we are
born with both the predisposition to good and the propensity to evil.93 This is
not, therefore, ‘Calvinist Sociobiology’, in Frans de Waal’s misleading phrase,
which denies the good.94 Our nature prompts us to both prudence (in the
sense of enlightened self-interest) and morality, and sometimes not even to
prudence (if we simply follow our desires as they come). This gives us
therefore a tension. It is not that prudence and morality are unconnected.
On RMH’s theory, for example, morality (excluding ideals) was analysed as
universalized prudence. But he agreed with Sidgwick that rationality does not
itself tell us to take the step to universalization. In Sidgwick’s terms, both
egoistic hedonism and universalistic hedonism are self-evident, but taken
together they are mutually inconsistent. The question remains open of why,
when prudence and morality conflict, we should choose morality. Foot and
Hursthouse do not accept much, if any, of this. But do they have the resources to
answer this question, which Chapter 1 called ‘the normative question’, following
Korsgaard? If not, the appeal to nature will fall into the same pattern as the appeal
to reason, or intuition, or the authority of one’s cultural norms. We will need to
know which reason, which intuitions, which cultural norms, and now which
promptings of nature we should follow.
Our nature gives us both too much and too little for the deduction of the
requirements of morality. We have so far considered why it gives us too much.
For the other half of this, the ‘too little’, there is in the same way a theological
source and a Kantian philosophical (though still theist) translation. The
theological source is the doctrine of sanctification, on one reading a restor-
ation to righteousness by the work of the Spirit.95 One philosophical transla-
tion is in terms of Kant’s picture of the two concentric circles of revelation
described in Chapter 1. In that chapter the second argument for the depend-
ence of morality upon religion was the argument from grace, and the doctrine
of grace belongs in the area of the outer circle immediately adjacent to the
inner one (the revelation to reason). Since we are not able by our own
resources to overcome the propensity to evil expressed in our fundamental
maxim, we need God’s grace in order to live a human life pleasing to God. But
we will not find in the facts of our nature, as Foot and Hursthouse construe
them, either this kind of assistance or an example of this kind of life. In this
sense, nature gives us too little for the deduction.

92
See Calvin, Institutes, I. 1. 8: ‘diffused into all parts of the soul.’
93
Rel. vi. 26–32. For his use of the term ‘nature’, see Ch. 3, n. 87.
94
Frans de Waal, Good Natured: The Origins of Right and Wrong in Humans and Other
Animals (Cambridge, MA: Harvard University Press, 1996), 17. But he is wrong about Calvin;
see Institutes, I. 5. 4: ‘yet that seed remains’.
95
Calvin, Institutes, III. 3. 19.
130 God’s Command

4.4.2. Good Roots and Good Wolves

Despite this objection, we should accept one central point from Foot and
Hursthouse. There is a natural goodness that is conducive to the good life, or
simply the good for both animals and plants. One example she used of
vegetable life is the roots of an oak tree. The roots play a part in the life of
the tree; they obtain nourishment. The development of roots is therefore a
norm, she said, rather than merely a statistical normality, like the noise made
by the rustling of leaves. It matters in the life of the organism, and its absence
would be a defect. ‘Oak trees have deep roots’ is what she called, following
Michael Thompson, an ‘Aristotelian categorical’.96 Goodness in the roots is
their ability to carry out this contribution to the life of the organism, and we
can deduce this goodness from this ability. Here is an acceptable form of
deductivism, from the roots providing nourishment to what Foot calls the
‘primary goodness’ of the roots. This is not yet, however, moral goodness.
A ‘secondary’ goodness is a contribution to something external—for example,
to human life. Perhaps a good apple tree produces fewer, larger apples, but this
is a merit for us, not for the life of the tree in itself (which maximizes its
reproduction by producing lots of smaller apples). We should accept that there
is a primary goodness in oak trees, in a particular oak tree, and in its roots.
By contrast, RMH denied the deduction of goodness from any description
(including ‘the categorical’). In his reply to the article by Geach discussed
earlier, RMH said:
[Geach is] the latest of a famous succession of thinkers who have systematically
confused ‘what a thing can typically or does typically do’ with the quite different
notion . . . ‘what it is specifically good for it to do’. Plato was of course the
principal culprit. The word ‘function’ has perhaps been used to cover all these
notions. The assimilation between them is only justified if we accept the assumed
premiss Natura (sive Deus) nihil facit inane [Nature (or God) does nothing in
vain]. Anyone who feels attracted by Geach’s use of this kind of reasoning should
first read Aristotle, Politics 1252a35, where a similar premiss is used in order to
justify slavery and the subjection of women (cf. also 1253a9).97
If we accept that there is such a thing as the primary goodness of a tree, we
need to say what it means to think of a tree as good (or the life of a tree as
good). It is helpful to return to the account of goodness as desirability in
Chapter 1, reached in response to Stump’s account of Aquinas. To say sin-
cerely that something is good is to express that one is drawn by it and to
endorse the claim that the thing deserves to draw one in that way. For
Aquinas, in Stump’s reading, goodness belongs to everything that is, and

96
NG 46.
97
RMH, ‘Geach: Good and Evil’, in Essays on the Moral Concepts, 37 n.
 Can We Deduce Morality from Human Nature? 131
degrees of being and degrees of goodness are coextensive. Here would be a way
to think of a tree as good: a tree is good because goodness belongs to
everything that is. But the notion of degrees of goodness is a great deal easier
for contemporary philosophy than the notion of degrees of being (though
ordinary language is less restrictive). Can we give an account without this
Thomist metaphysics of why life as a whole draws us and deserves to draw us?
The first section of this chapter discussed the relation between God and the
variety of life (the two kinds of lobelia). The four answers given there would
give reason to think that every kind of life that God creates is good. Along
these lines we can give a theistic account of the goodness of a tree, and
derivatively of the goodness of its roots. It may be that talk of the goodness
of a tree without such an account is like what Anscombe says about talk of
obligation without an account of a divine commander.98 But perhaps we can
be satisfied with at least a circular account of the goodness of an oak without
talking about God, if we think of this goodness as consisting in the range of
features possessed by mature oaks that are flourishing, and we think of this
goodness as what the oak is aiming towards. This term ‘flourishing’ indicates
that this is a circular definition, since flourishing is doing well. G. E. Moore was
right that we should not expect to be able to eliminate evaluation from the
account of any type of goodness. But this language of ‘aiming towards’ is the
language of final causation, and, while it is true that we make use of it
continually for organisms, in both lay and professional talk, it is not clear
whether it can be validated within the strict terms of the biological sciences.99
Can we make sense of the idea that animals have more value than plants
in general, though this may not be true in all cases?100 Yes, if there is value in the
things animals can do that plants cannot. There are indeed dangers in this kind of
hierarchy (which results in what Aristotle says about women and slaves), but
Aristotle could be right about plants and animals and wrong to deny that
all humans have the same basic value. On the view defended in Chapter 1, all
humans have the same basic value because they equally receive God’s call, not
because they are now equally capable of valuable activities. We can also say that

98
See Julia Driver’s entry under ‘Gertrude Elizabeth Margaret Anscombe’, in Edward
N. Zalta (ed.), Stanford Encyclopedia of Philosophy http://plato.stanford.edu/archives/win2011/
entries/anscombe/ (accessed 15 September 2014). In relation to Anscombe’s famous article
‘Modern Moral Philosophy’, Philosophy, 33 (1958), 26–42, Driver says: ‘One can conclude that
Anscombe is arguing that the only suitable and really viable alternative is the religiously based
moral theory that keeps the legalistic framework and the associated concepts of obligation.’
99
Thomas Nagel, in Mind and Cosmos (New York: Oxford University Press, 2012), proposes
a return to Aristotle’s central notion of natural teleology.
100
Can we confine natural goodness to things that are alive? Why should not rivers be good,
and mountains? We do not need to settle this question here. Peter van Inwagen, Material Beings
(Ithaca, NY: Cornell University Press, 1990), 81–97, following Aristotle, argues that only life has
the right kind of dynamic unity to be able to order material stuff into a persisting subject of
change, rather than a mere ‘heap’.
132 God’s Command
non-human animals have some of the same value as humans, being able to do
many of the same things. But whether some of them also have a divine call, we do
not know.
Even if we can give an account of the goodness of a tree, however, this is not
what Foot was talking about when she said that the roots have a ‘function’. She
distinguished ‘function’ from ‘adaptation’ as follows: ‘To say that some feature
of a living thing is an adaptation is to place it in the history of a species. To say
that it has a function is to say that it has a certain place in the life of the
individuals that belong to that species at a certain time.’101 But, if we are
considering individuals, we have to recognize that through microevolution
there is constant genetic variation, and moreover there is phenotypical vari-
ation from the same genotype. Foot tried to limit the variation by confining
her attention to features that ‘have to do, directly or indirectly, with self-
maintenance . . . or reproduction’ (emphasis added). But even the rustling of
the tree’s leaves in the wind, which she says is not functional, is indirectly
related to their falling and the tree’s annual life cycle. Here RMH’s point is a
good one. The individual plant can do and typically does do all sorts of things
that are not conducive to its good. It grows too tall, is not sticky enough, or is
too close to its neighbours. Even if we confine its good to its self-maintenance
and reproduction, the plants are in competition with each other, and not only
with other species; there are strong specimens and weak, and just as many
weak as strong. There is no deduction from a particular plant’s typical
performance to its doing well or from the typical performance at a time for
the set of members of a species to the species doing well.
Here Hursthouse has a corrective, conceding that in some cases it will be
‘quite indeterminate’ whether an individual x is overall a good x, and that even
an individual ‘perfectly endowed in every relevant respect’ may still not live
well given its circumstances.102 She adds some useful structure. She says that
there are two ends we share with plants—namely, survival and reproduction—
and two ends we share with non-human animals—namely, pleasure or the
absence of pain, and (for animals that live in society) the well-being of our
social group.103 These are the natural ends against which we can measure
whether some human life is a naturally good life, and there is no fifth such end.
We will come back to this in the final subsection of the chapter. Both Foot and
Hursthouse accept that these ‘natural facts’ about the various forms of life and
their functions are themselves value-laden and are not simply statistical. But,
while this does indeed make their position immune to the objection from the
nasty parts of our nature, it does so at a price. We are left without a way to say
why some dispositions to pursue these four ends are good and some disposi-
tions to pursue those same four ends are not. Even with plants, the result of

101 102 103

NG 32 n. OVE 204–5. OVE 197–201.
 Can We Deduce Morality from Human Nature? 133
Hursthouse’s corrective is to make the primary good of the oak frustratingly
indeterminate. We think immediately of some great specimen tree dominating
its surroundings. But the thesis about primary goodness has to apply to life in
general, including the acorn just beginning to germinate; bare survival and
reproduction are sufficient.
Now, we can go on to non-human animals. Foot said about wolves:
There is something wrong with a free-riding wolf that feeds but does not take part
in the hunt, as with a member of a species of dancing bees who finds a source of
nectar but whose behaviour does not let other bees know of its location. These
free-riding individuals of a species whose members work together are just as
defective as those who have defective hearing, sight, or powers of locomotion.104
Again, RMH did not like the deduction. He used to have on the wall of his
room at Corpus Christi College a diagram called ‘The Good Pig’ showing the
right cuts of meat for the butcher’s trade. He also said, against Geach, that
‘horse’ by itself is not a functional word:
Similarly, if ‘horse’ is used as a functional word, meaning ‘charger’, a horse that
throws its rider becomes eo ipso a bad one; but the horse might say to himself ‘I’m
not trying to be a horse in that sense; I’m only a solid-hooved perissodactyl
quadruped (Equus caballus), having flowing mane and tail’, and proceed to throw
his rider without offence to anything except the rider’s standards. . . . The horse
breaker’s art would be easy if one could turn horses into chargers by definition.105
What is at issue here is the distinction between what Foot called ‘primary’ and
‘secondary’ goodness. A particular kind of pig or horse is useful to humans, for
eating or riding. That is a secondary goodness. But the question is whether
there is a kind of goodness for the pig or the horse in itself. RHM said that
‘horse’ is not a functional word, like ‘screwdriver’, but this does not show that
there is not a primary goodness of horses. We have already conceded that it
makes sense to say there is. So far, Foot was right. RMH’s remark quoted
earlier, that we call things good because of their ‘relevance to fundamental
human needs’, suggests that he thought there was a primary goodness at least
of human beings.
One complication, however, is that RMH’s examples were of domesticated
animals, which have been bred so as to serve human uses. Foot’s examples
were of wild animals, the wolf and not the dog. Should we say that there is no
longer a natural goodness for dogs, or perhaps that their natural goodness is
still what would be good in the wild? Foot said that the deer in the zoo that
cannot run fast is defective, even though this may be no disadvantage for

104
OVE 16.
105
RMH, ‘Geach: Good and Evil’, in Essays on the Moral Concepts, 38. The definition of horse
is quoted from the OED.
134 God’s Command
defence or feeding or mating or rearing the young in the zoo.106 So defect or
natural goodness in an individual is relative not to the actual environment of
the individual, but to the normal habitat of the species. There are many
difficulties here. Consider melanism in moths. Apparently some kinds of
moth change the darkness of their colouring with increased pollution, so
that they remain inconspicuous against darker surfaces. With the individual
moth, what should we say? Maybe the pollution is like the zoo, so that we
should take the standard to be what life would have been without human
intervention. But individual animals are constantly moving into new habitats,
or their present habitats are changing in some respect. One key example is the
climate, affecting almost all the species on earth. Is it a defect in an individual if
there is a mutation that allows it to live at a warmer temperature? Again, the
climate change might be like the zoo, at least partially a product of human
intervention. Consider, for a final example, hypertrophism, as in the develop-
ment of enormous antlers, which may give individuals an advantage, but
which takes a species towards extinction. Is the individual with extra-large
antlers defective? Again, there seems no determinate answer. Traits can take
off in this way within a species. Should we say they are functional but not
adaptive? Competitive advantage within a group can be counter-adaptive to
the species. A biology textbook says: ‘In short, sex has led males to grow too
large, and burdened them with appendages that are too demanding for their
own ecological good.’107
The central case for the present chapter is the free-rider wolf. Is it defective?
One reason this is important is that the cooperation of wolves is the kind of
thing Frans de Waal suggests is a precursor or requisite of human cooper-
ation.108 Let us suppose, since we do not yet know whether this is so or not,
that free-riding in the wolf and cooperation are both genetically based (though
they may have environmental triggers). So Jack the wolf is born with a genetic
mutation that predisposes him to eat the kill even if he has not hunted for it
with the pack. Perhaps the trait is continually invading, and is either taken up
or lost within a population. There can be, so to speak, an arms race within a
population, in which whether a trait is adaptive within a group depends on its
frequency of distribution within the group, and the relation to external threat
from outside the group. It seems likely that there is no determinate answer to
the question of what the good incidence is within a species. If we index the
question of defect to the ‘natural habitat of the species’, this is constantly
shifting at the micro-level. The basic problem here is that what Foot called
‘Aristotelian categoricals’ (from Michael Thompson)—for example, ‘bees have
stings’—work much better with an essentialist conception of species, like the

106
NG 34.
107
John Cartwright, Evolution and Human Behavior (Cambridge, MA: MIT Press, 2000), 126.
108
De Waal, Good-Natured, 39.
 Can We Deduce Morality from Human Nature? 135
one Aristotle operated with. Foot was aware of this difficulty, and tried to
confine the use of ‘function’ and so ‘natural goodness’ to individuals. But she
then told us that the standard for the individual is the normal life and habitat
of the species. How is this supposed to work in the case of Jack the wolf? The
wolf form of life is, very probably, in flux, and so is the form of life in particular
packs. Whether it is to the advantage of an individual wolf to cooperate or to
defect depends on where the ‘arms race’ has reached at some particular time
and place. Again we need the modesty in Hursthouse about whether there are
determinate answers in many cases to questions about whether an x is a good
x, and indeed about the very notion of a species, since she concedes that ‘the
different modes of classification are in part determined by different interests
of ours’.109

4.4.3. The Good Promise-Keeper

For human natural goodness Foot gave the example of the anthropologist
Mikluko-Maklay, as reported in Kropotkin’s Memoirs of a Revolutionist, who,
on a field trip to study the indigenous peoples of the Malayan archipelago,
promised to a native in his service that he would never photograph him. One
day, when the native was asleep, Maklay was tempted to photograph him, ‘the
more so’, Kropotkin says, ‘as he was a typical representative of his tribe, and
would never have known that he had been photographed’. But Maklay
refrained because he remembered his promise.110 Foot commented about
this case that in giving a promise ‘one makes use of a special kind of tool
invented by humans for the better conduct of their lives, creating an obligation
that (although not absolute) . . . harmlessness does not annul’.111 She thought
Maklay would have been justified in thinking that taking the photograph
would not have done any harm. But, nonetheless, breaking the promise would
have been defective. She liked Geach’s remark that ‘humans need virtues as bees
need stings’.112 She thought there was a ‘natural-history story’ to explain why
the disposition to break a promise is defective, just as much as there is a natural-
history story to explain why it is a defect not to be able to walk or see. She used
Anscombe’s story about the need for the institution of promising if we are going
to be able to get each other to do the sorts of things that constitute the human
form of life.
By contrast, RMH described why making a promise creates an obligation in
his reply to John Searle, given in ‘The Promising Game’.113 His argument was
that, if a speaker says sincerely that all promises are acts of placing oneself
under (undertaking) an obligation to do the thing promised, he must himself

109 110 111 112

OVE 202. NG 47. NG 51. NG 44–5.
113
RMH, ‘The Promising Game’, in Essays in Ethical Theory, 131–44.
136 God’s Command
be expressing his own subscription to the rule of the institution of promising,
and thus stating a moral principle. There is no deduction, therefore, from a
fact (for example, a person has uttered the words ‘I promise’) to an obligation.
‘It is characteristic of words like “promise”, which have meaning only within
institutions, that they can be introduced into language only when certain
synthetic propositions about how we should act are assented to.’114
Foot thought there was a deduction of our obligation to keep our promises
from our human form of life. Keeping our promises is an instance of justice,
she thought, and she said that justice is one of the virtues that is an ‘Aristo-
telian necessity’. A just person ‘may not kill an innocent person even for the
sake of stopping someone else from killing a greater number’.115 Thus, for
example (though she did not discuss it), in the famous ‘cave’ case, where either
a mother must smother her crying baby or the whole group in the cave will be
killed, we would know what she should do by deducing it from this account of
justice, which is in turn deduced from the human form of life. Foot was not an
absolutist about keeping promises. Apart from killing the innocent, torture is
the only absolute prohibition she mentioned. Torture was also an absolute
prohibition for RMH, who spoke out of his own experience as a prisoner of
war of the Japanese in the Second World War. But, though Foot did not claim
to be able to say just when breaking a promise is forbidden and when it is not,
she made two suggestions. The first is that simply the absence of harm is not
enough to make breaking a promise permissible. She thought that Maklay
would have been justified in saying that no harm was done in taking the photo
(though this is doubtful). There was, on the other hand, she said, harm in not
taking the photograph, because anthropological research did not then have the
record of that typical representative of his tribe. But here she thought she
could say the harm was not enough to justify taking the photo, because there is
what she called, from Anscombe, a ‘stopping modal’ forbidding promise-
breaking, deduced from the human form of life.
The deduction does not work. As stated earlier, she treated our nature too
much as a single united package, and she was too optimistic in her account of
practical rationality as sensitivity to the reasons this package gives us. Consider
the following general statements about how we achieve our natural ends of
survival, reproduction, pleasure, and the advantage of our group: ‘Humans lie’,
‘Humans cheat’, ‘Humans steal’, and ‘Humans commit genocide’.116 Are these
statements Aristotelian categoricals? Can we rule them out as irrelevant
because they are not ‘directly or indirectly related to our survival and

114
Ibid. 135.
115
NG 12. Breaking a promise is merely to act ‘badly’ on p. 46, but it violates an obligation on
p. 51. We return in Section 8.4.1, to trolley cases.
116
Kant has a number of similar examples about the ‘character of our species’: human adults
tend to do things like deceive themselves and others, and to avoid exposure to the suffering of
others, Anth. vii. 331–2.
 Can We Deduce Morality from Human Nature? 137
reproduction’ (or our pleasure or the advantage of our social group)? The
accusation here is not that Foot was trying to deduce moral goodness from
biology or from the inclinations we supposedly share with the hunter-
gatherers who formed most of our evolutionary history. Other philosophers
have tried to do this and failed.117 For example, Larry Arnhart in Darwinian
Natural Right argued that the good is the desirable (as in Aquinas) and the
desirable is what is generally desired by human beings. By ‘generally desired’
he meant that these desires are found in most people in every society through-
out human history, and he thought evolution had given us these desires
because they enhanced our chances of survival and reproduction.118 He listed
twenty such desires, and his framework principle was that if a desire is general
in this sense, belonging to this list, then its fulfilment is good. He did not find
disinterested benevolence among these desires, and he concluded that it is
merely utopian, ‘beyond the order of nature’, and foisted upon us by religion.
What he did find was the desire for social status, for political rule (though this
is, he said, a natural male desire not a natural female desire), for war (again a
male desire), for wealth (that is, enough property to equip for a good life, and
to display social status), and for justice as reciprocity. This last extends, as
Hume concedes, only as far as utility. If we imagine a society in which we
know that those whom we exploit are not able to harm us because they are
weaker than us, justice as reciprocity will not have any purchase over us.119
It is instructive to compare Foot with Arnhart on these points. Foot said
that there is the same form of inference for humans and for wolves, from the
Aristotelian categoricals about a form of life to conclusions about goodness.
But she insisted that human good ‘must be recognized as different from good
in the world of plants or animals, where good consisted in success in the
cycle of development, self-maintenance, and reproduction’.120 Humans, for
example, refrain from reproduction, because the demands of work to be done
‘give reason’ to renounce family life. Unlike Arnhart, she was not willing to
say, simply, that the human good is happiness. She pointed out that Wittgen-
stein said at his end that he had had a wonderful life, but she said that he was
not, in any ordinary sense, happy. Happiness is the human good only if we
think of happiness in the way we discussed in relation to the letter-writers
earlier, for whom it was already too late for happiness. But this kind of
happiness is an ideal, and there is the same kind of difficulty as we found
with RMH’s treatment of ideals. RMH had a worked-out theory about moral
thinking in terms of basic-preference satisfaction, and then had trouble

117
See the discussion of Larry Arnhart in my God and Morality, 65–72.
118
Larry Arnhart, Darwinian Natural Right: The Biological Ethics of Human Nature (Albany,
NY: SUNY Press, 1998), 17, 30, 66, 81–2, 124, but he is no longer committed to these views.
119
See Jonathan Harrison, Hume’s Theory of Justice (Oxford: Clarendon Press, 1981), 276.
120
NG 51.
138 God’s Command
integrating ideals into it. Foot had a worked-out theory about moral goodness
in terms of natural facts and then had trouble integrating into it the distinction
between the natural traits we should admire and the natural traits we should
not. She gave ‘humans recognize rights’ as an example of an ‘Aristotelian
categorical’. She did not give ‘humans desire power over others’. On this point
her normative conclusion was far preferable to Arnhart’s. But there is a price
for this. We know with Arnhart where his conclusions come from, even if we
do not agree with them. He faced the nasty as well as the nice aspects of our
nature, and (given his premiss about goodness and what is generally desired)
he was consistent in what he said about how we should live. He was consistent
in the same way Aristotle is consistent. For Foot, by contrast, there was a gap.
The categoricals for plants and non-human animals are supposed to be
reached by saying ‘how for a certain species nourishment was obtained, how
development took place, what defences were available, and how reproduction
was secured’.121 For humans, if we ask merely these questions, we will get
answers in terms of deception and coercion, just as much as the recognition
of rights. Foot was right to want a different way to think about the human good.
But she did not give us a method for doing so that is ‘naturalistic’ in the way the
claim about the same ‘form of inference’ from categoricals to virtues implies.
One basic problem is that the four natural ends given by Hursthouse,
mentioned in Section 4.4.2, do not cohere. This means that our nature is not
harmonious in the way Hursthouse needs and claims.122 In the cave case, the
good for the mother from the survival of her group conflicts with saving the
life of her infant. This is plausibly construed as a conflict between the second
end (reproduction) and the fourth (the good of the social group). The case of
the depressing parent discussed in Chapter 1 is plausibly construed in the
same way as a conflict between ends. It is not hard to think of these sorts of
cases. Perhaps the most important is that moral goodness should include
caring about humans who are in need all over the world, and we have not
been given an argument for this from the four natural ends. Hursthouse wants
to reject the view that human nature is ‘just a mess’, because she thinks this
leads to moral nihilism and despair.123 But she does not consider the possi-
bility that we are not exactly a mess, but a mixture of the kind Kant describes.
This means that we are, as she denies, a ‘battleground’.124 The view that we are
a mixture fits well, and without despair, the argument from grace in Chapter 1.
There is a dilemma here for Hursthouse. Either the Aristotelian categoricals
need to be already screened by ethical principle, in which case we get a

121 122
NG 33–4. OVE 260.
123
OVE 261. Hursthouse is replying to Bernard Williams, ‘Evolution, Ethics and the Repre-
sentation Problem’, in Making Sense of Humanity (Cambridge: Cambridge University Press,
1995), 109.
124
OVE 265.
 Can We Deduce Morality from Human Nature? 139
‘deduction’ from nature only by this screening. Or we can allow that any
typical feature leading to the four natural ends is a virtue, but then we will not
get the deduction of a conclusion about moral goodness or the good human
life. It is better to allow that most of what we think constitutes a good human
life comes from our ideals, which are not deducible from the four ends at all,
though these ends are constraints upon our ideals.
Another way to put this dilemma is that Hursthouse has two theses that
conflict, when conjoined, with her admission that much of the work in
deciding how to live does not come from the four ends, and that there is no
‘fifth end’ characteristic of human animals from which to derive these deci-
sions.125 These two theses are, first, what I will call ‘virtue dominance’ and,
second, deductivism about virtue. Virtue dominance is the view that disagree-
ment about right and wrong is usually the result of ‘incorrect application of a
virtue or vice term, or an incorrect judgement as to what a virtuous agent
would do’.126 But, if the virtues are to be deducible from our nature (so that
our relation to the virtues is like the bees’ relation to their stings), then they
ought to give us a great deal more content about how to live than the
admission that there is no fifth end implies.
We should concede that our nature puts a constraint on what we should say
about a good human life and therefore about obligation. Foot and Hursthouse
are right that it makes sense to talk about a human specific good, at least in
ordinary speech, and so to talk about the kinds of human goodness that
contribute to it. Even so, these facts about the human good and human
goodness do not themselves obligate us. If we allow a more expansive notion
of our nature than they do, and we say like Scotus that it is our end to enter
into the love of God, there will be one exception to this: we will have a self-
evident obligation to love God and the neighbour (as argued in Section 4.1.3
and in Section 1.3). But none of the more specific obligations of the second
table will follow. Foot and Hursthouse for the most part discuss virtue rather
than obligation, and at least Hursthouse is open to the possibility that Kantian
deontology may be necessary in addition to virtue theory.127 For a divine
command theorist, it is God’s command that obligates. We should have the
faith, however, that God wants our good (our being co-lovers), and commands
us to live in a way that will be conducive to this end. The first section of this
chapter was about Scotus and the ‘fittingness’ relation between the moral law
and our nature. So, even though obligations are not (with one exception)
deducible from facts about human nature, those facts can serve as constraints
on what we should believe about how God has commanded us to live.

125
OVE 218. One good example is temperance; see OVE 227, where the virtue deducible from
our nature seems to prescribe abstaining from some food or other, but does not tell us whether or
not to eat meat.
126 127
OVE 243. OVE 119–20.
140 God’s Command
Does divine command theory itself derive an ‘ought’ from an ‘is’? Can we
deducing an ‘ought’ from the statement that God commands something? The
answer is ‘no’, but the defence of it is subtle. Consider Satan, who says in
Milton’s Paradise Lost: ‘Evil be thou my good.’ Satan does not deduce from the
fact that God commands something that he (Satan) has an obligation. To the
contrary, Satan thinks that God’s commanding something suggests he should
not do it. The person who believes in God deduces from God’s command that
she is under an obligation, but belief in God is already a commitment of her
will.128 It is true that God’s commanding something makes it obligatory, and
that this is the right criterion (according to divine command theory) for the
judgement that we ought to live a certain way. But we have to make what is the
criterion our criterion, by a decision of the will. We could, and many people
do, choose something else.
For Foot and Hursthouse, once we have pointed out the ‘natural-history
story’, for example, about promising, the question ‘why should I?’ loses sense.
It is like asking for a reason to act rationally. Sometimes Foot adopted the tone
of the headmistress. In the case of a person’s own interests, she said, it is ‘silly’
to disregard one’s own future without special reason to do so.129 In the same
way with morality, she thought ‘most people know it is unreasonable to take
benefits and give nothing in return’. This does not mean that she thought
morality should always override prudence (in the sense of enlightened self-
interest). If you have promised to help a friend, but, when the day arrives, you
are in bed with a heavy flu, you do not have to fulfil your promise, even though
you physically could do so.130 Now it is not clear in this case what morality
prescribes. For both Kantians and Utilitarians, the agent counts as one and no
more than one. Probably, as long as the promised help is relatively minor, a
plausible Kantian morality tells you to stay in bed. But the point is that in these
sorts of cases practical rationality can give us contradictory maxims, both of
which fit ‘the facts of human nature’, unless we have rigged those facts by
incorporating ideals into their specification. It is not silly to be torn, even to be
torn apart. When we bring the interests of others into the picture, especially
the interests of those not related to us by friendship or family, most of us in the
richer parts of the world fail most of the time. We simply do not think about
the impact our own lifestyles have on those who are suffering in the rest of the
world. Foot was herself not blind in this way. But she was too optimistic about
the rest of us. If, therefore, the morally good life is organized by the ideal of
including this wider care, we cannot deduce this from our nature unless we
build it in by already screening what we will consider ‘natural’ through the
moral law. Hursthouse ends with the need for hope that we can flourish

128
RMH did a similar manœuvre with ‘I’, which he claimed is a prescriptive term when
I identify some future preferences as mine; see Moral Thinking, 221.
129 130
NG 23. NG 79.
 Can We Deduce Morality from Human Nature? 141
together, and not at each other’s expense, and she knows that this hope ‘used to
be called belief in (God’s) Providence’.131 If we cannot rely on our nature to
produce this ethical commonwealth, because our nature is a mixture of good
and evil, then what is the ground of this hope? It must be something beyond
our nature, and God’s sovereignty is an answer to be considered, as we did in
the argument from providence in Chapter 1.
This has been a long chapter, and it may be worth summarizing what was
attempted in it. We looked at Scotus’s view that the moral law cannot be
deduced from our nature, even though it fits it very well. This required an
account of how the goodness of different species relates to God. An argument
was given for the claim that the good for the human species constrains God’s
commands to humans, even though those commands cannot be deduced from
this good. For example, there is a presumption that can be overridden that we
should not take a command to be from God if it is a command to take an
innocent human life or tell a lie. This argument was connected with the
four Barthian constraints mentioned in Chapter 2. We then responded to
Murphy’s dilemma about whether non-moral non-divine facts are inert. We
denied the claim of ‘consensus deductivism’ that we can fix the reference of
‘good’ by the evaluations of most of the people most of the time. This denial
derived from considering how competitive goods are differently treated by
Aristotle and by Jesus. We ended by looking at the views of RMH, who
attacked deductivism, and Foot and Hursthouse, who defend a form of it.
We considered three features of RMH’s view: his internalist account of value
judgement (which we accepted), his (implicit) rejection of value realism
(which we denied), and his failure to integrate ideals into his overall utilitarian
account. We looked at the implications of these features for deductivism. Then
we examined the view of Foot and Hursthouse that moral goodness or virtue
can be deduced from the human form of life. We accepted their central point
that there is a primary goodness of organisms. But there is still a central
objection to their deduction of the human good: humans are a mixture of the
good and the bad. Either we include our endorsed ideals as a screen for what
we count as ‘natural facts’, or we exclude them. In the first case, we get a
deduction of our conception of the good human life from these facts, but only
by fiat. In the second case, we do not get a deduction at all. In both cases we
need something else, either to justify the screen or to give us a way to discern
good from bad natural impulses without the screen. If we believe in God,
God’s command is a promising candidate.

131
OVE 264–5.
 5

Barth on Divine Command

INTRODUCTION

This chapter discusses what Karl Barth says about divine command, and is
for the most part about Christian theology. There will be some concepts here
that I will use without trying to explain, such as ‘election’ and ‘second
coming’. A chapter such as the present one would be unmanageable if it
tried to proceed without assuming familiarity with some basic doctrines. The
next chapter is about the treatment of divine command in some medieval
Islamic thinkers, and Chapter 7 discusses the treatment of the same topic in
some recent Jewish thinkers. The final chapter is about divine command and
evolutionary psychology.
This is not the place, however, for a comprehensive treatment of Barth on
divine command. That would be an enormous undertaking, since Barth’s
moral theology is centred on the doctrine of the command of God, and the
scholar attempting a comprehensive treatment is accountable not only to the
Church Dogmatics (about six million words in itself) but also to the intricate
history of Barth’s development. Barth is hard for a philosopher, particularly
for one trained in analytic philosophy. He works not so much by linear
argument (premiss, premiss, conclusion) as by the accumulation of slightly
different statements of the same point in a kind of spiral. The rhetoric is often
deliberately hortatory, rather than merely expository. Moreover, to the extent
Barth is engaging with philosophers who come after Kant, he has in mind
those of the continental tradition, and he uses their language. These differ-
ences in idiom are no doubt the reason that Barth has not, for the most part,
been read by analytic philosophers. This is a shame, because he has a great deal
of importance to teach us. For one thing, he has an astonishing knowledge of
twenty centuries of Christian reflection. My hope in writing this chapter is to
encourage more members of my tribe to try the experiment of reading him
and persevering through the initial discontent.
This chapter takes up three themes in Barth’s treatment of divine com-
mand, and it proceeds not so much by way of exegesis as by way of reflection
on some key ideas. The three themes are the particularity of God’s command
 Barth on Divine Command 143
(Section 5.1), our freedom in response to that command (Section 5.2), and our
access to the command (Section 5.3). All three of these themes have already
been mentioned in Chapter 2, in connection with ‘the four Barthian con-
straints’.1 In the relevant section of Church Dogmatics Barth is considering
respect for life as something for which we are created, and which we are
commanded to affirm. He observes that we can learn a good deal simply from
the fact that we are commanded. Commands, first, are given to responders of a
certain kind, to those who can hear and obey. This means that we are centres of
agency, and Barth emphasizes that we are individual or particular centres. The
commands are primarily to particular people in particular circumstances. He
gives the example of Jesus’s command to two disciples to go into the village
just opposite, where they will straight away find an ass and a colt, and to untie
them and bring them to him.2 This particularity is the topic of the first section
of this chapter. The commands are, secondly, to centres of agency whose
obedience consists in changing how things are, or in resisting change, and who
are therefore in time. They have to persist, in order to be obedient, through the
hearing of the command to the obeying of it. The commands are given,
thirdly, to free beings whose response is ‘up to’ them and is not merely caused
by the giving of the command. This is the topic of second section of the
present chapter. Finally, the commands are given to beings who can receive
words, and so are users of language, though there are complex questions here
about the source for the interpretation of the words in which the commands
come to us. The question of how we understand what we are being command-
ed to do is the topic of the final section of this chapter, and answering it will
require looking at Barth’s relationship to Kant.

5.1. BARTH ON PARTICULARITY

5.1.1. The Specific Individual

In the section of Church Dogmatics (III/4) from which these ‘Barthian con-
straints’ are drawn, Barth starts from respect for life, and then focuses on the
concrete or specific form of life of each individual.
If the respect for life required of man is an affirmation of life, this must consist
further in his resolute will to be himself. This concrete aspect of the divine
command derives, of course, from the fact that it is not a general truth, rule or
precept but always has a historical character and is always a particular challenge
to a specific individual. He who hears it perceives that it is he who is meant and

1 2
These constraints are taken from CD III/4. 4–8 (328–32). Matthew 21: 1–2.
144 God’s Command
whom it concerns—Thou! As this Thou he lives before God and for God, and
therefore with and for his fellow-man. And if he hears God’s command, he is
always summoned thereby to confess this Thou, to take himself seriously as the
Thou as which he lives before and for God, and in this sense to be himself.
Respect for life must also include this will. It must also be each man’s respect for
the individuality in which he may ‘be alive’ before and for God. It must be his
willingness and readiness to live his life as his own, i.e. to live it in the way in
which it is uniquely allotted and loaned to him by God, ‘according to the law of its
beginning’. This formula is not unambiguous. It does not concern the ‘I’ or ‘Self ’
which a man thinks he can find and possess and know in himself, but which he
can really possess and know only in a very limited and partial way. To raise this to
the level of a law, to affirm and take oneself seriously in this sense, to live in this
sense one’s own particular life, is the very thing which is not demanded but
prohibited by the command of God.3
In this passage we have a characteristic Barthian yes and no. He affirms that
there is for each one of us a specific and unique form of life before God to
which we are called.4 On the other hand, he separates his view from that of the
Romantics, and in particular from Schleiermacher, who (according to Barth)
proposed a self that we can find and possess and know in and by ourselves. On
the first view God is sovereign, and on the second view human beings are
sovereign; that is always, for Barth, the decision we are faced with.
We can divide the affirmative side, Barth’s yes, into two points. The first
point is that God is involved in the everyday details in every situation we
encounter, and has requirements for us in those details. God’s command is not
stuck at the level of high generality. Barth’s language is here reminiscent of
Buber in the 1957 Afterword to I and Thou:
One should beware altogether of understanding the conversation with God—the
conversation of which I had to speak in this book and in almost all of my later
books—as something that occurs merely apart from or above the everyday. God’s
address to man penetrates the events in all our lives and all the events in the world
around us, everything biographical and everything historical, and turns it into
instruction [in German, Weisung, or ‘direction’], into demands for you and me.
Event upon event, situation upon situation is enabled and empowered by this
personal language to call upon the human person to endure and decide. Often we
think that there is nothing to be heard as if we had not long ago plugged wax into
our own ears.5
The second point is metaphysically more ambitious. There is for each of us a
particular person God wants us to be. Barth uses the term ‘soul’ for this, but we
could equally use the word ‘life’, the life that we live by the Holy Spirit (the
Greek psuche means both ‘soul’ and ‘life’). One of the several biblical passages

3 4
CD III/4. 59 (385). See CD III/4. 56 (274) and n. 9 of the present chapter.
5
Martin Buber, I and Thou (1923), trans. Walter Kaufmann (New York: Scribner, 1970), 182.
 Barth on Divine Command 145
he collects under this heading is from Mark: ‘For what will it profit them to
gain the whole world, and forfeit their life? Indeed what can they give in
exchange for their life?’ (8: 36–7). Barth points out that there is, in the
immediately preceding verse, a losing of the soul or life that is required in
order to save it. He identifies the soul that is to be lost as the soul over which
we are sovereign, and the soul that is to be gained as the soul that is lived in
Christ, ‘and it is no longer I who live, but it is Christ who lives in me’
(Galatians 2: 20, and see John 14: 19, ‘Because I live, you also will live’).
Barth calls the self that is to be gained ‘the Thou–I’, which we are before
God (coram Deo). This is a ‘specific form of life which as such is the aim of
[our] history’. Barth also calls it ‘character’, and says that it is formed in the
struggle to be obedient situation by situation to God’s command. He warns us
against becoming enamoured of the roles that we see others play. Here is a
theme familiar from Kierkegaard and Heidegger on ‘the one’ or the crowd: ‘A
genuine chorister is better than a false soloist, and an honest pupil than a
supposed master on his own responsibility; and in any case there are no good
or less good roles before God but only the right ones as individually assigned
by Him.’ Each particular assigned form of life, Barth says, is a reflection, but
only a reflection, ‘of the singularity of the self-sufficient life of God’.

5.1.2. Haecceity

Scotus’s notion of a ‘haecceity’ or ‘thisness’ is helpful in understanding this

uniqueness to which we are called. Scotus departs from Aristotle’s position in
Metaphysics VII. 8, that Callias and Socrates are the same in essence, but differ
in matter.6 Scotus holds that what distinguishes two human beings, the ‘con-
tracting difference’, should be something positive that confers a higher degree of
unity and perfection, which explains why the individual cannot be divided into
parts in the way a species has members.7 The contracting difference should not
be just a quantity of matter, which for Aristotelians is less perfect than form. For
Aristotle, at least in this passage, the essence is the same for Socrates and
Callias—namely, ‘humanity’. But for Scotus there is an individual essence, we
might call it ‘Socrateity’, which is unique to Socrates. Socrateity is a perfection of
humanity, in an analogous way to that in which humanity is a perfection that
‘contracts’ animality (the genus). So Socrates has essentially animality and
humanity, but he has in addition his own unique essence.
There is, as mentioned in Chapter 1, a biblical idea that is a precedent for
Scotus. We are told in Revelation 2: 17 that God has for each of us a new name
written on a white stone, which God will give us in the next life but which we

6
Aristotle, Metaphysics VII. 8. 1034a7–8. 7
Scotus, Lectura II, dist. 3.
146 God’s Command
do not yet know.8 We can think of God as already calling us by a name that
expresses what God is calling us to be, even though we do not yet know this
name.9 Returning to Scotus, we can suggest more about the name. Scotus
holds that our happiness, the end towards which we are directed, is to be co-
lovers (condiligentes) entering into the love that is between the members of the
Trinity. The particular end, what Barth calls ‘the aim of our history’, is thus a
particular way of loving God, and the life of the redeemed in heaven is the
union of all these particular ways of loving. Scotus holds that the natural
inclination of the will is towards something particular, since the particular is a
perfection of the universal.10 We can suggest, then, that the name is a name for
this particular way of loving God that is unique to each of us.
This emphasis in Scotus on the particular endeared him to the nineteenth-
century poet Gerard Manley Hopkins.11 Hopkins imagines himself breathing
the same Oxford air Scotus breathed, walking past the same walls that
He haunted who of all men most sways my spirits to peace;
Of realty the rarest-veinèd unraveller . . . 12
Many of Hopkins’s poems express the idea of the ‘inscape’ that is present in
each thing and defines it, even though we see it only imperfectly and in
glimpses.
Each mortal thing does one thing and the same:
Deals out that being indoors each one dwells;
Selves – goes itself; myself it speaks and spells,
Crying Whát I dó is me: for that I came.

8
Chapter 1 mentioned the case of the name ‘Peter’ (meaning ‘rock’) given before Peter
denied Jesus three times, John 1: 42. See John 18: 15–27.
9
Barth discusses the individual’s name at CD III/4. 5 (329) (emphasis added): ‘God knows
who and what he may be. God calls him by his name. It must satisfy him to be always the
particular creature as which God addresses and thus acknowledges him, and to know himself as
such. There can be no doubt, however, that the Word of God, spoken by the divine I to the
human Thou, claims the supremely particular hearing and obedience of this specific man, and
thus reveals the individuality of his being and life.’ Barth’s particularism is also, no doubt,
influenced by Kierkegaard, Works of Love, 230: ‘Within the species each individual is the
essentially different or distinctive. . . . Indeed, if it were not so that one human being, honest,
upright, respectable, God-fearing, can under the same circumstances do the very opposite of
what another human being does who is also honest, upright, respectable, God-fearing, then the
God-relationship would not essentially exist.’ This theme in Barth is also clear in his discussion
of the void in CD III/1, and is repeated at CD III/4. 567: ‘By the fact that God separates this or
that human being as such from all other human beings, he limits him or her . . . to be this
particular human being. . . . It is precisely in this way that one finds oneself fully affirmed and
taken seriously by God.’
10
Scotus, Ord. IV, suppl. dist. 49, q. 9.
11
If I may be allowed a piece of personal history, I lived for a year in Gerard Manley
Hopkins’s rooms at Balliol, and read all his poetry. It was through Hopkins that I came to Scotus.
12
Hopkins, ‘Duns Scotus’s Oxford’, in Gerard Manley Hopkins, 40.
 Barth on Divine Command 147
I say móre: the just man justices;
Keeps gráce: thát keeps all his goings graces;
Acts in God’s eye what in God’s eye he is—
Chríst—for Christ plays in ten thousand places,
Lovely in limbs, and lovely in eyes not his
To the Father through the features of men’s faces.13
This particular essence is then the proper object of our love, even if we do not
understand it with our intellect, and this is true when we are properly loving
ourselves, and properly loving the neighbour, and properly loving God.

5.1.3. The Positions in a Moral Judgement

There is one more piece of philosophical equipment that is helpful before

returning to Barth and trying to understand how God’s command is both
particular (as we have been discussing) and universal. The piece of philosophy
is the distinction between the different positions in a moral judgement made
in Chapter 2.14 The background here is Kant’s first formula of the Categorical
Imperative. Kant says: ‘Act only in accordance with that maxim through
which you can at the same time will that it become a universal law.’15 One
way to understand this is that a universal law has to be stated in a way that
eliminates in principle reference to particular individuals or places or times.16
I may predict that, if I throw a particular stone at a particular window with a
certain velocity, the window will break. But the law that covers that event is
one that has to be stated without reference to this particular stone and this
particular window and this particular time. The law states that any stone of a
certain mass thrown with a certain speed will shatter glass of a certain degree
of fragility. In the same way, the test imposed by the Categorical Imperative
takes the maxim of my action (the prescription of the action together with my
reason for it), and asks whether I can continue to will the maxim when all
reference to myself and other particular individuals is eliminated. H. J. Paton
gave the following example, writing in 1946: ‘If it is the duty of a rational agent
to kill in certain circumstances, then it is the duty of every rational agent to kill
in these circumstances. It is fundamentally immoral to regard killing as a
special privilege of my own from which other men are excluded.’17 One
technical term used to express this idea is that moral judgements have to
be ‘universalizable’. This requires the eliminability in principle of singular

13
Hopkins, ‘As Kingfishers Catch Fire, Dragonflies Draw Flame’, in ibid. 51.
14
I have discussed this in more detail in The Moral Gap, ch. 6.
15
Gl. iv. 421.
16
This is the way RMH interprets the formula, in Moral Thinking, ch. 6. He is following
H. J. Paton, The Categorical Imperative (Tiptree, Essex: Anchor Press, 1946).
17
Ibid. 70.
148 God’s Command
reference to particular people or particular places or particular times, so as to
get what is properly speaking a universal law as the object of will.
This conception of morality is too restrictive, because it rules out as ‘non-
moral’ all sorts of judgements that are naturally seen as moral. We should,
rather, acknowledge that there are different ‘positions’ in a moral judgement
that a term can occupy, and that the requirement of universalizability does not
apply equally to the different term positions. We can distinguish four such
positions: ‘addressee’, ‘agent’, ‘recipient’, and ‘action’. This is not intended to
be an exhaustive list. The addressee is the person to whom the moral pre-
scription is addressed. For example, God tells Joshua before Jericho that seven
priests shall (where this is an imperative) bear before the ark seven trumpets of
rams’ horns.18 Joshua is the addressee, but he is not the agent. God is telling
him that someone else is to do something. Often the addressee and the agent
are the same, but not always. Also, it is often the case that the addressee is not
explicitly mentioned, but assumed from the conversational context. But dis-
tinguishing the addressee position from the agent position enables us to see
already that we can eliminate singular reference in one term position but not
the others. The recipient is the person to whom or for whom the action is to be
done. A judgement may have a term in the recipient position that is not
universalizable, or replaceable with a universal term. A lawyer asks Jesus
which is the greatest commandment in the law, and Jesus replies: ‘“You
shall love the Lord your God with all your heart, and with all your soul, and
with all your mind.” This is the greatest and first commandment. And a
second is like it, “You shall love your neighbour as yourself.”’19 In the greatest
and first commandment the term in the recipient position is not universaliz-
able. The command is not prescribing that the believer should love anyone
who is the same as God in universally specifiable respects. This is especially
clear in the Hebrew imperatives that stand behind this command: ‘I am the
Lord your God, who brought you out of the land of Egypt, out of the house of
slavery; you shall have no other gods before me.’20 For Barth, it is an important
point that the imperative (‘you shall’) comes after the indicative (‘I am the
Lord . . . who brought you’). There is, in other words, a particular relation
between God and God’s people, built up through a particular history. The Ten
Commandments are originally an expression of the obligations of God’s
people within the covenant that God establishes with them.21 Whether they
now apply to all human beings is a large theological question. The present
argument needs a relatively modest point. Even if the term in the agent
position in the commandment to love God with all one’s heart and soul and
mind is universal, and all human beings are commanded to do this (as Barth

18 19 20
Joshua 6: 4. Matthew 22: 36–40. Deuteronomy 5: 6–7.
21
Exodus 34: 28: ‘And he wrote upon the tables the words of the covenant, the ten
commandments.’
 Barth on Divine Command 149
believed), it does not follow that the term in the recipient position (‘The Lord
your God’) is required to be universal or universalizable.
We can make the particularist point by starting from the position of a moral
theorist who takes a position on this question stronger than the one we will
eventually support, and then qualifying it. Lawrence Blum has defended the
view he calls ‘direct altruism’, which denies that in order to be morally good
the maxim of an action must be (or be regarded by the agent as) universaliz-
able.22 He proposes that we can separate Kant’s first formula of the Categorical
Imperative, which was quoted above, from Kant’s idea, expressed in the
formula of humanity (or the formula of the end-in-itself), that we have to
treat humanity, whether in our own person or the person of any other, always
at the same time as an end in itself, and never merely as a means.23 There is a
natural interpretation of this formula that to treat another person as an end in
herself is to make her ends my own ends (to share them) just because they are
hers, as far as these ends are themselves morally permissible. But now we can
see that we can do this, if we love someone, even if we do not commit ourselves
thereby to making these ends our own ends whosoever’s ends they are.
Suppose I have a friend, Elizabeth. If I judge that I morally ought to help her
get rid of the bat in her house (she is terrified of bats), even though I am tired
at the end of a long day, I may be committed to judging that I should help her
in future circumstances of the same sort. But it is not clear that I am commit-
ted to helping anyone in these circumstances. There is a special relation
between the two of us that generates the obligation. Paton perhaps would
say that I am committed to the principle that friends should help each other in
these circumstances. But this general principle does not capture my commit-
ment, which is generated by the particular texture of my history with Elizabeth.
About the general obligations of different types of friendship, as Blum says,
I ‘may simply fail to give thought to whether [my] action has this sort of
universal validity’.24 The obligation is tied to the particular relation between
us, and the claim is that caring for Elizabeth ‘as an end in herself ’ is caring for
her for her own sake, or sharing her ends just because they are hers. If we ask,
‘What does it mean to care for her for her own sake?’, one answer is given by
the thoughts about haecceity described in the previous section of this chapter.
This does not mean that I stop caring for her as a human being. In Scotus,
the haecceity is a perfection of the common nature. But the care for her that
is internal to morality is not restricted to this common nature. In Kant, by
contrast, the answer might be ‘her practical rationality’ (if Kant could suppose

22
Lawrence Blum, Friendship, Altruism and Morality (London: Routledge & Kegan Paul,
1980), and see also his ‘Moral Perception and Particularity’, Ethics, 101/4 (1991), 701–25.
23
Gl. iv. 429. This does not mean here that this formula as Kant interprets it can be detached
from the first formula, but that the formula of humanity can be given a natural interpretation
according to which the formulas diverge.
24
Blum, Friendship, Altruism, and Morality, 89.
150 God’s Command
women have that sort of thing), and in Aristotle the answer might be ‘her nous’
(if Aristotle could suppose women have nous). John Lucas suggested that what
I love, when I love a person for her own sake, is a Leibnizian monad, possessing
an infinite number of features, partially ordered as regards importance and
knowability.25 Here is the Scotist proposal: What I love is her unique essence,
her ‘thisness’. This does not mean that I am not loving her, but it adds that I am
loving her for her haecceity.
In none of this, however, is it necessary to deny, as Blum does, that my
moral commitment to help Elizabeth is an exercise of reason. Blum is a
champion of the moral sentiments or emotions, which he insists have cogni-
tive as well as affective features, and he thinks he has to deny Kant’s tying of
morality to reason. But, while we could stipulate that caring about another
person’s unique essence is emotional rather than rational, or in competition
with rationality, there is no need to make this dichotomy.26
This takes us to the discussion of the last of the four positions in a moral
judgement, the ‘action’ position. The action is what the agent does to or for the
recipient. Here we do have the requirement of a universal term. There is a kind
of action that is being prescribed for the agent to do to or for the recipient, and
it would be inconsistent for the agent to do it for this recipient in one situation
and not in another that is the same in the relevant respects, where these can be
specified in universal terms. If, in helping Elizabeth get rid of a bat in her
house, I make the moral judgement that I ought to help her, then I am surely
committed to helping her the next time the same thing happens and I am at
liberty.27 One way to see this is to see that a moral judgement of an action is at
least an evaluation that the action is good. But a judgement of goodness
requires criteria, and the criteria are given by types, not by individuals.28 The
strawberry is good, in my prior example, because it is red and sweet, and so on.
When I make a moral judgement that I ought to help Elizabeth, I am not,
however, thereby committed to helping anyone else in such a situation, unless
the recipient is specified at least implicitly in universal terms. Sometimes

25
John Lucas, ‘Reasons for Loving and Being Loved’, in Freedom and Grace: Essays by
J. R. Lucas (London: SPCK, 1976), 64–8.
26
See MM vi. 28: ‘In acting I can, without violating the universality of the maxim [of
benevolence], vary the degree greatly in accordance with the different objects of my love (one
of whom concerns me more closely than another).’
27
Could God command me to do something for a person and then command me not to do
the same thing for the same person in relevantly the same circumstances? It is hard to see how
the first command and the second could both be loving. God commanded Abraham first to kill
and then not to kill his son. But the circumstances were different, because Abraham had shown
his obedience. This does not prevent God giving different commands for what I do to different
people in relevantly the same circumstances (since God sees the haecceity and I do not), unless
we tie ‘relevantly the same’ to anything that affects the well-being of the parties.
28
Criteria are epistemic. They tell me what to decide, given what I know. Even though there is
in the haecceity a particular (non-universal) goodness, I do not know it, and I cannot use it as a
criterion.
 Barth on Divine Command 151
moral judgements are universal in the agent and recipient positions (for
example, one ought to help one’s friends), and sometimes they are not. We
can call the ones that are not, ‘particular moral judgements’. The example of
my friend Elizabeth may suggest that we are dealing only with special rela-
tions, which are self-indexed in the sense described in Chapter 3. But, if the
object of my love is a haecceity, it does not have to be self-indexed to me in
order to be particular (see the second error of Kant described in Section 3.2.3).
The reason for insisting that particular moral judgements are still moral even
if they are not universal or universalizable in the recipient position (here we
depart from Kant) is that they can express caring for a person for her own
sake, and this is characteristic of a moral relation. There will usually be a fully
universalizable moral judgement that accompanies the particular moral judge-
ment. For example, ‘One ought to help one’s friends’ accompanies ‘I ought to
help Elizabeth’. But ‘I ought to help Elizabeth’ is, to use Scotist language, a
perfection of ‘One ought to help one’s friends.’

5.1.4. Barth on Universality

Now we can return to Barth and try to explain the sense in which divine
commands are universal as well as particular. He says:
Even the claim which is addressed to me is not for me alone, but of universal
validity. And I have to understand the universally valid claim as valid for me too
and applying to me. If I refuse to do this, then from the detached standpoint of
the individual and peculiar characteristics of my situation, my special case, I can
protect myself against the crisis which the existence of God’s command signifies
for me and brings down upon me. And in the last analysis who cannot claim in
every respect to be in a highly singular situation? . . . That the universally valid
command of God applies to me and affects me in a very definite way cannot be
taken to imply that I can treat it as conditioned by the peculiar factors of my
personal situation; that I can secure and fortify myself against its universal
validity as it certainly applies to me too. At this point, again, we have something
to learn from Kant—from his definition of the ethical as that which is adapted to
be ‘the principle for a universal law’.29
Is there any way to make this consistent with the themes in Barth we have
already been discussing? There is, but it will require making some distinctions

29
CD II/2. 655–6. Barth’s term ‘crisis’ is taken from the Greek krisis, which also means
‘judgement’. Barth says here that we can learn from Kant, but his point is not exactly the same as
Kant’s. The universality that Barth thinks we must acknowledge is not merely what is binding on
humans just as humans, but what is binding on the elect, to those chosen in Christ. To think of
the binding as on ‘me’ rather than on ‘us’ is in this context to try to evade the responsibility of
being chosen.
152 God’s Command
that Barth does not make as well as recalling some distinctions that he
does make.
First, there is a distinction he makes between instruction and reflection.30
Barth sometimes talks as though moral thinking is only particular, and as
though, since God’s judgement makes something right or wrong, our task is
not to work out what is right, but simply to repeat God’s judgement in our
own. This combination of views has seemed to many to be wildly implausible.
But the distinction between instruction and reflection helps us see that there is
a difference between the context in which we receive the divine command and
repeat it in our own will (this is ‘reflection’), and the context in which we
prepare ourselves to receive it by thinking through what we know about God
and ourselves (this is ‘instruction’). The Ten Commandments, for example,
are general to the extent that they give instruction for a whole range of
situations, and not just the situation of one agent at one time and place.
Barth wants us to see that thinking through these general prescriptions and
proscriptions is propaedeutic; it prepares us by forming us in advance to
receive the divine command for a particular situation. It does not, however,
give us exceptionless general principles that we then ourselves work out how
to apply. At least in the second edition of his commentary on Romans Barth
agrees with Scotus, ‘God may be honoured in behaviour that contradicts the
commandments of the second table’, and he does not formally retract this in
subsequent work.31 But nonetheless we learn about God and about ourselves
from these commandments, just as we learn from the proving of Abraham and
his son on Mount Moriah. We are now in a different position from Abraham
just because we now know that God does not want us to demonstrate our
devotion by sacrificing our children. Christians learn about the character of
God and themselves by immersing themselves in the stories of Jesus’s life,
death, and resurrection. Barth makes it clear that he wants us to do this kind of
preparatory thinking, because he himself writes hundreds of pages in which he
does it. He thinks through questions about abortion and suicide and armed
resistance to oppression, and almost all of this discussion is general, in the
sense that it applies to a whole range of situations and not just the command to

30
See McKenny, The Analogy of Grace, esp. ch. 6. See also Nigel Biggar, The Hastening that
Waits: Karl Barth’s Ethics (Oxford: Clarendon Press, 1993), 7–45.
31
Karl Barth, The Epistle to the Romans, trans. Edwyn C. Hoskyns (Oxford: Oxford Univer-
sity Press, 1968), 451. Barth does not formally take this back. However, in CD III/4, especially in
‘Respect for Life’, he does modify it, by making the distinction between boundary case (Grenzfall)
and exception (Ausnahme), and then by denying that the boundary case is an exception.
Unfortunately this is disguised in the English translation, which translates Grenzfall as ‘excep-
tion’. Barth goes to great length to interpret what look like exceptions (in the cases of abortion,
tyrannicide, self-defence, and war) as actually strange or paradoxical instances of the command
to protect life. I owe this point to Gerald McKenny.
 Barth on Divine Command 153
a particular person at a particular time and place. His writing is here
propaedeutic.
This distinction between the context of instruction and the context of
reflection does not, however, yet solve the problem. We cannot say simply
that divine command in the context of instruction is universal and in the
context of reflection is particular. In the long passage just quoted, in which
Barth talks about universal validity, he is talking about the context of reflec-
tion. The same is true when he says that ‘the goodness of the divine command
is something universal’.32 What does he mean? Here it is helpful to make use
of a distinction Barth does not make, between the good and the obligatory,
where ‘right’ belongs in the family of obligation. In the sentence just quoted
Barth talks about goodness, but often he talks about ‘goodness and rightness’
together.33 The first chapter of this book discussed the relation between the
good and the obligatory, distinguishing different kinds of priority relation
between them. It suggested that everything that is obligatory is good, but not
vice versa, and that the obligatory, but not the good, is grounded in divine
command. The good is related to God by resemblance or by drawing us to God
or in a variety of other ways. Everything God commands, we can say on this
view, is good, but not everything good is something God commands. In using
this distinction between the good and the obligatory, we go beyond Barth. But
we now have a tool for understanding what he does say about the universality
of the divine command. Whenever God issues a command to a particular
person in a particular situation, we can say, there is something good that is
commanded. God is giving us a route, we can say, towards the particular form
of life, or the particular way of loving God, that is our destination or calling
named on the white stone. Section 4.1.2 suggested that we have a presumption,
from our nature as called, against taking prescriptions to kill the innocent or to
lie as God’s commands. It is true that God can choose which good things to
command. But in interpreting the command, and deciding what to take as a

32
CD II/2. 711.
33
e.g. CD II/2. 634: ‘It is a question of the rightness and goodness of our choice between the
various possibilities of our existence . . . in the light of the command and decision of God.’ Barth
does have a distinction that is functionally equivalent to the distinction I am drawing between
‘good’ and ‘obligatory’. In his 1929 lectures on The Holy Spirit and the Christian Life: The
Theological Basis of Ethics, trans. R. Birch Hoyle (Louisville, KY: Westminster/John Knox, 1993),
7–8, he distinguishes creation order and God’s command to the individual. ‘I know, of course,
that I am this person, a man placed in this or the other set of circumstances with their internal
and external requirements, but I do not know at this moment the specific calling assigned to me
by God. I am well aware of the necessity for “order” and also that there are definite “orders”,
arrangements that are in force and regulate my life at this moment. These “orders” were laid
down by God at the creation, and even now they are directions for my living. But what, for
example, work, marriage, family, and so forth signify just now, in my particular case, as God’s
orders, I do not know. . . . For this reason it has to be told me through the second miracle of
God’s love, that is, God’s revelation.’ By the time of CD Barth is talking not about ‘orders’ but
about ‘spheres’ or ‘domains’. But the distinction remains.
154 God’s Command
command, we are within the narrative about God’s choosing that is given in
the tradition in which we stand. We cannot change the fact that we have been
given this story, though we can affect how we pass it on and how we interpret
it. The Abrahamic faiths have their own ways (sometimes different ways
within the same faith) to give authority to some interpretations over others.
It would be the task of a much larger project than this book to evaluate these
ways even within one tradition.
One feature of moral prescription is that we often say what we think a
person ought to do by simply mentioning a value that we think should prevail
in the particular situation. Proverbs work like this. ‘Too many cooks spoil the
broth’, we might say in some situation. We might also believe that many hands
make light work. But, in the present situation, the most salient value is the
agent’s independence from interference by others. When we say ‘Too many
cooks spoil the broth’, we are saying that in this situation that value is trumps.
God’s command can operate in this way also. God can make some value stand
out for us. There may well be competing values at stake. There usually are,
when we decide what to do. But this is the value that God wants us to realize
here and now. For example, in one story about Jesus’s disciples, they were in
an uproar because two among them, James and John, had asked for the
privilege of sitting at his right and left when Jesus came into his glory.34
Jesus responds by painting a picture of Gentile lordship, and then saying
about the Son of Man, by contrast, that he came not to be served but to
serve, and to give his life as a ransom for many. There is a command that
follows from these two pictures. James and John are being rebuked, but so are
the rest of the disciples for fighting for the same kind of position that James
and John were forthright enough to ask for. This is not the way they are to live.
There is a value here that Jesus makes stand out for them, by giving these
negative and positive examples, and this is the value of humility that they are
in the present situation failing to realize. This value is something universal. We
might describe it as ‘thinking others better than oneself ’.35 But it is up to God’s
command whether or not this is the value that is the most important value to
realize in some situation. Sometimes some other value, perhaps self-respect,
may be more important, and a particular young person needs to be told not to
let anyone despise his youth.36
This sort of picture of the universality of divine command fits what Barth
says about Kant. The goodness of thinking of others as better than oneself is a
universal value, applying to everyone. Jesus presents this value to the disciples

34
Mark 10: 41.
35
Philippians 2: 3. Paul describes this as humility, in Greek tapeinophrosune, and his word for
‘others’ is allelous, more literally ‘each other’. Section 4.2 mentioned the difference of this virtue
from any of the virtues on Aristotle’s list.
36
1 Timothy 4: 12.
 Barth on Divine Command 155
in the action of washing their feet, and then tells them to do likewise.37
I cannot escape from this claim on me by urging that my own case is special.
This is Barth’s emphasis in the long passage quoted from him at the beginning
of this section about the universal validity of the divine command. It is also
Kant’s emphasis in his discussion of why the Categorical Imperative is
categorical.
If we now attend to ourselves whenever we transgress a duty, we find that we in
fact do not will that our maxim should become a universal law—since this is
impossible for us—but rather that its opposite should remain a law universally:
we only take the liberty of making an exception to it for ourselves (or even just for
this once) to the advantage of our inclination.38
This is no doubt the passage of Kant that Barth had in mind in saying that ‘we
have something to learn from Kant—from his definition of the ethical as that
which is adapted to be “the principle for a universal law”’.39 The structure of
Barth’s ethics is distinctively Kantian in this respect; moral obligation is
established independently of what we desire.40 This does not mean, for either
Kant or Barth, that desire is finally opposed to obligation. But both thinkers
deny that we can deduce what morality requires from a consideration of what
we desire, even what we would desire if we were fully informed about the non-
evaluative facts.
We can now link this point about universal validity with the distinction
drawn in the previous subsection of this chapter between different positions in
a moral judgement. The action position is occupied by the term that describes
what the agent is prescribed to do to or for the recipient, and there is a
requirement of universalizability for the term in this position in a moral
judgement, though not for the terms in the other three positions. We can
now say more. What is required morally is a good action, where universal
standards or criteria specify the goodness, and this goodness is in principle
knowable in advance within the context of instruction (as opposed to the
context of reflection).41 It does not follow from this, however, that we know in
the context of instruction what our obligation is, because it is God who tells us
which good kind of thing we are now to realize, to which particular recipients.
When God does command this, however, we are no longer at liberty to plead

37 38
John 13: 4–17. Gl. iv. 424.
39
CD II/2. 656. In the second edition of Romans, 493, Barth says that Kierkegaard must be
corrected by Kant, because we cannot escape the criterion of universal validity. David Clough
discusses this passage well in Ethics in Crisis: Interpreting Barth’s Ethics (Aldershot: Ashgate,
2005), 38–41.
40
CD II/2. 650.
41
I take the phrase ‘standards or criteria’ from McKenny, The Analogy of Grace, 268–9. But
I am using ‘criteria’ to mean the descriptions on the basis of which a person makes an evaluation,
as when she says a strawberry is good because it is red, sweet and ripe. See Section 4.3.1; also my
God and Morality, 212–17, and God’s Call, ch. 1.
156 God’s Command
special circumstances that give us an exception. Putting the matter this way
allows us to see how Barth could say both that the command is ‘a particular
challenge to a specific individual’ and that it has universal validity.42
Two other Barthian themes fit this analysis. The first is Barth’s claim that
the divine command always comes to us ‘in the garment’ of a claim from
someone or something else we encounter in our natural, social, and cultural
existence.43 One thing Barth has in mind is what Jesus says in the parable of
the sheep and the goats: ‘Just as you did it to one of the least of these my
brothers and sisters, you did it to me.’44 God’s command comes to us, as
Kierkegaard puts it in Works of Love, with a forwarding address.45 In terms of
Jesus’s parable, God presents me with a good action type such as ‘feeding the
hungry’, ‘clothing the naked’, ‘visiting those in prison’, in connection with
some particular person with a claim to be fed, clothed, or visited now by me.46
In other words, we always encounter the divine command in connection with
some human being or something else with a claim on us. But, if God
determines obligation in this way, we might wonder what role is left for our
own moral deliberation? Do we not become entirely passive, merely fulfilling a
command that comes to us in full detail from the outside? That is the topic of
the rest of this chapter.
The second theme from Barth that fits the analysis here is that the universal
validity of the divine command means that we have to be willing to make it
public. This is like Kant’s publicity condition, which requires that the maxim
of my action be communicable to, and then endorsable by, all members of the
kingdom of ends. The Categorical Imperative itself is, for Kant, a public
principle, and any attempt to make oneself exempt from this public morality
is morally evil. There is no private morality, for Kant. For Barth, if I am going
to obey a prescription as a divine command, I am claiming to follow the same
commander whose commands establish the covenant obligations for all
human beings. ‘I may and must hear his command,’ Barth says, ‘but his
command applies to us all. . . . Even in the necessary testing of my conduct

42
CD III/4. 59.
43
‘The claim of God’s command always wears the garment of another claim . . . an object with
its question, the compulsion of a necessity of thought, one of those hypotheses or conventions, a
higher or more primitive necessity of life, a necessity which in itself seems to be that of a very
human wish or very human cleverness, a summons coming from this or that quarter, a call that
man directs to himself—all these can actually be the command of God veiled in this form, and
therefore genuinely participate in the corresponding authority and dignity’ (CD II/2. 584–5,
trans. Clough, in Ethics in Crisis, 127). See also McKenny, The Analogy of Grace, 8.
44 45
Matthew 25: 40. Kierkegaard, Works of Love, 161.
46
In Chapter 2, I discussed the notion I called ‘providential proximity’, that God puts us next
to the people God wants us to help. I do not mean to imply, however, that we have to know who
the recipient is. God can put me next to the residents of some village in Zambia, when my son
goes there for a year to work for an aid organization. Even though the residents of the next village
have the same needs, and even though I do not know who the residents of my son’s village are,
this is now, so to speak, my village.
 Barth on Divine Command 157
I cannot overlook or forget the fact that I am never alone and never will be.’47
Barth’s view is that we receive divine commands as members of a body, and
the context of instruction has given us common ground in which to test for
each other whether we genuinely have heard a divine command or not. We
have the duty to testify to the choices and decisions we have made, and to be
open to the refutation of our claim. We are usually to hear God’s command
through our relations of mutual accountability with other people.

5 . 2. BA R TH ’S ACCOUNT OF HUMAN FREEDOM

5.2.1. Three Pictures of Freedom

The second part of this chapter is about the human response to divine
command, and the extent to which we can say that this response is free.
Barth says: ‘As election is ultimately the determination of man, the question
arises as to the human self-determination which corresponds to this deter-
mination.’48 If God determines us to be what we are, how can we determine
ourselves? Barth makes this question acute, and indeed sometimes he seems
more to relish the difficulty of the question than painstakingly to work out his
solution to it. Barth likes paradox more than most analytic philosophers do.
Nonetheless he does propose a solution.49 His central term is our ‘responsive-
ness’ (Verantwortlichkeit), and this lies behind the claim at the end of
Chapter 1, drawn directly from David Kelsey but indirectly from Barth, that
our dignity lies in our reception of the divine call. Barth says that our
responsiveness is ‘the most precise description of the human situation in the
face of the sovereign decision of God’.50
The first thing to observe is that Barth denies that there is a dichotomy
between divine freedom and ours, or between divine sovereignty and human
autonomy. Indeed, he suggests that the postulation of such a dichotomy was
the work of the serpent in the Garden of Eden. For Barth there is not in this
context a zero-sum game, not a competition, but a complementarity. Kathryn
Tanner makes the point well. There is, she says,
a direct rather than an inverse proportion between what the creature has, on the
one hand, and the extent and influence of God’s agency, on the other. . . . Talk of
the creature’s power and efficacy is compatible with talk about God’s universal

47 48
CD II/2. 655. CD II/2. 510.
49
See John Webster, Barth’s Moral Theology: Human Action in Barth’s Thought (Grand
Rapids, MI: Eerdmans, 1998), esp. chs 5 and 6. See also his Barth’s Ethics of Reconciliation
(Cambridge: Cambridge University Press, 1995).
50
CD II/2. 641.
158 God’s Command
and immediate agency if the theologian follows a rule according to which divinity
is said to exercise its power in founding rather than suppressing created being,
and created being is said to maintain and fulfil itself, not independently of such
agency, but in essential dependence upon it.51
We can see this point in Barth’s treatment of the first petition of the Lord’s
Prayer, ‘Hallowed be your name’.52 God’s name will finally be hallowed, and
that is God’s work in the second coming of Christ, but, ‘if this prayer is part of
man’s invocation of God, it does, of course, entail necessarily that there should
be a corresponding willing, acting and doing on man’s part’.53 Barth is saying
that God determines what we are, by electing us in Christ, but we have to
acknowledge this, or determine ourselves in correspondence to this. To put
this paradoxically, this is to say that we become what we already are. But it is
not necessary to put this paradoxically. We can proceed by giving three
pictures of the relation between God’s determination of us and our self-
determination. The first is a humble analogy of playing a piano duet with a
master, and it comes not from Barth (who distrusts such analogies), but from
the present author’s own experience. The second is a picture of the relation
between our loving and God’s loving that derives from Kierkegaard’s descrip-
tion in Works of Love of the quiet lake and the hidden spring.54 The third is
Barth’s own picture of how God’s life and ours relate when we pray to God.
Suppose I am not a very good pianist, but I play well enough so that I can get
through, let us say, one of Mozart’s sonatas for four hands. Someone comes for
dinner who plays the piano professionally, and is in fact a master of her
instrument. She proposes that we play a duet together. When we get started,
after my initial embarrassment, I find that I am playing far better than I normally
do. In fact, I am playing in a power not my own, swept along by the power of her
playing. In the slow passages I take the delicacy of my phrasing from her, and
I find my fingers running through the fast passages as though possessed. It is,
however, still me playing. One sign of this is that I could, no doubt, stop the flow
at any minute by simply relapsing back into my lesser and more timid self.
Another sign is that, even at my best, my playing is still derivative from hers. But
nonetheless what she wants, in making the proposal that we play, is something
she cannot do just by herself; she cannot play all four hands.
There is a combination here that is suggestive of how things are in our
relation to God. She does it and I do it. But this is not mere collaboration,
because I am playing in her power (as well as my own), and the same is not

51
Kathryn Tanner, God and Creation in Christian Theology (Oxford: Blackwell, 1988), 85.
52
Matthew 6: 9. Barth’s treatment comes in the lectures posthumously published as The
Christian Life: Church Dogmatics IV/4. Lecture Fragments, trans. Geoffrey W. Bromiley (Edin-
burgh: T & T Clark, 1981), and gives us an indication of what would have been the content of the
unfinished part of Church Dogmatics on the ethics of reconciliation.
53 54
Barth, The Christian Life, 156. Kierkegaard, Works of Love, 9–10.
 Barth on Divine Command 159
true symmetrically the other way round. My part is to correspond to her. But
presupposed in this idea of ‘my part’ is a kind of responsibility, and presup-
posed in this notion of responsibility is the ability to do otherwise. Barth
distinguishes between freedom properly so called and what he calls ‘natural
freedom’. Gerald McKenny puts the point this way:
For Barth, freedom in its most proper sense is the confirmation in our decision of
God’s decision for us. It is the freedom of the human covenant partner of God. By
contrast, natural freedom is the neutral capacity to choose one action or another.
Barth explicitly denies that this capacity to choose is freedom in the proper sense.
Yet he also argues that the freedom to confirm God’s decisions in our own
decision presupposes this ‘natural freedom’.55
There is a similar combination in Scotus, though the interpretation of Scotus
here is disputed.56 There is an indeterminacy of defect, he says, which matter
has in the absence of form (as in a mere heap of sand). But, suppose we
imagined fire, for example, acquiring the ability not only to heat things but to
cool things. Then it would have what he calls the indeterminacy of super-
abundance. It would have enough form, so to speak, for opposite effects (and
not merely in the way the sun has the capacity both to melt ice and dry mud).
Humans have this kind of indeterminacy of superabundance, because we have
the capacity to rank (as described in Chapter 3) our basic motivations in two
opposite ways. We can give priority either to the affection for advantage (the
self-indexed drive for our own happiness and perfection) or to the affection for
justice (the drive towards what is good in itself, independently of any relation
to us). This kind of indeterminacy is what puts us in the domain of freedom
rather than mere nature. On a Kantian view we are no longer in a neutral
position, because of the bondage of the will, and grace is necessary in order to
restore the possibility of our ranking duty over happiness. But there is also
something for us to do in choosing the right, and the grace here that makes
this choice possible does not itself determine the choice. The grace is not
irresistible. But this account of neutral freedom allows a different kind of
freedom, which Scotus calls ‘firmness’, and the change from mere indetermin-
acy to firmness is a change in the same direction as the change from the
indeterminacy of defect to the indeterminacy of superabundance. Firmness is
where virtue has become ‘second nature’ for a person. In such a case, we have
two efficient causes of an action (‘efficient co-causality’), of which the free act
of self-determination is one and the habit of the virtue is a second. Firmness of
this kind is, for Scotus, a perfection of freedom.57

55
McKenny, The Analogy of Grace, 12.
56
For a fuller treatment, see my God and Morality, ch. 2.
57
For an illuminating treatment, see Bonnie Kent, ‘Scotus on Virtues’, in Thomas Williams
(ed.), The Cambridge Companion to Duns Scotus (Cambridge: Cambridge University Press,
2003), 352–76.
160 God’s Command
There is a similar pattern in Kant. He has both the conception of ‘freedom
of choice’ (Willkühr) and the conception of the rational will (Wille). The
rational will is free, but its freedom is of a different kind from mere freedom
of choice, which is a neutral switching device that enables us to give priority
either to the evil maxim (which prefers happiness to duty) or to the good
maxim (which prefers duty to happiness). The rational will chooses only in
accordance with the moral law. As with Scotus, Kant is not here saying that we
should prefer not to have the drive towards our own happiness. The question
is, however, how we rank it. Barth is thus following a familiar pattern of
thought, in which there are two freedoms, one ordered towards the other. This
is one way among others to understand Paul’s teaching: ‘For freedom Christ
has set us free.’58 Christ gives us one kind of freedom in order that we might be
able to come to a second, higher kind. The first is a necessary condition, a
prerequisite, but not yet a sufficient condition for the second.
The second picture of the relation between our self-determination and
God’s determination of us is the picture of love given us by Kierkegaard in
Works of Love, in his description of the quiet lake and the hidden spring.
Kierkegaard says:
Love’s hidden life is in the innermost being, unfathomable, and then in turn is in
an unfathomable connectedness with all existence. Just as the quiet lake originates
deep down in hidden springs no eye has seen, so also does a person’s love
originate even more deeply in God’s love. If there were no gushing spring at
the bottom, if God were not love, then there would be neither the little lake nor a
human being’s love. Just as the quiet lake originates darkly in the deep spring, so a
human being’s love originates mysteriously in God’s love. Just as the quiet lake
invites you to contemplate it but by the reflected image of darkness prevents you
from seeing through it, so also the mysterious origin of love in God’s love
prevents you from seeing its ground. When you think that you see it, you are
deceived by a reflected image, as if that which only hides the deeper ground were
the ground. . . . In the same way the life of love is hidden, but its hidden life is in
itself motion and has eternity within itself. Just as the quiet lake, however calm its
surface, is actually flowing water, since there is the gushing spring at the bottom—
so also love, however quiet it is in its concealment, is flowing nevertheless.59
Kierkegaard has two preoccupations in this passage, one ontological and the
other epistemological. The ontological point is the dependence of the lake and
its motion upon the spring, though there is also a disanalogy here because the
quiet lake will dry up if and when the spring ever stops; but the life of love has
an eternal spring and will never dry up. The epistemological point is that we
can see the lake, but we cannot see the spring. One way to take the ontological

58
Galatians 5: 1. Paul may well mean, however, that it is for its own sake that Christ has given
us freedom, meaning just one kind of freedom, not two.
59
Kierkegaard, Works of Love, 9–10.
 Barth on Divine Command 161
point, a way that Kierkegaard does not intend, is that we become part of God
when we love.60 The picture would then be that the water represents love,
and it flows from God, who is love, and when we love we enter into that
eternal stream. This is probably not Kierkegaard’s intention, because he is so
frequently concerned to deny that we are part of God, who dwells within the
universe and within us. Judge William, for example, who is a picture of the
ethical life and not the life of faith, goes deeply inside himself and finds God,
and Kierkegaard intends this picture as a critique.61
If this is not Kierkegaard’s intention with the picture of the quiet lake and
the hidden spring, what is? He intends that we are the lake. We are indeed
created from outside, just as a lake’s water comes from a hidden source; but the
point of the picture is that the motion of the water comes from the spring. The
lake may seem still, but it actually has movement and is itself flowing towards
the outside. The biblical picture of living water is relevant here (where ‘living’
means flowing, as opposed to still or stagnant). Jesus says: ‘Those who drink of
the water that I will give them will never be thirsty. The water that I will give will
become in them a spring of water gushing up to eternal life.’62 Kierkegaard’s
intention is that there is a loving that God does, and it communicates life or
movement to us, so that we love in a way that corresponds to it.
This way that God loves is a kind of self-limitation, by relation to what is not
God. Barth repeats this idea when he says: ‘The biblical witness to God sees his
transcendence of all that is distinct from himself, not only in the distinction as
such, but furthermore and supremely in the fact that without sacrificing his
distinction and freedom, but in the exercise of them, he enters into and
faithfully maintains communion with this reality other than himself.’63
God’s being-in-self, or God’s ‘aseity’, is God’s being-for-the-other already
within the Trinity. But in creation and then in redemption God freely becomes
vulnerable to us.64 Barth and Kierkegaard are here maintaining a distinction
between God’s kind of love (in Greek, agape) and merely human loves
(various forms of eros). Kierkegaard in Works of Love is more single-minded
in his suspicion of eros, whereas Barth sometimes expresses respect for the
joyful affirmation of life in the ancient Greeks and in Nietzsche.65 But both

60
See CD IV/2. 309.
61
Sren Kierkegaard, Either/Or II, ed. and trans. Howard and Edna Hong (Princeton:
Princeton University Press, 1987), e.g. 209–10, in which a young man despairs and finds in
that despair his eternal validity and the eternal human being. But the point is pervasive in the
Judge’s two huge letters, and requires the context of the whole, and indeed of Kierkegaard’s
whole pseudonymous authorship.
62 63
John 4: 13. CD II/1. 303.
64
Philippians 2: 6 (author’s translation): Christ ‘being in form God, did not consider equality
with God something to be grasped, but emptied himself, taking the form of a servant.’ This is the
idea of God appealed to in Section 3.3.4, in the revision of Porter–Aquinas’s version of
eudaemonism.
65
McKenny, The Analogy of Grace, 157–8.
162 God’s Command
Kierkegaard and Barth maintain the distinction between the two kinds of love,
and it is agape that is the motion towards the outside communicated from the
hidden spring to the quiet lake. This love is, Kierkegaard says in another image
in the same chapter, ‘eternity’s bond’ with temporality.66 The fact of being
loved in this way has the force of a claim on us that we should love each other
in this way. This, too, is Barthian. Law or command is, he says, the form of
gospel or grace, because what we have been given enables us freely to be what
the command informs us we ought to be.67
This is the ontological point. But for Kierkegaard there is also an emphasis
on the epistemological point. The picture of the hidden spring comes in the
first chapter of Works of Love entitled ‘Love’s Hidden Life and its Recogniz-
ability by its Fruits’. In the long quoted passage, he emphasizes that we cannot
see the spring, the source. When we think we see it, we are deceived by the
reflected image, ‘as if that which only hides the deeper ground were the
ground’. Curiosity leads us to want to dig down and force our way in to find
the source, but Kierkegaard’s idea is that this will result only in making it
harder for us to see.
Certainly, since God is spirit, we should not expect our eyes to see God’s
working, though we may see the fruits of it. But Kierkegaard’s point goes
beyond this, and goes too far. He is deeply suspicious of any claim to a distinct
phenomenology or experience of receiving divine command. Christian love is,
he says, sheer action, the fulfilling of the law; it is not ‘that hidden, private,
mysterious feeling behind the lattice of the inexplicable that the poet wants to
lure to the window’.68 There is a similar critique in Barth, who says of
eighteenth-century Pietism that it
is inconceivable without the undercurrent of mysticism, enthusiasm, ecstasy,
inspiration and occasionally of theosophy and occultism of every kind. . . . From
this perspective, the phenomenon of the ‘inner voice’ as a substitute for previous
authority . . . is to be taken as only one element among many. The substitute for
authority has its basis in the much more important substitute for the sacrament:
because there can be felt, undergone, experienced an inner baptism of the spirit,
an inner spiritual eucharist, an inner perception of the word, because there can be
a direct personal converse of the soul with God . . . there is that prophecy, that ‘It
seems to me’ with its immediate authority as a final resort.69
Standing behind both Kierkegaard and Barth here is Kant, who has the same
misgivings about the experience of the presence and command of God. He
warns: ‘If the delusion of this supposed favorite of heaven reaches heights of

66
Kierkegaard, Works of Love, 6. We will see a similar theme in Rosenzweig in Section 7.3.2.
67
CD II/2. 587, 593, 602. But the claim is pervasive.
68
Kierkegaard, Works of Love, 99.
69
Karl Barth, Protestant Theology in the Nineteenth Century: Its Background and History,
trans. Brian Cozens and John Bowden (Grand Rapids, MI: Eerdmans, 2002), 107.
 Barth on Divine Command 163
enthusiasm, to the point of imagining that he feels the special effects of faith
within him (or even has the impertinence of trusting in a supposed hidden
familiarity [in German Umgang, or “intercourse”] with God), virtue finally
becomes loathsome to him and an object of contempt.’70 Kant’s epistemology
gives him reason for this misgiving. ‘This feeling of the immediate presence of
the highest being, and the distinguishing of it from any other, even from the
moral feeling, would constitute the receptivity of an intuition for which there
is no sense [faculty] in human nature.’71 Kant is putting together two issues
here. The first is whether we have any receptivity for a divine communication
and the second is whether we can always tell whether we are receiving it. The
final section of the present chapter argues that we do have the receptivity, but
we are not infallible about our reception, even though we do have some
markers for it internal to the tradition.
The third picture of the relation between our self-determination and God’s
determination of us comes from Barth’s account of prayer. This is not just a
picture, if that means a model or an example. Barth thinks that prayer is where
the reception of God’s command takes place. Prayer is therefore not just one
place among others where we can see the characteristic freedom of human
response to God’s command; it is the central place. Barth focuses on the kind
of prayer he calls ‘invocation’. It is an action ‘for which man finds himself
empowered only by the free grace of God’, and, on the other hand, it is ‘an
authentically and specifically human action, willed and undertaken in a free
human resolve. . . . No less serious in his place than God in his, man must be
present and at work in it according to the measure of his human capacity.’72 In
prayer there are two agents, two subjects, and not just one.
The God who is known as ‘our Father’ in Jesus Christ is not this supreme being
who is self-enclosed, who cannot be codetermined from outside, who is con-
demned to work alone. He is a God who in overflowing grace has chosen and is
free to have authentic and not just apparent dealings, intercourse, and exchange
with his children. . . . He will work only in connection with their work.73
One scriptural text that is reflected in this passage is Paul’s teaching in Romans
that the Holy Spirit helps us when we do not know how to pray. ‘Likewise the
Spirit helps us in our weakness; for we do not know how to pray as we ought,
but that very Spirit intercedes with sighs too deep for words. And God, who
searches the heart, knows what is the mind of the Spirit, because the Spirit
intercedes for the saints, according to the will of God.’74 God, in the person of
the Holy Spirit, prays with us when we pray. But this does not mean that God

70
Rel. vi. 201. The context makes it clear that Kant is thinking of people who, just from the
unique quality of their experience, think they know what God is telling them to do.
71 72
Rel. vi. 175. Barth, The Christian Life, 42.
73 74
Ibid. 103. Romans 8: 26–7.
164 God’s Command
is simply praying to Godself. We have the responsibility of corresponding to
the divine will. There is here a kind of divine self-limitation. God sometimes
works in connection with our work. (Barth’s ‘only in connection with their
work’ is too strong, since not all God’s work concerns us.) We have to repeat in
our will God’s will for our willing. This is not necessarily to repeat God’s
overall will in our willing, for God has supervision of the universe and we
do not.75 But the repetition, or, in Barth’s term, the ‘correspondence’, is to
God’s will for our will, expressed in the divine command. God’s self-limitation
is sometimes to include our agency in the production of the effects of the
divine will.

5.2.2. The Canaanite Woman

Perhaps it is worth retelling one biblical story as an example of the relation

between our self-determination and God’s determination of us. This is the
kind of thing Barth does all the time, and does very well. The story is that of
Jesus and the Canaanite woman.76 At the beginning of the chapter, Jesus is
debating with the Pharisees about the purity laws that distinguish between
Jews and Gentiles. He quotes from Isaiah: ‘These people [the people of Israel]
draw near to me with their mouths and honour me with their lips while their
hearts are far from me.’77 The passage goes on to say that God is going to
accomplish a marvellous work: ‘Shall not Lebanon in a very little while become
a fruitful field? . . . On that day the deaf shall hear the words of a scroll.’ The
passage starts with the unfaithfulness of Israel and moves to God’s generosity
to those outside Israel immediately to the north. In the story we are now
considering this is where Jesus immediately goes, to the region of Tyre and
Sidon. The passage from Isaiah is no doubt in his mind. A woman comes out
to meet him, crying, ‘Lord, son of David, have mercy on me! My daughter is
suffering horribly from demon-possession.’ She knows about him that he is
Jewish (‘son of David’) and that he can do miracles of healing. Behind her
address, there is another relevant background, the reported history of the
treatment of the Canaanites by the people of Israel.78 The important point is
that Christians have to read this history in the light of the whole narrative of
Scripture, and in particular in the light of the life of Jesus, who told us to love
our enemies.79 This is one focus of the present story. It shows that God is
including this Canaanite within the covenant (not obliterating her), and

75
Scotus discusses this at Ord. I, dist. 48. He is commenting on Peter Lombard, who raises the
example of a good son who is anxious that his father should live, when it is God’s good will that
he should die.
76 77
Matthew 15: 21–8. Isaiah 29: 13–18.
78
I say ‘reported’, because I am not trying to pass judgement on the historicity of Joshua.
79
Matthew 5: 44.
 Barth on Divine Command 165
through her as a beginning, including the Gentiles. But Jesus’s immediate
response to her is silence. Matthew emphasizes this: ‘Jesus did not answer a
word’, not even to acknowledge her presence with a ‘Shalom’. Why? Here we
have to speculate, but there is a pattern we can see in the Gospels, in which
Jesus takes the people who come to him for help through something especially
hard for each one. Section 4.1 described the interlinked stories of Jairus, the
ruler of the synagogue, and the woman with the bleeding.80 It is as though
Jesus sees into their souls, and sees the obstacle that has to be overcome,
almost opposite for each of them. Perhaps something similar is happening in
our present story.
Jesus does not answer a word, but the disciples fill the gap. ‘Send her away,’
they say, ‘for she keeps crying out after us’. Note that this is not quite true. She
was calling out not after them, but after Jesus. They take Jesus as belonging to
them, and as needing to be protected from aliens and strangers who would try
to take advantage of his kindness. But now Jesus speaks: ‘I was sent only to the
lost sheep of Israel.’ This is what the disciples were implying, but, more
importantly, it is something the Canaanite woman has to see. She has to see
that she is asking for an extension, an inclusion, a widening of God’s mercy.
She is, in effect, asking to enter into something that already existed, but did not
include her. She sees that the initial answer is, no. And she now comes close to
him, and kneels, and repeats her plea: ‘Lord, help me!’ And Jesus says what
seems harsh: ‘It is not right to take the children’s bread and toss it to the dogs.’
The dogs under the table should not get the bread, and, as he puts the matter,
the bread would have to be taken away from the children and given to the
dogs. This is, so to speak, zero sum. And this is how the blessing was under the
Abrahamic covenant, like Isaac’s blessing being taken away from Esau and
given to Jacob.81 But she will not accept this. She will accept that this is how it
was: ‘True, Lord,’ she says. But this is not how it has to be. There can be
enough for the children and enough for the dogs. She responds with wit: ‘Even
the dogs eat the crumbs that fall from their masters’ table.’ To reach this point
she has to accept that there was a covenant in which she was not included. Not
only was she not included; her people were driven from their land, and killed.
She carries with her that history (somewhat as Palestinians now identify
themselves as excluded from the same land). She has to accept that this
good thing that she wants, the healing of her daughter, has to be the gift of
the God of the people who hated her people. Perhaps Jesus looked into her
soul, and saw an impediment that had to be defeated by being gone through,
something bitter and painful. He finally answers her: ‘“Woman, you have
great faith! Your request is granted.” And her daughter was healed from that
very hour.’ Perhaps not only was her daughter healed, but she was healed also.

80 81
Mark 5: 21–43. Genesis 27: 35.
166 God’s Command
The relation is rather like the piano duet, or like Socrates’ relations with his
interlocutors in the early dialogues. Socrates seems to know what the inter-
locutor is going to say, but still cares strongly that he say it, and say it sincerely.
In our present story the momentum in the conversation comes from Jesus; but
there are also steps the woman has to take that Jesus wants from her.
This is speculative. But, if the reading is plausible, it shows us something
about how God’s effective grace is related to our choice and decision. Jesus
comes to the region of Tyre and Sidon where she is. She responds with desire
and longing for something he can give her. He points to an impediment. She
confesses and opens herself to his instruction. He gives her the terms in which
the solution can be found (‘bread’ and ‘children’ and ‘dogs’), but she has to
find the arrangement of the terms that provides a solution. He then does the
sign and wonder. There is always, first, something he does, and then some-
thing she does in response, enabled by what he does.82 And through all of this
there is a fulfilment of God’s purpose that Isaiah already foretold.

5.3. BARTH AND OUR ACCESS TO THE COMMANDS

5.3.1. Barth and Kant

The third theme for this chapter on Barth is Barth’s view of what kind of access
we have to the divine command. We can approach this topic indirectly, by
considering first Kant’s view of our access. Barth is a subtle reader of Kant, but
he gets Kant wrong at a few key moments. We will understand Barth better by
first looking at his relation to Kant’s moral theology in general, then reviewing
Kant’s own view of our access to the divine command, and then finally (in the
last subsection of this chapter) comparing Barth’s view to Kant’s on this more
specific topic.
There is a historical irony in this comparison. American evangelicals
reacted negatively to Barth in large part because they thought him too close
to Kant. Thus Cornelius Van Til called Kant’s philosophy ‘Criticism’, and he
said: ‘There must of necessity be a death-struggle whenever Christianity and
Criticism meet. And they meet at every front since the days when the
Copernicus of philosophy took his regular walks in Königsberg. It is Criticism,
too, which in the persons of Barth and Brunner meets and attacks the historic

82
It is interesting to compare Paul’s conversion story in Acts 9, where there is a sign and
wonder (the revelation and blinding) that comes before the response. Perhaps, like the case of the
woman with the bleeding and Jairus, there are opposite temptations through which the Canaan-
ite woman and Paul are being conducted. Paul has to be brought low by a power he has been
trying to destroy.
 Barth on Divine Command 167
Christian faith under cover of an orthodox-sounding theology.’83 Van Til took
Kant to reduce God to a regulative as opposed to a constitutive principle.84
These terms ‘regulative’ and ‘constitutive’ are indeed Kant’s terms. A regulative
principle is one that helps us organize our experience, but does not claim the
existence of anything; we are merely helped by thinking as if something existed.
A constitutive principle makes an actual existence claim. And Kant does say in
the first Critique that from within the theoretical use of reason the idea of God
has a regulative but not a constitutive use.85 But Kant also says that within the
practical use of reason this idea becomes constitutive.86 Barth gets Kant wrong
here in the same way as Van Til does. He says that, for Kant, ‘to speak of
existence or non-existence is per se not to speak of God’.87 The irony is that, if
Barth and Van Til had seen what Kant was doing here, they would have been
both less hostile to Kant and less hostile to each other.
Barth’s reaction to Kant is, as we would expect, a complex yes and no. The
key reason for his no is Kant’s Christology.88 Kant translates within the
boundaries of mere reason using the moral concepts, so that all reference to
particular persons is excluded. He is left with Christ as ‘humanity in its moral
perfection’. Barth concludes (as did Van Til) that in Kant’s philosophy of
religion ‘man is the measure of all things’ (including God).89 Barth sees in this
the beginning of a straight line to Albrecht Ritschl and human divinization.90
He says, again reflecting Van Til’s view: ‘The critical philosophy of religion
cannot therefore speak of revelation.’91 But the two concentric circles in

83
Cornelius Van Til, The New Modernism (London: James Clarke, 1946), 27.
84
For a fuller treatment, see my ‘Karl Barth, American Evangelicals, and Kant’, in Bruce
McCormack and Clifford Anderson (eds), Karl Barth and American Evangelicalism (Grand
Rapids, MI: Eerdmans, 2011), 73–90. There is, in the same volume, an excellent article by George
Harinck, ‘How Can an Elephant Understand a Whale and Vice Versa? The Dutch Origins of
Cornelius Van Til’s Appraisal of Karl Barth’, which explains more of the history here. Other
American evangelicals who took a similar view are Carl Henry, God, Revelation and Authority, iii
(Waco, TX: Word Books, 1979), and Gordon H. Clark, Karl Barth’s Theological Method (Nutley,
NJ: Presbyterian and Reformed Pub. Co., 1963).
85
KrV A644 = B672.
86
KpV v. 135. I will say more about what this means in Section 5.3.2.
87 88
Barth, Protestant Theology, 261. CD II/2. 651.
89
Barth, Protestant Theology, 290.
90
Here is another example of this. Barth says that for Kant historical faith (which accepts a
historical Christ) is ‘dead in itself ’; it is not a living, not a salutary, faith (ibid. 270). Here too Van
Til agrees with this reading of Kant, in The New Modernism, 81. (See Henry, God, Revelation and
Authority, iii. 432.) But in this passage to which Barth refers, Kant is quoting from James 2:17: ‘So
faith by itself, if it has no works, is dead’, and he is criticizing historical faith without corres-
ponding moral action (Rel. vi. 111). He is not saying that historical faith if lived together with the
moral disposition is dead or inert. Kant’s view of the historical part of the faith is that it has a
moral meaning that is valid for everyone, and that is why Christianity is ‘to be sure [zwar]
destined to be the religion of all’ (The End of All Things, viii. 339). The Cambridge translation is
misleading: ‘because Christianity, though supposedly destined to be the world religion’.
91
Barth, Protestant Theology, 270.
168 God’s Command
Religion—namely, historical revelation and the revelation to reason—are both
described as revelation. And ‘pure rationalism’, which Barth is right to attri-
bute to Kant, allows special revelation.92 One final point. Barth, like Van Til,
quotes Kant’s dictum that ‘the human being must make or have made himself
into whatever he is or should become in a moral sense, good or evil’.93 Barth
takes it that this removes the need for God’s grace. But Kant says three
sentences later: ‘Granted that some supernatural cooperation is also needed
to his becoming good or better.’ Kant means that we are to be called morally
better only to the extent that we are responsible for the improvement, but this
does not mean that we are solely responsible for our improvement.94 The
divine supplement that enables the revolution of the will in us is the effect of
grace. Kant’s view is that we can make room for it in our belief, even though
we cannot make use of it in our maxims of practical reason (because it is
something God does, not something we do) or theoretical reason (since it is
beyond the reach of our understanding).95 This is the argument from grace in
Chapter 1.
All this is part of Barth’s no to Kant’s moral theology (even if he does not get
Kant right here). What is Barth’s yes? He thinks that Kant himself ‘disturbs’
the ‘closed and rounded quality’ of his philosophy of religion by acknowledg-
ing human radical evil; this results in a ‘clash’ that makes Kant inconsistent
with himself.96 We might put the result this way: Barth says yes to Kant’s no. It
is interesting to compare Barth’s treatment of other figures, such as Novalis,
where we find the same interpretative strategy of positing an internal clash. It
is indeed a typical deconstructive strategy in the reading of a text, found in
many of Hegel’s heirs, such as in Derrida’s reading of Plato.97 The resulting
picture is odd, however, because Kant starts from radical evil. It is better to see
his philosophy of religion as a coherent unit, if we can. And to a remarkable
extent we can do this by being careful to distinguish at every point where Kant
is, so to speak, on the diagram of the two concentric circles. Barth himself
acknowledges that Kant’s account is not reductive, that he does not mean to
eliminate the outer circle in favour of the inner one. Here Barth is a forerunner
of the recovery of what I call Kant’s ‘vertical dimension’ in the last thirty years

92
Allen Wood denies this in ‘Kant’s Deism’, in Philip Rossi and Michael Wreen (eds), Kant’s
Philosophy of Religion Reconsidered (Bloomington, IN: Indiana University Press, 1991), and
I have replied to him in a review of the book in ‘Kant’s Philosophy of Religion Reconsidered’,
Faith and Philosophy, 11/1 (1994), 138–44.
93
Rel. vi. 44. Nicholas Wolterstorff calls this ‘the Stoic maxim’, in ‘Conundrums in Kant’s
Rational Religion’, in Rossi and Wreen (eds), Kant’s Philosophy of Religion Reconsidered, 40–53.
See Barth, Protestant Theology, 284, and Van Til, The New Modernism, 82. I have discussed this
passage in more detail in ‘The Place of Kant’s Theism in his Moral Philosophy’, 300–14.
94
This is Kant’s standard use of ‘moral’, e.g. Rel. vi. 21.
95 96
Rel. vi. 53. Barth, Protestant Theology, 290.
97
See Jacques Derrida, Dissemination, trans. Barbara Johnson (Chicago: University of Chicago
Press, 1981).
 Barth on Divine Command 169
of Kant scholarship. Indeed Barth claims for himself the space Kant leaves for
the biblical theologian, a space ignored by much Kant scholarship before the
last thirty years.98 We can think of Barth as having the same picture of the two
concentric circles, but putting biblical revelation in the centre, and the reve-
lation to reason on the outside. But then we should say that there is a ‘clash’ or
a ‘contradiction’ only if Kant tries to put his talk of the effects of grace inside
his inner circle, the circle of the revelation to reason; and Kant does not do
this. He puts them in what he calls the ‘parerga to religion within the
boundaries of pure reason; they do not belong within it yet border upon it’.99

5.3.2. Kant on Conscience

Having given the framework now of Barth’s relation to Kant’s moral theology
generally, we can look at Kant’s view of our access to the divine command. The
best place to see this is Kant’s account of conscience, and in particular a passage
about conscience from the Metaphysics of Morals that I will call ‘the conscience
passage’, and that we have discussed already in Section 2.3.2.100 When a person
is reflecting about the rightness or wrongness of what he has done, Kant suggests
he will think in terms of a judicial court, and the ‘consciousness of an internal
court in the human being (before which his thoughts accuse or excuse one
another) is conscience’.101 Kant argues that the idea of the judge is also the idea of
a figure outside the individual. There is an odd relation of the inside and the
outside here that is worth exploring.
We can compare the conscience passage with a passage from the third
Critique.102
Consider a person at those moments in which his mind is disposed to moral
sensation. If, surrounded by a beautiful nature, he finds himself in peaceful and
cheerful enjoyment of his existence, he feels a need to be thankful to someone for
it. Or if, on another occasion, he finds himself in the same state of mind under the
press of duties which he can and will satisfy only through voluntary sacrifice, he

98
Barth does not, however, occupy this space as peripheral, in the way Kant describes, but as
the centre. At CD II/2. 513, he is probably referring back to Kant’s discussion at Rel. vi. 9, and
reversing it.
99
Rel. vi. 52. Kant is not entirely consistent in his attitude to the question whether the outer
circle is, in the history of this world, to be replaced. See my The Moral Gap, 73–4.
100
MdS vi. 438–40. Much of the material in this subsection of the chapter is dealt with
more fully in my ‘Conscience and the Moral Epistemology of Divine Command Theory’.
101
Kant here describes conscience as retroactive, making a judgement about what has been
done; but I am going to widen the account to include also making a judgement about what one
might do in the future. Kant uses the term also in this wider way—e.g. MC xxvii. 354–5, a passage
that also describes, like the conscience passage, the different roles in the courtroom of defendant,
judge, prosecutor, and defence counsel.
102
KU v. 445.
170 God’s Command
feels a need to have done something that was commanded and to have obeyed an
overlord. Or if he has in some heedless way acted contrary to his duty, although
without having become answerable to other people, nevertheless a strong self-
reproach will speak to him as if it were the voice of a judge to whom he must give
account for his action.
In this passage we have a background picture of three roles played by the
divine sovereign that are analogous to roles played by human sovereigns,
though in the human case the roles can be divided between different people.103
These are the legislative, the executive, and the judicial roles, as mentioned in
Section 1.2. For the understanding of conscience the central role is the third.
This voice of the judge, Kant thinks, is heard by every human being. He thinks
it is possible to stun oneself, or put oneself to sleep, but not indefinitely; we
cannot help coming to ourselves, and hearing the voice. We can also avoid
heeding the voice, and we will do so to the extent that we are under the evil
maxim that subordinates our duty to our happiness. Nonetheless we cannot
escape the voice. It is, Kant says, like our shadow when we plan to run away.
In the conscience passage, Kant recognizes something odd about the rela-
tion of the inside and the outside. The business of conscience is a person’s
relation to herself, but Kant says we have to see ourselves as constrained to
carry on this business at the bidding of another person. So we have to divide
ourselves up.104 One thing that is ‘peculiar’, in Kant’s term, is that the self as
judge has to be pictured as having qualities that are inconsistent with being
human at all.105 Kant says pointedly that there are two possibilities here: this
pictured judge ‘may be an actual person or a merely ideal person that reason
creates for itself ’.106 As described in the previous subsection, his term for this
kind of mere idea is a ‘regulative’ principle, as opposed to a ‘constitutive’
principle that makes an existence claim. But even this ideal person has to be
conceived as ‘a scrutinizer of hearts’ (this is part of God’s role as judge, which
humans cannot carry out because they do not see the heart), who ‘imposes all
obligation’ (this is God’s role as legislator), and who ‘has all power to give
effect to his laws’ (this is God’s role as executive). In short, we have to picture
this merely ideal person as God, carrying out the threefold function of the
sovereign of the kingdom of ends, and therefore not as a human being, a mere

103
From the practical point of view we need enough unity of purpose so that these three roles
cohere in a single set of prescriptions, and, as in Psalm 85: 10, justice and peace embrace. This
does not show, from a theoretical point of view, that we do not have more than one being, or, as
Hume might suggest, a committee. But Chapter 1 referred to a text, Miscarriage, where Kant
does try to show this from a practical point of view.
104
There is a somewhat similar self-division in Kant’s discussion of atonement in Rel. vi.
72–8. In both cases, the human being has to be the same, numero idem, but also different, so that
(in the discussion of atonement) the new person after the revolution of the will can take the
punishment deserved by the old person under the evil maxim.
105 106
MdS vi. 439. MdS vi. 438.
 Barth on Divine Command 171
member of the kingdom of ends.107 The thought in the conscience passage is
not that I should think of myself as God, but that I should entertain within
myself the thought of what an actual God would prescribe, and the verdict that
an actual God would reach, and then I should make that verdict my own; I can
repeat in my own will the jubeo or veto (that is, the legislative function), and
I can repeat the verdict (that is, the judicial function), even though I cannot
repeat the omnipotent supervision (the executive function).108 I can reach this
thought about repetition, ‘in following out the analogy with a lawgiver for all
rational beings in the world’.109 This is not yet to say that there exists such a
lawgiver, which would be the other possibility (that the judge is an ‘actual
person’). Rather, the claim is that ‘human beings are merely pointed in the
direction of thinking of conscientiousness (which is also called religio) as
accountability to a holy being (morally lawgiving reason) distinct from us
yet present in our inmost being, and of submitting to the will of this being, as
the rule of justice’.
In passages like this Kant makes a regulative practical use of the idea of God.
It is quite possible for someone to ‘hear the voice’ without believing that there
exists a being outside herself whose voice it is. Kant is not proposing that the
phenomenology of conscience should be taken as a mystical experience, and
he is opposed to ‘enthusiastic’ claims to have ‘occult intercourse with God’.110
But his strategy, as with the regulative speculative use in the third Critique, is
to connect up the regulative practical use with a constitutive (existence-
claiming) use. We can turn to this next, in a passage from the Dialectic of
the second Critique, the section called ‘On the Primacy of Pure Practical
Reason in its Connection with Speculative Reason’.111 There are propositions

107
Gl. iv. 433–4.
108
Robert M. Adams has objections to the thought even if construed in this way, as ‘thinking of
what an actual God would prescribe’. See Adams, Finite and Infinite Goods, 255–6. But Kant can
concede that the motivational power of what an actual God in fact prescribes is greater than the
power of the merely regulative thought.
109
MdS vi. 440.
110
Rel. vi. 201, from the translation by Theodore M. Greene and Hoyt H. Hudson, Religion
within the Limits of Reason Alone (New York: Harper, 1960).
111
KpV v. 119. Kant is discussing in this passage propositions in which pure speculative and
pure practical reason are ‘united in one cognition’. By ‘primacy among two or more things
connected by reason’ (note: one and the same reason) he says that he means ‘the prerogative of
one to be the first determining ground of the connections with all the rest’. The interest of
speculative reason is the cognition of the object up to the highest a priori principles of its
constitution, and the interest of practical reason is the determination of the will with respect to
the final and complete end, which Kant has determined is the highest good—namely, the union
of happiness and virtue. If speculative reason had primacy, we would not assume to exist
‘anything beyond what can accredit its objective reality by manifest examples to be shown in
experience’. But Kant holds that pure practical reason has the primacy, and so we must accept
what ‘belongs inseparably to the interest of practical reason’, even if it extends beyond the
capacity of speculative reason to establish it affirmatively. Where there is no contradiction,
speculative reason has two tasks, first to accept these propositions and second ‘to try to unite
172 God’s Command
about what is the case—for example, ‘God exists’—which belong from their
content to speculative reason, but no experience can count as giving a manifest
example of them. Kant argues that practical reason requires belief in these
propositions, because their truth is necessary for the real possibility of the
highest good, which is the end that morality itself gives to us, the union of
happiness and virtue. This is what Chapter 1 called the argument from
providence.112 Speculative reason has first to accept these shared propositions
and then to unite them ‘as a foreign possession handed over to it’ with its own
concepts. There is a similar connecting operation within practical reason,
when we acknowledge that it is the same God, already pictured regulatively
in conscience and in religion (the recognition of our duties as God’s com-
mands) but without an existence claim, who is now justifiably believed to
exist.113
Kant says throughout the published corpus that we have to recognize our
duties as God’s commands. One conspicuous place is at the beginning of
Religion, part four, where Kant says that religion is (subjectively considered)
the recognition (Erkentniss) of all our duties as divine commands.114 This
definition of religion is similar to the passage from the third Critique that
I referred to earlier, in which Kant says that a person who feels himself under
some hard duties ‘will feel within himself a need that in performing them he
will also have carried out something commanded, and have obeyed some
sovereign’.
We can return now to the conscience passage from the Metaphysics of
Morals. Here too we have a regulative principle within the practical use of
reason. Conscience gives us the voice we hear, pronouncing a verdict on
whether we have done our duty or failed to do it. Kant wants us to do the
same kind of uniting or connecting up as in the speculative case, so that we
end with the same God who is pictured merely regulatively now the object of a

them, as a foreign possession handed over to it, with its own concepts’. This takes us to the
connecting-up operation mentioned above.
112
Andrew Chignell’s work is important in understanding this. See, e.g., ‘Belief in Kant’,
Philosophical Review, 116 (2007), 323–60. I am very grateful to him for extensive comments on
an earlier draft. See also Patrick Kain, ‘Practical Cognition, Intuition and the Fact of Reason’, in
Benjamin Lipscomb and James Krueger (eds), Kant’s Moral Metaphysics (Berlin: DeGruyter,
2010), 211–30.
113
The term ‘justifiably’ here should not be taken to imply an epistemic justification in Kant’s
own restrictive sense of ‘knowledge’. There is another difference, also. The regulative use within
the speculative employment of reason (e.g., the ens realissimum of KrV A576 = B604) may be
more ambitious than anything validated by reason’s practical employment.
114
Rel. vi. 154–5. This is a hard passage. I am grateful to Patrick Kain for thinking through it
with me, though he should not be held to agree with my conclusions. I think that Kant means
that we have to have an assertoric faith that the highest good will obtain, and that this requires in
the end an assertoric faith in the existence of a God who has the attributes necessary to secure the
consistency of happiness and virtue. But we should not pretend to be able to secure objective
reality for this idea through theoretical cognition.
 Barth on Divine Command 173
justified existence claim. This can be seen in those passages where Kant makes
it clear that the God who sustains the real possibility of the highest good is also
the God whose commands are our duties, and who judges our failures to
comply with our duties. Sometimes these are distinguished as legislative and
judicial functions, and the maintenance of the moral order of the cosmos is the
executive function, God thus happily combining the three humanly separable
functions of government. The executive function is the key to the connecting-
up operation just described. Kant has the argument from providence, in the
Dialectic of the second Critique where we have just been looking at it, but also
at the end of the first and third Critiques and at the beginning of Religion, that
we have to believe in the existence of a Supreme Being to carry out this
function. But then, in this same passage from the second Critique, he extends
the status given to the executive function to the legislative and judicial. Reason
requires us to believe in the existence of a Supreme Being to prescribe the law
and promulgate it, and to hold us accountable to it.115

5.3.3. The Tradition’s Resources for Assessing the Command

We can now return to Barth, and compare his view with Kant’s on the nature
of our access to the divine command. We can think of him playing the role
that Kant assigns to the ‘biblical theologian’ (as opposed to the ‘philosophical
theologian’, who is Kant himself). The biblical theologian has a whole range of
resources for answering the question ‘How do we tell what is a divine
command?’, resources that are not available to Kant within the boundaries
of mere reason. Kant does not take himself to have a phenomenology of the
reception of what I will call ‘direct’ divine prescription.116 A direct divine
prescription is one that is a gift, standardly received in prayer, but there is a
spectrum here. A prescription can present itself immediately as an ‘extraor-
dinary’ divine gift (perhaps there is an auditory sensation), or much more
often it presents itself as an ‘ordinary’ part of our reflection that we then
recognize as God speaking to us, rather than our simply working out what to
do unaided. Both of these may be God’s direct commands or counsels or
permissions. The contrast is with working out what we ought to do by some
ethical decision procedure such as the Categorical Imperative test, and then
attributing the resulting prescription to God because we think God has
authorized the procedure. Perhaps an example of an ordinary direct prescrip-
tion will help. Suppose a man and his fiancée are trying to settle on the vows
for their wedding, and he prefers the traditional language of ‘thee’ and she

115
The first point, about the executive function, is given at KpV v. 129, and the second point,
about the other two functions, at KpV v. 131.
116
I am grateful to John Pittard for clarifying the thought here.
174 God’s Command
prefers the fresher ‘you’. Perhaps they have reached an impasse, and this
dispute has come to symbolize all sorts of deep differences between them.
They decide to pray, and the idea comes to him: ‘I will say “thee” and you will
say “you”.’ They recognize this as a divine counsel.
As quoted earlier, Kant is opposed to the idea of ‘occult intercourse with
God’, and he is no doubt reacting here primarily to the claims of ‘extraordin-
ary’ experience. But his critique applies to ordinary direct prescriptions as well
as extraordinary ones. The problem is epistemological, though it is not about
‘knowledge’ in Kant’s technical and restrictive sense of that term, in which we
can know only what we could possibly experience with the senses. Kant has
told us that we are to recognize our duties as God’s commands. But his
conception is that morality, determined under the procedure of testing a
maxim under the Categorical Imperative in its various formulations, tells us
what our duties are, and we then attribute these to the divine commander. The
present question is the other way round. Suppose I receive what seems to be a
direct divine command—for example, to take my son, whom I love, up Mount
Moriah and sacrifice him. Kant discusses this example both in Religion and in
Conflict of the Faculties.117 In Religion, his view is not that God could not give
such a command, but that we should never take it to be God’s command,
because it conflicts with something that is demonstrably certain, that an
innocent should not be killed. Adams, in Finite and Infinite Goods, takes the
same position: Abraham should have said, ‘It’s not God telling me to do
that.’118 In Conflict Kant goes beyond Religion in a sceptical direction, and
this is typical of the differences between the two works. In Conflict Kant says
that we cannot apprehend any command from an infinite being and be
acquainted with it as such. In Religion he says not this, but rather that, if we
did receive a command, especially a command in conflict with a clear pro-
nouncement of our practical reason, we would have to be uncertain about
whether it was God’s communication to us. Similarly, in Conflict he says that
we can be sure that the voice is not God’s. In Religion he says that we cannot be
sure that it is God’s. The more cautious statement in Religion is more in line
with Kant’s other commitments. To claim certainty or sureness in such
matters goes beyond the proper limits of the understanding. But, in any
case, Kant’s position even in Religion is hard to fit with the story in Genesis,
in which Abraham is commended for not withholding his son, his only son,
and God gives a blessing, ‘and by your offspring shall all the nations of the
earth gain blessing for themselves, because you have obeyed my voice’.119 The
benefit of supposing that Abraham was obeying God is that it is consistent
with the status of Abraham within all three Abrahamic faiths. Kant’s position

117 118
Rel. vi. 186–7, and SF vii. 63. Adams, Finite and Infinite Goods, 277–91.
119
Genesis 22: 18. The Hebrew for ‘obeyed’ contains the root shema (‘hear’, as in ‘Hear,
O Israel, the statutes and ordinances that I am addressing you today’ (Deuteronomy 5: 1).
 Barth on Divine Command 175
here is hard to square with his own programme as described in the first preface
to Religion, where he says that his task is philosophical theology, which leaves
biblical theology as it is.120
Kant is not engaged with the question how we tell whether God is command-
ing something, except that he thinks we should not say God is commanding
something contrary to the moral law. But any divine command theorist has to
have something to say to someone who thinks God is telling him, for example,
to hijack an aircraft and fly it into a skyscraper. All three Abrahamic faiths can
show conspicuous and horrible examples of adherents doing things out of what
they took to be obedience to God. A divine command theorist does not,
however, have to take Kant’s position about Abraham. Here is one suggestion,
made after reflecting on a text of Maimonides, though it does not remove all the
difficulties.121 The suggestion is that God in this encounter replaces child
sacrifice with animal sacrifice. God’s intention in providing animal sacrifice
is, in mercy, to accommodate our need to sacrifice something, but to end the
pagan practice (in which perhaps Abraham grew up) in which people sacrificed
their own children to demonstrate their devotion to their gods. Maimonides
does not propose that the system of animal sacrifice in the Pentateuch is for his
own time, or even that it would be for his own time if the temple were restored.
If we read the story as God’s deliberately bringing Abraham through a transi-
tion from the pagan culture he grew up in, we can see that revelation is
progressive. We are now in a different position from Abraham, exactly because
we come after Abraham’s story with Isaac or Ishmael. We know now, as
Abraham did not at the beginning of the story, that we are not to demonstrate
our devotion to God by killing our children. There is thus now a resource
internal to the tradition for constraining what we should take to be a divine
prescription.
So we can ask what constraints there are in the historical revelation that
Kant is translating within the boundaries of mere reason for answering a
question Kant is not himself engaged with (except by giving one negative
condition), the question: ‘How do we tell what purportedly direct divine
prescriptions are genuine?’ Many of these constraints will be from the content
of the narrative, or from procedures internal to the tradition about checking
purported divine communication with other members of the faith community.
But there are at least five features of the phenomenology of receiving a divine
prescription that Barth describes: clarity and distinctness, external origin,
familiarity, authority, and providential care. Together, these five features give

120
Rel. vi. 9.
121
Maimonides, Guide, iii. 32. Why did not God simply tell Abraham that killing one’s first-
born was wrong, and killing an animal was all right? This is speculation. Perhaps Abraham had
to form the intention to do the act himself in order to see fully what was wrong with his ancestral
practice.
176 God’s Command
us a significant additional constraint on what we should take to be divine
prescription. An outside observer, who is not herself a member of the traditions
in question, can still agree that divine prescriptions appear a certain way within
these traditions. She can agree to that, even if she does not believe in God or gods
at all. There is, to be sure, a kind of circularity in the account that follows, and
the end of this subsection will address the question whether this circularity is
vicious. The present suggestion repeats something like Kant’s strategy described
in the previous subsection. The phenomenology, like a Kantian regulative
principle, does not itself license the conclusion that a divine prescriber exists.
But we can connect it up with this existence claim. If we believe in such a
prescriber, then we can find in the phenomenology a useful set of constraints on
what we should take to be a divine prescription.122 These constraints should not
be taken as each by itself sufficient to rule out a divine source, but their force is
cumulative. I am also not claiming that the constraints have equal force.123 In
what follows, we will focus on commands as Barth does, but a fuller treatment
would look at the different phenomenology of the different types of divine
prescription distinguished in Chapter 2. Precepts and prohibitions can feel
different in their reception from permissions and from counsels.
The first feature of the phenomenology is that what one takes to be a
command has a certain kind of clarity or distinctness.124 Often, in the moral
life, there is a blur or cacophony of indistinct evaluative impressions, with
none of them standing out clearly marked as deserving attention. Someone
who is receiving and obeying a direct divine command tends to hear it
distinctly; if it is a divine prohibition, it has, to use a visual analogy, a black
line around it. Or, to use an aural analogy, a divine precept has a resonance to
it, though this does not mean it has to be loud. Different parts of the Christian
tradition put different emphasis on whether Christians experience receiving
divine command as something different from the application of general God-
given principles or whether ‘ordinary’ Christian moral deliberation just is our
application to particular situations of the general commands given us by God.
Very roughly, this is a difference between Protestant and Roman Catholic

122
This leaves open a large question. It is controversial whether a belief that God or gods exist
and relate to us by command is the sort of belief that requires justification by some other more
basic belief or set of beliefs. Arguably it does not (though I would distinguish it here from more
theologically particularist claims such as Trinity or atonement), but this chapter is not the right
place to defend this claim. It is worth adding that the fact that the belief in moralizing high gods
is less pervasive than religion (as discussed in the final chapter) is relevant to determining
whether such belief is properly basic.
123
The fifth is the most important, but I do not have a worked-out theory of how to
rank them.
124
These are Descartes’s terms in Meditations, synopsis 13, and they are misleading if they
suggest an analogy with mathematics. In what follows the essential points are that the commands
stand out from the general blur, and they are persistent.
 Barth on Divine Command 177
moral theology.125 Barth is Protestant in terms of this distinction. He empha-
sizes the particularity of divine command. The concept of command, he says,
brings with it the concept of ‘the God who has a personal life and therefore
acts and speaks directly and concretely’, so that ‘the command has a concrete
content’.126 This is true, even though for Barth the command is always the
intersection of the vertical, God’s activity, with the horizontal, the continuity
of our own lives and concepts. He thinks God’s command standardly comes to
us in prayer.127 He holds that the command is self-interpreting, but he does
not deny that we sometimes experience it as ambiguous, or faint. This,
however, is our difficulty, created by our constant tendency to turn away
from God. The situation of a person listening obediently will be one in
which the command is clearly and distinctly heard. Barth puts this point by
saying: ‘From God’s side nothing is hidden at this point.’128 But it does not
follow that, from our side, what we receive is completely unambiguous.
Barth draws a distinction between two German terms for knowledge,
kennen (to know by acquaintance) and wissen (to be fully conversant with),
and claims that we have the first but not the second in our reception of the
divine command. Only to a limited extent can we now have conscience
(Gewissen), which Barth defines as ‘the totality of our self-consciousness
insofar as it can receive and then proclaim the Word and therefore the
command of God that is given to us, insofar as we . . . can become co-knowers
[Mit-Wisse] (suneidotes, conscientes) with God. That Word and command
that is given to us is as such the promise that we can become this.’129 Barth is
here postponing our knowledge in the sense of wissen of God’s command to
our final state, when we can know even as we are known. Here and now, we do
not see the ‘Thou–I’, which we are before God. We do, however, have
knowledge by acquaintance, enough to go on, so that we can be held account-
able. Here is an analogy.130 I may know that my wife wants me to buy
something for dinner because she has had an extra-long day of teaching,
even though she has not given me this request, and we do not have a standing

125
James M. Gustafson, Protestant and Catholic Ethics: Prospects for Rapprochement (Chicago:
University of Chicago Press, 1978), 44–6. But this is only a rough distinction. I am grateful to
Brian Besong for pointing me to various counter-examples, such as Cardinal Newman in the fifth
chapter of Grammar of Assent (London: Assumption Press, 2013) and Mother Teresa’s ‘locution’
telling her to go to Calcutta.
126
CD II/2. 38, 667–9.
127
See the end of Section 5.2.1. See also Biggar, The Hastening that Waits, 7–45, who argues
persuasively that Barth does not mean to exclude a place for human moral deliberation. See
also Oliver O’Donovan, Finding and Seeking (Grand Rapids, MI: Eerdmans, 2014), 188–90,
who says that Barth is guilty of ‘the refusal of a role to intelligence’, because he is too worried
about compromising theological distinctiveness.
128
CD II/2. 704.
129
CD II/2. 667. See McKenny, The Analogy of Grace, 283.
130
I owe this to Gerald McKenny in private communication.
178 God’s Command
agreement that covers this case. This is part of my relationship with her, and
I can be confident of the request, even though I have had no sensory experi-
ence of it. Barth, like Kant, is worried about ‘occult intercourse’. He does not
want to base our access to divine command on some special ‘mystical’
experience, though he does not deny the possibility of ‘locutions’ such as are
reported by some of the saints. But he thinks that God is nonetheless able to
communicate to us the divine will in our prayerful reflection. Moreover, the
command is heard as persistent and not easily shaken. It resists our attempts
to ignore it. Kant translates this feature for the case of his conception of divine
command by talking about how we can put ourselves to sleep, but not
indefinitely. He says we can avoid heeding the voice, but we cannot escape
it; it is like our shadow when we plan to run away.
If we perceive God as the source and ourselves as the receivers, we should
expect this combination of clarity and persistence in the perceived source and
continual resistance in the recipient. On the analogy with human legislation,
the job of the legislator is to promulgate clearly; without this, the citizen is not
properly held accountable to the law.131 The citizen may or may not pay
attention to the law. Even if she does not, however, that is no excuse. She may
not have noticed that the speed limit has been reduced on a stretch of road
going through what looks like open country; but that does not save her from
the resultant ticket, unless she can show that the sign announcing the change
of speed limit was in some way obscured by something (like an overgrown
bush) that was not her own fault. But our failure to hear the command clearly
often is our fault, and, as Buber says, we have plugged wax into our own ears.
We are not, therefore, excused from obedience by our lack of hearing.
The second feature of the phenomenology of receiving direct divine com-
mand is that the command presents itself as having an external origin, either
immediately or mediately. This is what Kant explores in the conscience
passage in relation to the roles of legislator and judge. One ‘peculiar’ feature
of conscience, as he describes it, is that the self as judge has to be pictured as
having qualities that are inconsistent with being human at all. Kant suggests
that we have to think of ourselves as ‘accountable to a holy being, morally
lawgiving reason, distinct from us yet present in our inmost being, and of
submitting to the will of this being, as the rule of justice’. In the tradition that
Kant is translating within the boundaries of mere reason, and in traditions like
it in the relevant respects (including traditions within Judaism and Islam), this
is indeed how divine command is pictured. It is also true that, as with the first
feature, there is a distinction within all three Abrahamic faiths between those
who emphasize the reception of direct divine command as experienced sep-
arately, in my term as ‘extraordinary’, and those who emphasize its continuity

131
See Aquinas, ST I–II. 90. 4: ‘Without promulgation there is no law at all.’
 Barth on Divine Command 179
with ‘ordinary’ moral deliberation. Barth approves of Kant on this point of the
externality of the demand. He says that Kant ‘has expressed the essential
concern of Christian ethics’ in saying that obligation comes to me from
outside myself, as the command of some source other than myself.132 But
for Barth the essential point is Christological. Christ calls us to participation,
and we should expect that this call would appear to be from a source that is
already good in a way that we are not yet capable of being.133
Finally, however, it is worth noting that, as with Kant’s regulative principles,
we are not licensed to move from the premiss that something appears to us a
certain way (namely, as externally sourced), to the conclusion that it is that
way. It is perfectly possible (and it sometimes happens) that what we are
‘receiving’ is an illusion.
The third feature of the phenomenology is that the command comes in a
familiar voice.134 A person learns through experience to trust this voice. The
first encounter may be strange. The boy Samuel thought it was Eli calling, not
God.135 But the recipient obeys the command on one occasion, even if she is
not given the reason for it, and she subsequently sees the fruit of this
obedience. She takes this memory with her into her next encounter, and trusts
that obedience a second time will produce blessing as it did the first time. This
means that the proper reception and recognition of divine command stand-
ardly requires practice, and all three Abrahamic faiths have accounts of what
kind of practice is involved. If we want to receive the command and hear it
clearly, we need to learn what this voice is like by consistently listening for it,
and this takes a life of discipline. Then we can more easily separate it out from
the buzzing confusion of mental contents competing for our attention, just as
we separate a face we love from a crowd.
Barth has a good deal to say about instruction, as opposed to reflection
(which is where the encounter with the divine demand takes place).
Section 5.1.4 described this distinction. Barth’s view here has not always
been recognized, because of his emphatic no to casuistry.136 But it is important
to see that what he dislikes about casuistry is any attempt to determine God’s
will in advance of the encounter. He is not an opponent, indeed he is
convinced of the necessity, of a lifetime of preparation for the encounter, so
that one is in a state of readiness for it. ‘To examine ourselves means . . . to

132
CD II/2. 649–51.
133
See McKenny, The Analogy of Grace, 282: ‘The capacity to decide what corresponds at any
point in our lives to the fulfillment of electing grace in Christ comes to us only in confrontation
with the gracious God.’
134
This is consistent with saying that the voice may be recognized in something unfamiliar,
or something not familiarly associated with this voice. For example, Augustine hears ‘tolle, lege’
from a children’s game in a nearby house (Confessions VIII. Xii. 29). I am grateful to Brian
Besong for this point.
135 136
1 Samuel 3: 4–5. See Biggar, The Hastening that Waits, 40–5.
180 God’s Command
prepare ourselves for the encounter with our Judge.’137 For Barth, the Lord’s
Supper is the focus of this self-examination (following Paul’s exhortation in
1 Corinthians 11). But he spends hundreds of pages of the Church Dogmatics
doing preparation for the encounter in careful thought. It involves thinking
through the history of God’s dealings with people in general, and with oneself
in particular, and thinking through the different ‘domains’ in which this
dealing takes place. Much of this preparatory work sounds a lot like natural
law theorizing, but the key difference is that Barth wants to leave open God’s
verdict on any particular decision. Like Scotus, Barth denies that we can
deduce God’s will for some particular situation from antecedent truths about
creation, but he does not deny the fittingness of this will to this creation.
Section 5.1.4 analysed this fittingness in terms of the goodness (as opposed to
the obligatoriness) of the item in the action position of a moral judgement. In
the present context, the important point is that discerning this goodness is a
matter of practice, a disciplined ordering of the loves that we have, so that we
can see clearly what goods are at stake.
Again, we are not licensed to move from this premiss to the conclusion that
there is a divine being addressing us. It is always possible that we have simply
habituated ourselves to taking a certain kind of purely internal prompting as
authoritative. Nonetheless this feature is important in highlighting that what is
normative is a lifetime of response, not the immediate response to a single
momentary input.
As with the first two features, we would expect to find a feature like this in
the picture of communication from a divine ruler who has in mind the well-
being of the kingdom’s members over their lives as a whole. When Kant
describes the head of the kingdom of ends as securing the real possibility of
the highest good, he is translating the idea of a ruler who requires righteous-
ness and speaks peace, so that, in the words of the Psalmist, ‘righteousness and
peace will kiss each other’.138 This peace, or shalom, and this righteousness are
both features of whole lives lived in obedience, not of mere discrete episodes of
doing what we are told to do. Christian tradition leads us to expect that we
would receive these commands often and cumulatively, since God has in mind
a long-term relationship that has a character shaped by our interactions.
The fourth feature of the phenomenology is that the commands carry about
them a sense of conviction or authority. There is a comparison here with
William James, who attributed this feature to what he called ‘mystical experi-
ences’. ‘They carry with them’, he said, ‘a curious sense of authority for
aftertime’.139 Barth is not defending a Jamesian view that there is a common
core to all religious experience. But the phenomenology of receiving a divine
command has, for Barth, this same character of perceived authority. This is

137 138
CD II/2. 640. See III/4. 5–6. Psalm 85: 10.
139
William James, Varieties of Religious Experience (New York: Modern Library, 1902), 371.
 Barth on Divine Command 181
not at all surprising. One is likely to take as authoritative what one takes to be a
divine command (unless, like Milton’s Lucifer, one has said ‘Evil be thou my
good’, and one has taken evil to be what God forbids). Barth says: ‘We
subordinate ourselves to what he wills and orders, and our action is brought
into line with his command.’140 He draws a dividing line here between Kant,
who preserves this sense of external authority, and Fichte, who does not. Thus
he describes the step from Kant to Fichte as ‘the true fall of German Ideal-
ism’.141 For Fichte, on this account, the moral law is that in which and by
which I posit myself, because of the identity Fichte posits between the subject
and the whole. But:
If there is an ought, it must not be the product of my own will, but touch from
outside the whole area of what I can will of myself. It must lay upon me the
obligation of unconditioned truth—truth which is not conditioned by myself. Its
authority and power to do so must be intrinsic and objective, and not something
which I lend to it.142
This authority is what God exercises as the judge, ‘who sees both our actions
and also ourselves, the heart from which they proceed’.143 Like Kant in the
conscience passage, Barth is here emphasizing the externality of the judge.
The point for now is just that people within the tradition Kant is translating
do have this experience of hearing an authoritative voice that presents itself as
externally sourced. This is the evaluative correlate of persistence, as described
in relation to the first feature of the phenomenology. Persistence, we might
say, is a matter of power; the ‘voice’ is not easily ignored. But this could be true
of some annoying jingle that one cannot get out of one’s head. The point of
authority as opposed to mere power is that one perceives the voice as deserving
to be attended to, or heeded, whether it is in fact attended to or not.144
The fifth and most important feature of the phenomenology is that the
commands appear to be from a loving or merciful source.145 Adams adds this
feature to his account of divine command, following Paul Tillich’s account of
what Tillich calls ‘theonomy’.146 Obligation, on Adams’s account, is consti-
tuted by the commands of a loving God, and ‘respect for divine authority
motivates, largely because it coheres with, organizes, supports, and is sup-
ported by goods that we care about for their own sakes’. Adams’s account of
Abraham on Mount Moriah was rejected earlier. But a substantial restriction

140
CD II/2. 657.
141
Karl Barth, Ethics, ed. Dietrich Braun, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark,
1981), 86. See CD II/2. 651. Whether Barth is fair to Fichte is a different question.
142 143
CD II/2. 651. CD II/2. 637.
144
See Butler, Fifteen Sermons, 45: ‘Had [conscience] strength, as it had right: had it power, as
it had manifest authority, it would absolutely govern the world.’
145
See Aquinas, ST I–II. 90. 4: a law ‘is nothing other than a certain dictate of reason for the
common good, made by him who has the care of the community and promulgated’ (emphasis added).
146
Adams, Finite and Infinite Goods, 274–5.
182 God’s Command
on what we take to be a divine command can be derived explicitly from the
tradition itself. This was the reading of the Abraham story proposed at the
beginning of this subsection, and it puts us in a different position from that of
Abraham himself. Anything that we take to be a divine command has to be
consistent with the character, the providential care, of the God who is sup-
posed to be communicating with us. Discerning this depends partly on the
content of the command, but partly also on whether the experience of
receiving it is an experience of being loved. For Christians, the central reve-
lation of this character will be the life of Jesus, including the Sermon on the
Mount and the Golden Rule that lies at the heart of it. For Muslims it will be
the revelation in the Qur’an of God’s mercy.
Barth repeatedly emphasizes that our reception of the divine command and
our responsiveness to it are conditioned on our election, on God’s gracious
action towards us. ‘God elects himself to be gracious toward man, to be his
Lord and Helper, and in so doing He elects man to be the witness of his
glory.’147 This election is in Christ, who is both divine and human, so that ‘in
his person God has acted rightly towards us. And in the same person man has
also acted rightly for us.’148 It would take us too far afield to unpack the
theology here. For the purpose of the present chapter, all that is needed is the
point that law, the ought, is already gospel, the good news of what Christ has
done for us. The divine command is sanctifying, indicating how we are (in
Scotus’s terms) to repeat in our will God’s will for our willing, and that we will
be able to do this (as Jesus’s command to Peter enabled him to get out of the
boat and start walking on the water).149
It is natural to reply, ‘Well, that didn’t seem to be enough on 9/11, or during
the crusades.’ But it is too much to require that a moral restriction on what is
taken to be divine command in fact be persuasive to all those to whom it
applies. If what we want is effectiveness, it is more likely to be effective to call
on a person’s own religious traditions than to rely on some religiously neutral
statement of a moral constraint. But there may be no way that somebody who
has a perverse understanding of God’s will can be stopped by moral suasion;
perhaps force or the threat of force is the only thing that will work.
There is an apparent problem of circularity that affects most of the discus-
sion in this final subsection. ‘Of course,’ it may be objected, ‘if you construe
yourself as under the rule of an omnipotent, omniscient and omnibenevolent
being with legislative, executive and judicial authority, you will experience in a
certain way what you take as “being commanded”; but this is only because you
have built this theology into your construction of the experience.’ Seeing our
way here is very tricky.150 It would be wrong to argue that the experiences just

147 148 149

CD II/2. 510. CD II/2. 538. Matthew 14: 28.
150
There is a similar problem addressed by William Alston in his account of religious
experience, ‘Religious Experience and Religious Belief ’, Nous, 16 (1982), 3–12. Alston tries to
 Barth on Divine Command 183
described demonstrate that there is a being of the kind described in the
tradition Kant is translating. That would be viciously circular in the way the
objection says it is. However, the present proposal is something different.
Suppose we ask, ‘What resources do we have within the tradition Kant was
translating within the boundaries of mere reason for limiting what we take to
be a direct divine command?’ One place to look is the experience of receiving
such a command, as this experience is understood within this tradition. The
claim here is that this experience is understood in terms of a certain theology,
and in particular in terms of the divine commander having legislative, execu-
tive, and judicial authority. These five phenomenological features taken to-
gether do put significant moral limits on what we will properly take to be a
divine command within this tradition. Objecting that the experience is con-
structed in terms of the theology is hardly persuasive, because that is just the
point. The experience, constructed in just this way, puts significant moral
limits on what those within this tradition and others like it in the relevant
respects should construe as a command of God. This is true, even though
the phenomenology does not itself license the conclusion that such a God
exists.151

establish that beliefs based on religious experience have an epistemological status on a par with
beliefs based on ordinary perceptual experience. An objection might be that with perceptual
practice we can have checking by others, and prediction, and that capacity for perception is
universal among normal human beings, and the concepts used for objectifying perceptual
experience are likewise, at least roughly, universal. Religious experience is unlike ordinary
perceptual experience in all these ways. But Alston replies that, if there is a God, who is pure
Spirit, and ‘wholly other’, and only present to the awareness of humans under certain difficult-to-
obtain conditions, then these asymmetries are exactly what we would expect. It would not be
reasonable to object that we have simply constructed the experience in terms of the theology.
151
I am grateful to Michael Bergmann, Brian Besong, and Patrick Kain for comments on an
earlier draft of this section of the chapter.
 6

Divine Command in Some Medieval

Islamic Thinkers

INTRODUCTION

This chapter is about the concept of divine command in three medieval

Islamic thinkers. It strays, like the previous chapter, beyond the territory in
which the author can properly claim to be expert. Many of the works of the
central figure discussed in this chapter, al-Maturidi of Samarqand (d. 944),
have not been translated into English. Some of the texts quoted will appear
here in English for the first time, and I have relied upon other people for the
translation from Arabic.1 Other texts have appeared in secondary work in
English, and I rely especially on Mustafa Cerić and Ulrich Rudolph.2 Because
al-Maturidi’s work is unfamiliar in the West, I have made more extensive
quotation of him than of any other figure discussed in this book. This chapter
locates al-Maturidi as taking a plausible mediating position between an ex-
treme form of divine command theory (in al-Ash‘ari) and an extreme form of
natural-reason theory (in the Mu‘tazilites, especially ‘Abd al-Jabbar).
Despite reservations, I have undertaken this part of the project because the
concept of divine command is central outside the Christian tradition as well as
within it, and there is a great deal to be learnt from the comparison. Within
medieval Islam, and within contemporary Jewish appropriations of medieval
Judaism (as argued in Chapter 7), there is very much the same range of
options in understanding the relation between a sovereign God who gives us

1
I am especially grateful to Humeyra Karagozoglu Ozturan, who provided many of the initial
translations, and to Frank Griffel, Stephen Ogden, and Geoffrey Moseley, who both translated
and saved me from many errors. They are not accountable for the errors that remain.
2
Mustafa Cerić, Roots of Synthetic Theology in Islam: A Study in the Theology of Abu Mansur
al-Maturidi (d. 333/944) (Kuala Lumpur: International Institute of Islamic Thought and Civil-
ization, 1995) (henceforth Roots). Ulrich Rudolph sent me part of a manuscript, Al-Maturidi and
Sunni Theology in Samarkand (Leiden: Brill, 2015) (henceforth Theology). I have also been
helped by the introduction by Fatholla Kholeif to his edition of Abu Mansur al-Maturidi, Kitab-
al-Tawhid (Beirut: Dar el-Machreg Editeurs, 1970) (henceforth Tawhid).
 Divine Command in Some Medieval Islamic Thinkers 185
commands and our own reason, as we try to determine how to live our lives.
This book assumes, without arguing for it, that the three Abrahamic faiths
worship the same God, though they say very different things about this God.
The doctrines of the Trinity and of the twofold nature of Christ are the central
examples of the difference. Jews do not accept these, but Christians do not
usually conclude that Jews do not worship the same God. Christians should
say the same about Muslims. A useful side effect for a Christian of examining
divine command in Judaism and Islam is that new light gets shed on areas of
the Christian’s own faith that had tended to get obscured. Psalm 119, for
example, acquires fresh meaning, and doctrines of divine concurrence. But
there are also places where the Christian will benefit from seeing the important
consequences of the absence of the Christian doctrine of the work of Christ.
The chapter covers just three thinkers, and has no pretension to be talking
about Islamic ethics as a whole. The scope is relatively modest, limited by the
author’s competence. But there is something like an obligation, if one thinks
one has something useful to say about divine command, to relate this to the
faith of over a billion people for whom divine command is a central concern. It
is the confinement to a discussion of Christianity that requires justification,
not the inclusion of a discussion of Islam.
This chapter locates al-Maturidi against the background of a dispute
between Mu‘tazilites and Ash‘arites about three questions. The first is whether
acts and persons have intrinsic value (or whether that value is to be under-
stood only as a divine willing or commanding), and what kind of access we
have to that value. The second question is whether human beings have
freedom of choice in what they do, or whether our actions are only the product
of divine causation. The third question is whether there is any proper use of
human reason independent of divine revelation, or whether the proper use is
only derivative from what we are given in the Qur’an and the Traditions.
These three questions are continuous with themes in Chapter 5. The first
question, about our access to value, presented itself to Barth in terms of his
concern that we not regard ourselves as ethical sovereigns. The second ques-
tion, about human freedom, arose in his insistence that in prayer, when we
receive the command, there are two subjects interacting. The third question,
about reason and revelation, came to Barth in terms of his relation to Kant and
his accepting the role Kant assigned to the biblical theologian.
There are many differences between Mu‘tazilites (especially between the
schools from Baghdad and from Basra), and this chapter relies mainly on the
texts of ‘Abd al-Jabbar (from Basra, d. 1025), who gives the fullest account.3
There is, however, a chronological awkwardness here. Al-Ash‘ari (d. 935) and

3
The system is described in George Hourani, Islamic Rationalism: The Ethics of ‘Abd
al-Jabbar (Oxford: Clarendon Press, 1971), and Sophia Vasalou, Moral Agents and their Deserts:
The Character of Mu‘tazilite Ethics (Princeton: Princeton University Press, 2008).
186 God’s Command
al-Maturidi are roughly contemporaries, though there is no evidence that they
met. They are both responding to people they call ‘Mu‘tazilites’, and indeed al-
Ash‘ari started off as a Mu‘tazilite under the tutelage of al-Jubba’i of Basra
(d. 915), but they do not by and large name names. ‘Abd al-Jabbar lived almost
a hundred years after them, and they are therefore not responding to his version
of the arguments (which is, in many cases, a refinement of them).4 In the case of
each of the three questions, the chapter starts with the Mu‘tazilite position, and
continues with the Ash‘arite response. It then locates al-Maturidi between the
two, taking something from each side. But both al-Ash‘ari and ‘Abd al-Jabbar
also see themselves as taking middle positions, and indeed we should expect this
because the Qur’an itself recommends this strategy: ‘Thus We have made you to
be a community of the middle [road]’ (2: 143). ‘Middle-ness’ is not itself truth-
marking. Everything depends on what the extremes are between which the
middle ground is being claimed. But, for our purposes, it will be helpful to
see analogies and disanalogies between the kind of middle ground staked out
by al-Maturidi and the kind of middle ground staked out by Duns Scotus, as
described in earlier chapters. We come back to this at the end of the chapter.
Two other general comments will be helpful in what follows. The first is
about the understanding and interpretation of law, which is chronologically
antecedent in Islam to questions in theology (kalam) about the relation
between divine command and human reason. Of the four main Sunni tradi-
tions or schools of jurisprudence (Hanafite, Shafi‘ite, Hanbalite, and Malikite),
al-Ash‘ari comes out of the Shafi‘ite school and al-Maturidi from the Hanafite
school. The last of these is the school that gives the most leeway of the four to
legal reasoning that is not itself derived from the Qur’an and the Traditions.
Al-Maturidi’s full name or identification makes reference to Abu Hanifah (d.
767), the founder of the school, demonstrating the importance of the geneal-
ogy: ‘Abu Mansur Muhammad Ibn Muhammad Ibn Mahmud al-Hanafi al-
Mutakallim al-Maturidi al-Samarqandi’. The Hanbalite school, by contrast, is
the most conservative in terms of the attempt to confine legal reasoning to
what can be derived from the Qur’an and the Traditions.5 Al-Ash‘ari con-
structs his own ‘middle’ position as being between the Hanbalites and the
Mu‘tazilites. The Malikite position will not further concern us, though Ibn
Rushd (Averroes) belonged to this school, and an enquiry into divine com-
mand in Islam of larger scope than the present chapter would have to give it
equal space. To say that al-Maturidi belongs to the Hanafite school is not to
say that he continually invokes Abu Hanifah; in fact, in the Kitab al-Tawhid

4
In fact Kholeif (Tawhid, p. xvii) speculates that ‘Abd al-Jabbar may have been influenced by
al-Maturidi. There are striking similarities in their treatment of taqlid (blind obedience) and the
errors of the non-Islamic sects.
5
This school also denies the necessity of theology (kalam) on the grounds that it is not in the
Qur’an and the Traditions. The current Wahhabi school descends from the Hanbalites.
 Divine Command in Some Medieval Islamic Thinkers 187
(Book on Unity), his major work and our source text here, he refers to Abu
Hanifah only four times. But his place in this school already puts him closer
than al-Ash‘ari to the Mu‘tazilite position on all three questions.
The second general point before we get to the three questions is that some
influential secondary sources associate divine command theory in Islam with
fundamentalism, and oppose it to enlightenment. The project of defending the
Mu‘tazilites within Islam is correspondingly seen as rescuing Islam from
obscurantism and hostility to the modern world. Charles Malik, the Lebanese
lawyer who was a principal author of the Universal Declaration of Human
Rights, is an example of this tendency, writing from outside Islam.6 So is
George Hourani, also writing from outside, and more recently Mariam
al-Attar from inside.7 But so far as the thesis of the present book is correct,
there is no conceptual requirement to connect divine command theory with
fundamentalism, Christian or Muslim or Jewish. The term ‘fundamentalism’
is itself prejudicial here. But, as Chapter 1 argued, divine command can give us
an account of the ground of human dignity in a way that simply making
human dignity ‘a truth of reason’ cannot. As a meta-ethical theory, divine
command theory does not tell us what the commands of God in fact are. But it
gives no ground for inferring that these commands will be any less or any
more liberal than the prescriptions generated by the various versions of
natural law. Having said that, however, it is also true that a theory that has
an honoured place for both revelation and reason will find conversation with
other traditions easier to sustain.

6 . 1. I N T R I N S I C V A L U E

The first question is whether acts and persons have intrinsic value (or whether
that value is to be understood only as a divine willing or commanding), and
what kind of access we have to that value.8

6
See Tim Winter, ‘Response to Christopher Insole’, Studies in Christian Ethics, 25/2 (2012), 222.
7
George Hourani, Reason and Tradition in Islamic Ethics (Cambridge: Cambridge University
Press, 1985). Though he does not speak as a Muslim, this is the advice he gives to Muslims
(p. 276), ‘to start over again at the points where the early jurists and the Mu‘tazilites left off, and
work to develop a system of Islamic law which would openly make use of judgements of equity
and public interest, and a system of ethical theology, which would encourage judgements of right
and wrong by the human mind, without having to look to scripture at every step’. Mariam
al-Attar, Islamic Ethics: Divine Command in Arabo-Islamic Thought (Abingdon: Routledge,
2010), 142: ‘[Divine Command Theory], consciously or unconsciously, provided the philosoph-
ical basis for religious fundamentalism’, and ibid. 167: ‘Divine Command Theory is the most
dangerous theory anyone can defend, because of its possible harmful consequences.’
8
The word ‘intrinsic’ is full of difficulty. Christine Korsgaard, in ‘Two Distinctions in
Goodness’, repr. in Creating the Kingdom of Ends (Cambridge: Cambridge University Press,
188 God’s Command

6.1.1. ‘Abd al-Jabbar

The best place to start is with ‘Abd al-Jabbar’s account of negative valuation.
His term is qabih, and this is hard to translate. It is contrasted with hasan,
which is normally translated ‘good’.9 But qabih has a whole range of senses,
from ‘wrong’ to ‘ugly’.10 Since ‘Abd al-Jabbar is normally talking about acts,
perhaps we should translate his two terms as ‘right’ and ‘wrong’, even though
this will be a strain in certain contexts (noted in the text). There are two
qualifications that need to be made to this proposal. The first is to say that he
does not distinguish between two normative families of terms (the family of
value and the family of obligation) in the way that Chapter 1 urged the strategy
of distinguishing these two families of terms in order to deal with the worry
about arbitrariness. He does, however, quite clearly have an account of
obligation, and an account of right acts that are not obligatory.11 A second
qualification to the proposal is that if we translate qabih as ‘wrong’, we will
have to distinguish it from a narrower term zulm, which is often translated
‘wrongdoing’ but is better translated ‘injustice’, and is defined, as we will see, in
terms of bringing undeserved harm.
Here is how ‘Abd al-Jabbar defines ‘wrong’: ‘What we have established on
the definition of a wrong act all boils down to this: it is the kind of action
that, if it occurs in one way or another as a conscious act of an agent or one
that is merely tolerated by him, he deserves blame unless there is something

1996), 249–74, urges us not to merge the distinction between extrinsic and intrinsic goods and
the distinction between final and instrumental goods. If we follow this advice, this still leaves us
with the question whether a good that is necessarily maintained in existence by the will of God
can be an intrinsic good. G. E. Moore (in Principia Ethica, 187), proposed an isolation test for
intrinsic goods, that we have to consider whether a thing is such that, if it existed by itself in
compete isolation, we should judge its existence to be good. But it is not clear that the necessarily-
God-maintained good could exist in complete isolation, so as to be the object of the required
thought experiment. We should say, rather, that a normative property can be intrinsic even if it is
necessarily given not just its existence but its goodness by God, just as an animal can give its
DNA to another and the recipient will still have that DNA in itself. Chapter 1 defended the idea
that the good that is the individual’s destination is itself both a relation and a kind of
intrinsic good.
9
The topic is often referred to in the nominal form, ‘husn and qubh’.
10
There is very much the same range of senses for the Attic Greek terms kalon and aischron.
I am influenced in my choice here by a private communication from Frank Griffel. Vasalou,
Moral Agents, usually uses ‘good’ and ‘evil’. We will encounter the same difficulty in Chapter 7,
in translating Maimonides.
11
The strategy of distinguishing the two families could be rephrased in ‘Abd al-Jabbar’s own
terminology: God does not make something right by commanding it, but God chooses which
right things to command, and thereby makes them obligatory. But he does not say this. He says
that God can do (though He does not do) what it is obligatory not to do. ‘Abd al-Jabbar also does
not acknowledge a class of normatively negative things that are not blameworthy to commit.
There is such an account in Gregory Mellema, Beyond the Call of Duty: Supererogation,
Obligation, and Offense (Albany, NY: SUNY Press, 1991).
 Divine Command in Some Medieval Islamic Thinkers 189
preventing that particular desert.’12 So a wrong act is one that deserves blame,
but there are two qualifications that need to be made to this. If the agent is, for
example, unconscious or asleep, and so does not know the act will occur and
therefore does not ‘let’ it occur, the agent is not blamed. But the act can still be
the kind that would have been blamed if the agent had had the relevant
knowledge, and in such a case the act is still wrong.13 The second independent
qualification is that there is no ‘restricting’ reason. ‘Abd al-Jabbar’s view is
that the wrong that would otherwise characterize an act can in many cases
be neutralized or overridden by an equal or greater right-making property,
so that it is ‘as if ’ the wrong did not occur. He says: ‘In cases where the act
leads to benefit and injury, comparison of them is necessary. If the injury is
greater, it is as if the act does not lead to benefit . . . If the benefit is greater, it
is as if the injury does not exist.’14 The second qualification is thus that there
is no such neutralizing or overriding right-making property in the act that
deserves blame.15
With this account of ‘wrong’ in place, we can then distinguish four other
kinds of value predication. The contrary of ‘wrong’ is ‘obligatory’, where the
person who omits the act (if he is able to do the act) deserves blame. There are
cases in which there are several alternative ways of avoiding blame, and then
what is blamed is not the omission of any particular one of them, but the
failure to do even one of the alternatives. There are other cases, however, in
which there is no alternative to the act in question for the agent who wants to
avoid blame.16
Distinguished from both wrong and obligation are two other kinds of right,
the ‘merely right’ and the ‘recommended’. Cases of the ‘merely right’ are
breathing the air, or eating harmless food, where the agent does not deserve
praise or blame. These are not, however, simply neutral in value; the breathing
and eating are (non-praiseworthy) right things to do (though in English we
might say more naturally that they are good things to do). Cases of the
recommended are praying or fasting beyond what is required, where the
agent is praised for the act but would not be blamed for the omission. It is
instructive to compare the discussion in Chapter 2 of the types of command

12
‘Abd al-Jabbar, al-Mughni fi Abwab al-Tawhid wal-Adl, ed. I. Madkour et al. (Cairo, 1962–),
VI. i. 26 (henceforth Mughni). The translation of just this sentence is by Frank Griffel.
13
We would more naturally say that the act is still bad.
14
Mughni, XIV. 26–7, discussed by Hourani, Islamic Rationalism, 73–4.
15
Hourani makes the comparison with W. D. Ross’s account of ‘prima facie evils’ (Reason
and Tradition in Islamic Ethics, 104). Ross’s account comes in The Right and the Good (Oxford:
Oxford University Press, 1930), 19–20. But the comparison is in many respects misleading. In
particular, the comparison suggests that ‘Abd al-Jabbar must mean by ‘grounds’ something like
what Ross does. In what follows I will try to show that this is anachronistic and leads to
unnecessary philosophical problems. See Vasalou, Moral Agents, 27–8.
16
The reader may compare Kant’s distinction between perfect and imperfect duties, at
Gl. iv. 421.
190 God’s Command
recognized in the Christian scholastics (precepts, prohibitions, permissions,
counsels, and directly effective commands).
‘Abd al-Jabbar holds that the right and wrong acts distinguished in the
system just described are evident to human reason in their right and wrong
character. They are ‘known immediately’, independently of revelation.17 Reve-
lation does indeed inform us of the obligations we already have, but these
truths are known by reason when they are revealed, and this knowledge by
reason is primary in justification.18 These standards that we learn from reason
apply also to God. ‘The Eternal Glorious One is able to do what would be
wrong if He did it.’19 Because God in fact only commands and does what is
right (though He could do what is wrong), we can use these standards to judge
what God is and is not commanding us to do.
‘Abd al-Jabbar claims that there are ‘aspects’ (wujuh) by which wrong acts
are wrong and right acts are right, and that we can discern these aspects with
our reason. Hourani sometimes translates the Arabic term here by ‘ground’.
The English term ‘ground’, however, is problematic as a translation, and
Hourani has been misled by his sense of the analogy between ‘Abd al-Jabbar
and the work of W. D. Ross, the distinguished British Aristotelian and
intuitionist. For Ross the ground of an evaluative property is the descriptive
property upon which the evaluative property is ‘consequent’, as the goodness
of a strawberry is consequent on its ripeness, redness, juiciness, and so on.
Chapter 1 used the term ‘supervenient’ rather than ‘consequent’. But the list of
‘aspects’ that ‘Abd al-Jabbar gives does not fit at all well with this account. His
list is ‘injustice, uselessness, lying, ingratitude for a favour, ignorance, willing
wrong, commanding wrong, and imposing unattainable obligations’.20 Hour-
ani is left with the task of showing how this list is not viciously circular, which
it would be on his account if the items in it were themselves value-laden,
because they could not then be grounds in the required sense. But this project
is hopeless. Consider just the items ‘willing wrong’, ‘commanding wrong’, and
‘imposing unattainable obligations’. The same is true with ‘injustice’. ‘Abd
al-Jabbar gives, as a short definition of this term, ‘wrong injuries done to another
person’.21 A longer definition is ‘the essential nature of wrong is any injury
without benefit exceeding it or repulsion of harm greater than it, which is not

17
Mughni, VI. i. 18, see also XIV. 152–3. See Section 6.3.1, and Vasalou, Moral Agents, 5.
18
Mughni, VI. i. 47, and XIV. 23.
19
Ibid. VI. i. 127. See also ibid. VI. i. 125: ‘If the status of the agent made a difference to the
ethical value of the act, it would be admissible to say that injustice by prophets or angels or even
God was not wrong.’ In this chapter and the next, but not elsewhere, I refer to God as ‘He’,
masculine and capitalized, since this is the practice in the translations I am using.
20
Ibid. VI. i. 61. The list of right ‘aspects’ is not parallel to the list of wrong ‘aspects’, because
the first consists largely of aspects that can be overridden (such as ‘benefit’), and the second
does not.
21
Ibid. VI. i. 50.
 Divine Command in Some Medieval Islamic Thinkers 191
deserved and not thought to have any of these [right-making] aspects’.22 Here
we need to focus on the terms ‘benefit’, ‘harm’, and ‘deserved’. These are all
evaluatively laden. A ‘benefit’ is a good (bene) received. Perhaps ‘Abd al-Jabbar
can restrict benefits to pleasures and harms to pains, where pleasure and pain
are given value-free accounts (though the project seems unlikely to succeed).
But ‘deserve’ is not going to be susceptible to the same treatment. He says: ‘It is
established by reason that it is characteristic of blame to be corresponding to
wrong and doing mischief, so as to be requital to it, and it is characteristic of
praise to be corresponding to beneficence in the same manner.’23 Hourani says
that ‘the concept of “correspondence” is supposed to be factual in a sense that
is free of value’.24 But he himself realizes that the concept of correspondence
here is almost certainly that of fittingness, an evaluative concept. He has got
himself unnecessarily into this difficulty by trying to force ‘Abd al-Jabbar into
Ross’s mould.
Even the ‘aspect’ of lying may well be value-laden on ‘Abd al-Jabbar’s
account. ‘Lying’ and ‘wrongdoing’ are aspects that necessarily bring wrong
with them, on his account, unlike ‘injury’, which may bring wrong or right
depending on the situation (for example, depending on whether it is deserved).
He distinguishes between the ‘aspect’ of an act (such as those listed above) and
the ‘genus’ (jins) of an act.25 The genus does not make an act wrong.26 ‘Entering
a house’ is a genus of act, as is ‘bowing in prayer’. But entering a house with the
owner’s permission is right (or all right), and without it is wrong. Bowing in
prayer to God is right and to Satan is wrong. But ‘injustice’, he says, is not a
genus of act. This is because injustice is named together with the bad, though
Hourani would disagree.27 But lying is an aspect, not a genus. Hourani is
embarrassed for his author here, because he thinks the Mu‘tazilite’s view is
contradictory. ‘Abd al-Jabbar holds that lying necessarily brings wrong with it,
but he also holds that a small lie may be exempt from blame, on account of the
good past deeds of the speaker and the amount of praise he has earned.28

22 23
Ibid. XIII. 298. Ibid. XIII. 346.
24
Hourani, Reason and Tradition in Islamic Ethics, 102.
25
Hourani translates jins as ‘species’. But the etymology is from the Greek genos, ‘genus’, and
this fits the sense better as well, since the jins is made more specific by the consequences and
circumstances. See al-Attar, Islamic Ethics, 124–5. See also Scotus, Quodlibet q. 18, who makes a
similar point about the act’s relation to its ‘circumstances’.
26
Mughni, VI. i. 59.
27
See NE II. 6. 1107a9–17. See Scotus, Ord. III, suppl. dist. 38, where there is a similar account
of a kind of lying that is ‘named together with the bad’.
28
He discusses at Mughni, VI. ii. 342 Kant’s familiar case of the murderer who comes to the
door, and ‘Abd al-Jabbar bites the bullet: ‘In our opinion it is unsound for them to say that it is
right that a person should tell someone who seeks to murder a believer, and who asks “Is he at
home or not?”, “No, he is not at home” even if it is a lie, to save him from being killed.’ There is,
however, another text, quoted at the beginning of Section 6.3.1, suggesting that it is not necessary
knowledge that lying with the intention of repelling harm is wrong. ‘Abd al-Jabbar and Kant are
in similar difficulties about the case of lying. The claim about small lies is at Mughni, VI. i. 19.
192 God’s Command
Perhaps this point about small lies can be accommodated consistently if we
suppose that he is interested primarily in evaluating agents rather than actions.
There is a kind of lying that is characteristic of a blameworthy kind of person,
and to ascribe this kind to a person is not to describe it in a value-neutral way,
but already to condemn it.
The aspect of injustice is not to be attributed to God’s acts, according to ‘Abd
al-Jabbar, but that is not because there is some difference between aspects as
ascribed to humans and to God. ‘If the status of the agent made a difference to the
ethical value of the act, it would be admissible to say that injustice by prophets
or angels or even God was not wrong.’29 He allows that we might seem to judge
God’s acts differently from our own, when we judge that His goodness is
consistent with causing pain to children; but in fact there is a difference of
circumstances here, because we are assuming that God compensates the children
in the next life, and so in fact the same standard is being applied.30 A key
difference between the three authors we are considering is that they disagree
about whether God could do something wrong, even if He does not in fact do so.
There are two more preliminary matters before we move on to al-Ash‘ari
and al-Maturidi. First, the previous chapters of this book have proceeded as
though there were an affinity between natural law theory and eudaemonism.
One value of studying Islamic medieval moral theology is that we can see
a school where this pairing does not obtain.31 The Mu‘tazilites, and ‘Abd
al-Jabbar in particular, hold that the right in all of its aspects attracts us in
itself, intrinsically, not because it leads to a benefit for us as agents of the
action. Of the two features we have discussed that make an act wrong
whenever they occur—namely, injustice and lying—both are defined in
terms of injuries to another person. Thus injustice is a ‘wrong injury done to
another person’.32 The example is given of Zayd’s injury to ‘Amr, when we
know that ‘Amr does not deserve it, and that Zayd has no right to punish him,
no claim to be acting in self-defence, and no intention to benefit ‘Amr
ultimately.33 ‘Abd al-Jabbar recognizes that his opponents will claim that
people do not avoid injustice and lying intrinsically, but only because of
some benefit to themselves. He replies that people will do wrong for the sake
of some benefit, but they will do right without any benefit to themselves. Even

29
Ibid. VI. i. 125.
30
Ibid. VI. i. 179. See Hourani, Islamic Rationalism, 68.
31
There is an example of the opposite pairing of divine command theory and eudaemonism
in al-Ghazali. See Faith in Divine Unity and Trust in Divine Providence, trans. David Burrell
(Louisville, KY: Fons Vitae, 2001), 62. The texts of al-Ash‘ari suggest a similar pairing to al-
Ghazali’s, but are not completely clear on this.
32
Mughni, VI. i. 50, VI. i. 18. See ibid. XIII. 298: ‘The essential nature of wrong is any injury
without benefit exceeding it or repulsion of harm greater than it, which is not deserved and not
thought to have any of these [right-making] aspects.’
33
Ibid. XIII. 306.
 Divine Command in Some Medieval Islamic Thinkers 193
a heartless man would warn a blind man against falling into a well. If the
objection is made, ‘That is just to prevent his own distress’, ‘Abd al-Jabbar
replies that it is possible to act without thinking about one’s own interest at all.
‘So if it is possible that all the [extrinsic] reasons for which acts are done may
be removed, and he still guides the wanderer, it is necessary to hold that he
guides him for its rightness. . . . For it is impossible to say that he does it for no
reason at all.’34
The second preliminary matter is ‘Abd al-Jabbar’s explicit arguments against
divine command theory. There is no need to go through them all, but it is useful
to see the continuity with the questions within Christian theology we have
already discussed and the questions within Judaism we will discuss in Chapter 7.
Divine command is central to all three Abrahamic faiths, and it is an idea
that creates much the same difficulties in all three. There are at least seven
explicit arguments in ‘Abd al-Jabbar, and they are not equally good. Here
are four, with a brief reply to each in a footnote referring to longer replies
elsewhere in the book. The first is that commands do not imply obligation (as
the divine command theory asserts that they do, at least in the narrow sense of
‘command’ distinguished in Chapter 2). He understands commands as inform-
ing the recipient of the command that the commander wants something done.
Chapter 2 rejected this view on the grounds that it reduces imperatives to an
indicative indicating that someone wants something. ‘Abd al-Jabbar’s point is
that, if we understand commands this way, commanding, even by God, cannot
make something obligatory. If an act is obligatory, he says, a command might
indicate that fact, and it might indicate that the rightness of the act is the
commander’s reason for the command; but ‘it is impossible for [the act] to be
obligatory because of the command, because it is the function of an indication to
disclose the condition of the object indicated, not to put it in that condition.’35
He quotes the Qur’an (16: 90), ‘Surely God bids to justice and good-doing and
giving to kinsmen’, and comments that God is here referring to these things as
real virtues, indicated by the command but not produced by it. This sort of
objection is frequently made by those who cannot see what normativity is added
by a command, even a divine one. Either, they think, the thing commanded is
already right or it is not; the commanding cannot change it from one to the
other, though it can inform us of a character that the act already has.36

34
Ibid. VI. i. 225. Section 3.3.3 discussed Jean Porter’s point that the agent does not have to
think about happiness in order to be guided by it. But ‘Abd al-Jabbar is making the stronger
point that the reason from happiness can be removed. See the discussion of Lawrence Blum in
Section 5.1.3.
35
Mughni, XIV. 22. See Vasalou, Moral Agents, 71–2, where she argues that for ‘Abd al-
Jabbar motivation may frequently be mixed, including both what I called ‘self-indexed’ and ‘non-
self-indexed’ components.
36
The best way to disarm this objection is to point out that normativity is not all alike. ‘Abd
al-Jabbar has a class of right acts that are not obligatory. Accordingly we can say that God does
194 God’s Command
On ‘Abd al-Jabbar’s second objection, the account of wrong as what is
forbidden by God does not fit our normal language. ‘If God were to do wrong
that would be evil of Him, yet we do not say of Him that he was forbidden,’
and, if God were to forbid, for example, showing gratitude to benefactors,
being just in our dealings with others, and acquiring religious belief and
knowledge, all these things would for a divine command theorist become
wrong.37 Here ‘Abd al-Jabbar is committed to the view that, if God were to
do or to command wrong, it would still be wrong, and if He were to forbid
right, it would still be right. In contemporary discussions the example is often
torturing babies, which would still be wrong (the objection goes) even if God
commanded it; so God’s prohibiting something cannot be what makes it
wrong.38
The third objection is that, if divine command theory were right, we could
not know our obligations without knowing that they were commanded by
God; but, ‘Abd al-Jabbar says, ‘the sane man knows his obligation even though
he does not know that . . . there is a commander’.39 ‘Abd al-Jabbar considers
the reply that, by analogy, we can know things exist without knowing that God
created them, even though they do exist by God’s creation. But he responds
that I cannot, according to the account of obligation in divine command
theory, know something as wrong without knowing it as prohibited by God.40
Finally, he objects that the divine command theorist has a problem with
understanding the goodness of God. ‘Acts from the Exalted could not be right,
if rightness in our acts arose only following a command, for commands do not
happen to Him, in the same way as they say that acts from him are not wrong,
because prohibition of Him is impossible.’41 ‘Abd al-Jabbar is pointing here
to what he thinks is an unfortunate consequence of the divine command

not make something right by commanding it, but God does select which right things to make
obligatory for us by commanding them. Also, we need to avoid what Austin calls ‘the descriptive
fallacy’, as described in Section 4.3.1. Some of the Ash‘arites do seem to have made a distinction
between good and right. See al-Attar, Islamic Ethics, 75.
37
Mughni, VI. i. 28, 104.
38
One reply to this is to point again to the distinction between the good and the obligatory.
God has reasons, we may say, from the goodness of what God commands, and we have partial,
but only partial, access to those reasons. This was argued in Section 1.3 and Section 4.1.2. As far
as God has revealed these reasons (which to some extent constrain us in determining what God is
commanding), they do not encompass torturing babies. There will be more to say about this
when we come to al-Maturidi.
39
Mughni, VI. i. 45.
40
Section 4.2 referred to Adams’s point that we can know something is water without
knowing that its molecules are constituted by two hydrogen atoms and one oxygen. ‘Abd al-
Jabbar does not know of the distinction between the meaning of a term and the metaphysical
constitution of the kind (e.g. water) picked out by the term. Even if we drop the language of
‘constituting’, as we should because ‘right’ is not a natural kind, we can keep the distinction
between what a term for a characteristic means and what makes a thing have that characteristic.
41
Mughni, VI. i. 107.
 Divine Command in Some Medieval Islamic Thinkers 195
theorist’s defence of God against the imputation of doing wrong. If, we say,
God’s acts are not wrong because God is not commanded, then we cannot say
God’s acts are right either. But we need, and the Qur’an gives, standards of
value intelligible to us in terms of which we can praise God for doing right.42

6.1.2. Al-Ash‘ari

Al-Ash‘ari’s response to the Mu‘tazilites comes in his Highlights of the Polemic

against Deviators and Innovators (Kitab al-Luma),43 and The Elucidation of
Islam’s Foundation (al-Ibanah).44 He has many complaints against them, and
we will focus on the ones directly relevant to the theme of divine command.
He was himself initially trained by a prominent Mu‘tazilite (al-Jubba’i, d. 915),
and, according to one story, was persuaded to attack them (though not to
abandon the methods of theology) by three dreams in which Mohammed
himself spoke to him and commanded him to defend Islam as it had trad-
itionally been taught. In chapter seven of the Kitab al-Luma al-Ash‘ari dis-
cusses the imputation of justice and injustice to God, and concerns himself in
particular with the question of whether God is unjust in relation to un-
believers, since He wills their perversity. Al-Ash‘ari says:
Whenever He . . . is gracious to some and not to others, and creates men knowing
well that they will disbelieve—all that is justice on His part. And it would not be
wrong on the part of God to create them in the painful punishment and to make it
perpetual. Nor would it be wrong on His part to punish the believers and to
introduce the unbelievers into the Gardens. Our only reason for saying that He
will not do that is that He has informed us that He will punish the unbelievers—
and he cannot lie when he gives information.
Al-Ash‘ari recognizes that what he has just said is controversial, and he
immediately mounts a defence.
The proof that He is free to do whatever He does is that He is the Supreme
Monarch, subject to no one, with no superior over Him who can permit, or
command, or chide, or forbid, or prescribe what He shall do and fix bounds for
Him. This being so, nothing can be wrong on the part of God. For a thing is
wrong on our part only because we transgress the limit and bound set for us and

42
One reply to this point is to say that ‘good’ means ‘attracting us and deserving to attract us’
(where both of these conditions are necessary), and that we can say that God and God’s acts are
the paradigm case of what is good in this sense. This was argued in Section 1.3.
43
Abu al-Hasan al-Ash’ari, The Theology of al-Ash‘ari, trans. Richard J. McCarthy, SJ
(Beyrouth: Imprimerie Catholique, 1953) (henceforth Kitab).
44
Abu al-Hasan al-Ash’ari, Elucidation of Islam’s Foundation (al-Ibanah), American Oriental
Series, vol. 19, trans. Walter C. Klein (New Haven: American Oriental Society, 1940) (henceforth
Elucidation).
196 God’s Command
do what we have no right to do. But since the Creator is subject to no one and
bound by no command, nothing can be wrong on His part.45
Al-Ash‘ari is not here proposing a definition of ‘wrong’. But he is committed to
the view that there is no standard for wrongness among human beings other
than God’s setting a bound or limit for us, and there is no one to set a bound or
limit for God, so that (contrary to Mu‘tazilite doctrine) there is no such thing
as a wrong that God could do.
The objector then asks whether this means that lying is wrong only because
God has declared it to be wrong. Al-Ash‘ari replies: ‘Certainly. And if He
declared it to be right, it would be right; and if He commanded it, no one could
gainsay Him.’46 This does not mean, however, that God can lie. There is a
difference, al-Ash‘ari maintains, between what God can do and what God can
command. Thus God can command us to pray and to be submissive, but this
does not mean that God can pray or be submissive. God cannot lie, but that is
not because it is wrong, but simply because this is not a power God can have.47
It is like the power to be ignorant, which is another power God cannot have.
Al-Ash‘ari holds that our human perception of what is wrong is a recep-
tion of God’s command, and not (as for the Mu‘tazilites) a faculty of reason
independent of revelation. This is a point about Mu‘tazilite moral epistem-
ology, and not their moral ontology. God controls who hears the command
and who does not. Al-Ash‘ari uses the Qur’an extensively to make this point.
He groups the Mu‘tazilites and the Qadarites (who think humans have the
power, qadar, to determine their own acts) together. He replies to them both:
‘It may be said to them: Has not God said: “Their hearts and their ears hath
God sealed up, and over their ears is a covering.”’48 Al-Ash‘ari goes on:
‘God . . . seals the hearts [of the infidels] and locks them against the truth and
hardens them, as when the Prophet of God, Moses, wished ill to his people
and said, “O our Lord, confound their riches and harden their hearts, that
they may not believe until they see the grievous torment.” ’ The divine
initiative here is not just blocking reception but opening it up. Al-Ash‘ari
relates a story about Mohammed: ‘It may also be said [to the Mu‘tazilites]:
God said to His Prophet, “and had We not settled thee, thou hadst well nigh
leaned to them [the unbelievers] a little”’.49 He draws the moral as follows.
Now inform us concerning that settlement—does God do it, or what is like it, to
the infidels? Wherefore, if they say no, they abandon belief in the qadar; but if
they say yes, the answer is: Then, since the settlement keeps the Prophet from

45 46
Tawhid, 169–70. Ibid. 171.
47
Ibid. 172. The reason is that lying would be motivated by weakness or incapacity, which
God does not have.
48
Elucidation, 114. He is quoting Qur’an 2: 6.
49
Elucidation, 115. He is quoting Qur’an 17: 76. The Arabic translated ‘Had We not settled
thee’ can also be translated ‘Had We not made thee stand firm’.
 Divine Command in Some Medieval Islamic Thinkers 197
leaning to them, if God does it to the infidels, they must be settled and kept from
infidelity; but, since they are not dissociated from infidelity, it is simply not true
that He gives them any such settlement as he gave the Prophet—which settlement
kept him from leaning to the infidels.
Al-Ash‘ari here presents a dilemma to the Mu‘tazilites. According to the
Qur’an, knowledge of the command comes with a gift of power to the faithful.
The dilemma is that the Mu‘tazilites have to say whether God gives the infidels
the same sort of gift. If they say no, they are no longer maintaining that we
humans have the power to determine our acts.50 If they say yes, then they have
to say how the ‘settlement’ produces for the Prophet the state of being settled,
but for the infidels it does not produce this. Al-Ash‘ari’s conclusion is that,
since it does not produce this result, this means the divine settlement is not
given to them.
Al-Ash‘ari acknowledges that the Mu‘tazilites may think that they have a
reply, also from the Qur’an. Here is just one example. The Mu‘tazilites may
quote, ‘the Qur’an was sent down as guidance to man, and explanation’,51 and
they may say that this requires that the Qur’an is guidance both to the infidels
and to the faithful. Al-Ash‘ari replies: ‘The answer is: The verse has a particular
meaning, because God has already explained to us that He guides those who
fear Him, and told us that He does not guide the infidels. The Qur’an does not
contradict itself, and so it is necessarily true that His words “guidance to man”
mean the faithful and not the infidels.’52 Al-Ash‘ari here distinguishes between
cases where a verse of the Qur’an has universal reference and where it has
particular reference—that is, reference to some particular group of human
beings. His principle of interpretation is that the Qur’an interprets itself, so
that we can legitimately choose a particular meaning over a universal meaning
if there are texts elsewhere that prohibit the universal meaning. In the case of
the present verse, ‘guidance to man’, if interpreted universally, would contra-
dict verses in which guidance is restricted to the faithful. One of his frequent
refrains against the Mu‘tazilites is that they are not careful about this principle
of interpretation, and pick out verses independently of the sense of the text
as a whole.

6.1.3. Al-Maturidi

We begin with a part of al-Maturidi’s work that sounds very like the Mu‘ta-
zilites, the very opening lines of the Kitab al-Tawhid.53

50
Al-Ash‘ari’s argument here seems to beg the question about whether one could have the
power to act without acting.
51 52
Qur’an 2: 181. Elucidation, 123.
53
The section of Mughni that resembles this passage is XII. 123.
198 God’s Command
Al-Shaikh Abu Mansur [this is al-Maturidi himself], may Allah have mercy on
him, said: furthermore, we find that all people, with all their different religious
opinions and sects, agree on one statement, namely, that whatever one holds to be
true, is valid, and, consequently, that whatever others than him hold, is invalid.
[This comes from the fact] that they all agree that each one of them has his own
predecessors [Salaf] whom he follows. Therefore, it is taken for granted that blind
following [taqlid—that is, following without independent justification] excuses its
embracer from holding the opposite view on the same question. This, however,
only accounts for the multiplicity of the number [that is, the number of his
relevant predecessors]. The only way out of this is if one of them has his ultimate
argument based on reason by way of which his truth can be known and if he has a
demonstrative proof by way of which he can persuade fair-minded people to
accept his truth. Therefore, the one whose source of religion compels the realiza-
tion of this view, is right, and thereupon, each one of them ought to learn the
truth which the former finds in his religion.54
Al-Maturidi lived in a region, Transoxiana, on the eastern borders of the
Muslim world (across the river Oxus), in which Islam was in competition
with all sorts of other religions—for example, Zoroastrianism, Manicheanism,
Daysanism, Marcionism, Judaism, and Christianity, some of which long ante-
dated Islam. He makes in this passage a point about life together with people
with whom one has religious disagreement, and his situation is like our own in
this respect. Each party will hold that its own belief is valid, and its opponents’
beliefs are invalid. Each party, moreover, will have its own tradition, handed
down by authoritative transmission, and will assume that following this
tradition, just because it is the tradition, is legitimate. The only way to get
agreement in such a situation is for one party to have reasoned proof, which
can persuade any fair-minded person. If it does have such proof, the other
parties ought to submit. Al-Maturidi takes this as an argument in favour of
kalam, theology. But this shows that his conception of theology is not confined
to working out the implications of authoritative texts. As we will see in
Section 6.3.3, he thinks that reasoning about our relation to God can have
merit (though subordinate merit) independent of special revelation.
Later in the work he goes further:
If one says: ‘If it is permissible that God command to man that which he does not
understand by his reason, why is it not permissible, then, that He speak to him
that which he does not understand either?’, it is to be said to him: there is no
difference between these two, and, so, it is not right to treat them the way you
have mentioned; there is nothing that God would command to man unless He
caused his reason to understand it, and likewise, there is nothing that God would

54
Tawhid, 3–4, translation by Fathalla Kholeif. See Roots, 67–9. The idea of ‘demonstrative
proof ’ should not be taken in a strict Aristotelian sense (apodeixis), but in the sense of a
compelling argument.
 Divine Command in Some Medieval Islamic Thinkers 199
speak to man unless he also caused the way through which he can understand it.
Therefore, if man is short of understanding of the bearing of the command, he is
excused from it.55
Here al-Maturidi acknowledges that God gives to human reason an under-
standing both of the divine speech in general in the Qur’an, and of divine
commands in particular. If God did not give this understanding, he says,
humans would be excused from complying with the commands. This is a
point we can find also in ‘Abd al-Jabbar, where it is deduced from God’s
justice. But this agreement of al-Maturidi with the Mu‘tazilites needs to be
heavily qualified. Al-Maturidi also holds that we very often do not know
whether something is wise or foolish, just or unjust. The central Mu‘tazilite
error, he thinks, is to suppose that God’s actions are like human actions.56
Al-Maturidi does not deny the Mu‘tazilite claim that God has a reason for the
divine command, but he does deny that we always have access to it, even in
principle.57 How can we hold these two parts of al-Maturidi’s view together,
that God causes our reason to understand His commands, and that very often
we do not know God’s reason?
There is a key text that gives us a way to hold these ideas together. Al-
Maturidi says:
God created people responsible. God made people [such that they] know how to
divide right and wrong. God made wrong action ugly to our rational understand-
ing. He made right action beautiful to our rational understanding. He put the
action of choosing the beautiful rather than the ugly in people’s understanding.
He put the action of preferring something that deserves praise rather than
something that deserves blame. For that reason God calls people to prefer one
action rather than another as being fitting to the features that have been put into
their nature, and the good features that have been given to them. And God made
it ugly to act otherwise, so that conscious living things cannot accept this
rationally. God determined everything that surrounds human beings to be either
harmful, i.e. to be avoided, or beneficial, i.e. to be desired, so that it would be clear
that the results of action are to be desired or avoided. God created human beings
with features that make us hate some things and make us tend toward some
things. Moreover He made some actions beautiful to our reason. One set of
actions human beings can hate and another set will give good results. But He
made ugly some actions, because of their bad results, that human beings tend
towards by nature. Therefore He made people with a nature that enables them to
stand things that are hard for their nature for the sake of their results which give
pleasure, and enables them to bear suffering for the same goal. Because of the fact

55
Tawhid, 137. See Roots, 69–71.
56
Tawhid, 343. The Mu‘tazilites hold, he says, to the principle of al aslah, that even God has
to act in a way that is for the maximum benefit of His creatures.
57
See Theology, ch. 3: ‘Al-Maturidi emphasizes . . . that the innermost being [kunh] of divine
wisdom is not conceivable to us.’
200 God’s Command
that human reason will avoid bearing difficulties, God tests responsible people,
and for that reason He encourages people to right actions and right morals and he
commands us to prefer legal actions and to avoid illegal actions.58
Al-Maturidi here gives us a composite picture of human nature. We have both
a rational understanding that responds with attraction to the right and with
repulsion to the wrong, and we have a tendency towards what is bad in its
results. Both are properly described as belonging to our nature. He is not here
simply talking about an aversion to physical pain, but a tendency in our reason
to avoid bearing difficulty and to prefer illegal actions. This is a key point. Like
the Mu‘tazilites, al-Maturidi can affirm that God gives us in creation a rational
understanding, which responds to the right. He even repeats the Mu‘tazilite
account of the right as what deserves praise and the wrong as what deserves
blame. But this does not mean that our actual decision-making about what to
do accurately tracks what is in fact right and wrong. To the contrary, we tend
towards what is in fact, in its results, wrong, because our human reason avoids
bearing difficulty. This is why we need testing, and why God gives us com-
mands and encouragement, to counteract this tendency. When al-Maturidi
says that God causes us to understand His commands, he is referring to God’s
creation in us of the rational understanding that is attracted to the right and
repelled from the wrong; but when he says that very often we do not know
God’s reason, one explanation is our natural tendency to avoid bearing
difficulty.
An example he gives of this deplorable natural tendency is that we do not
like taking bad-tasting medicine. This example allows us to see how he is
making a distinction like the Mu‘tazilite distinction between genus and aspect.
Al-Maturidi says:
As a conclusion of these things, it is possible to say of something that it is both
injustice and justice. I say again that the main problem for distinguishing these
two different concepts is the ignorance of human beings. So far we have shown
that in the same thing there can be wisdom and folly. But human perception and
human understanding may not understand this. For a person’s judgement that
something is wise or foolish, just or unjust, on the grounds of his own thinking
system, may not always be true. It is clear that it is impossible for people to
understand two contradictory properties in one thing; this is not the kind of
information that people can get through their perception. We show that it has
been proved that the Mu‘tazilites do not understand that there are some cases of
folly in the creation of both beneficial and harmful objects, because in one respect
a thing can be harmful and in another beneficial, for example bad-tasting
medicine. The Mu‘tazilites’ idea is that if an action does not provide benefit for
another, it does not have wisdom, but I prove that this is not true. There is no
harmful thing from which it is impossible for a person to derive benefit. This

58
Tawhid, 351.
 Divine Command in Some Medieval Islamic Thinkers 201
benefit could be that it gives warning, or it is an instrument for taking a lesson, or
for remembering God’s blessings, or to avoid something bad happening. There
are many possibilities here, but it would take a long time [to go through them].59
In this passage al-Maturidi points to the same range of phenomena that we
found described by ‘Abd al-Jabbar in terms of the genus of an action. The
Mu‘tazilite holds that the genus does not make a thing wrong.60 ‘Entering a
house’ is a genus of act, as is ‘bowing in prayer’. But entering a house with the
owner’s permission is right (or all right), and without it is wrong. Bowing in
prayer to God is right and to Satan is wrong. ‘Abd al-Jabbar is saying that
when we remove the owner’s permission, we add the aspect of injustice,
though ‘injustice’ adds something to ‘wrong’ (since injustice is wrongful
injury) and is not simply the same thing. But al-Maturidi analyses the phe-
nomenon differently. Of the same thing, he says, we can predicate benefit and
harm, justice and injustice, wisdom and folly. The medicine tastes unpleasant,
and drinking it is so far folly, but drinking it is also conducive to health, and so
far wisdom. When we add the aspect of justice, or wisdom, we are already in
the sphere of the right, and the right still needs an account; this account, for al-
Maturidi is in terms of God’s decree. ‘If, then, the beauty of wisdom and justice
is established as a general principle as well as the ugliness of foolishness and
injustice, God must be described with every and each action He creates by
wisdom and justice or grace and righteousness because it has been established
that He is good, generous, self-sufficient and knowing.’61
There is a similar point in a later Ash‘arite, al-Juwayni (d. 1085).
Or else it is argued [by the Mu‘tazilites]: ‘A thing is wrong only by something
other than scripture and other than wrong [itself].’ Then if they say that, one
answers them: If the thing is not wrong by itself, and its wrongness is not related
to its connection with [divine] prohibition, then it is impossible that an attribute
should be wrong because of another attribute [i.e. the ‘aspect’], unless that
[second] attribute is an attribute for wrong either in its essence or as a qualifying
attribute.62

59 60
Ibid. 346. Mughni, VI. i. 59.
61
Tawhid, 217 (emphasis added), discussed in Roots, 218–19. We will see in Section 6.2.3 that
this divine decree will be either the ‘absolute decree’ or the ‘detailed decree’, in the distinction to
be drawn there. Al-Maturidi continues: ‘Because there is no harmful action at all from which one
cannot receive benefit either through the way of guidance or admonition, or that there cannot be
in it the reminder of benefaction or the warning for retribution, and that it might not lead to the
knowledge of the one to whom the creation and the command belong, and many other things
which would take us long to mention.’ Note that all these possibilities are in relation to God, so
that ‘God must be described with every and each act He creates.’ See Theology, ch. 3: ‘[God] holds
Himself to norms that He has conclusively established.’
62
Abi’l Ma’ali al-Juwayni, Kitab al-Irshad, ed. M. Y. Musa and A. A. Abd al-Hamid (Cairo,
1950), 267. This passage is discussed in Hourani, Reason and Tradition in Islamic Ethics, 132–3,
but he takes the reference of the final clause to be to the first attribute, whereas it seems to me
more natural to take it as referring to the second, the relevant ‘aspect’.
202 God’s Command
This is an obscure passage. But the Mu‘tazilites are described as holding that
what makes a thing wrong is not Scripture but what I have called the ‘aspect’,
for example ‘injustice’, which is not simply the same as wrong itself. Al-
Juwayni replies that if the thing is only made wrong by its aspect (‘another
attribute’), and it is not wrong because of God’s prohibition, then it cannot be
made wrong by the aspect unless that aspect is itself wrong, either (using
Aristotle’s logical language) in its essence or as a quality. In other words, the
aspect ‘injustice’ cannot make an act wrong unless ‘injustice’ is already named
together with the wrong. But, if it is already wrong, then it is divinely
prohibited, according to the divine command theorists we are considering in
this chapter. To say that the action is made wrong by the aspect and that
therefore it is not made wrong by God’s prohibition, as the Mu‘tazilites do, is
simply to beg the question.
Al-Maturidi sums up his views about right and wrong as follows:
Our opponents [the Mu‘tazilites] can say that ‘actions have right and wrong in
themselves’. If the right and wrong of an action happen in themselves, we should
know that God is more helpful for every being than it is for itself, because every
being is ignorant of its own essence. If the existence of right and wrong without
any creator were possible, it would be possible to have existence without a creator.
In this statement we would be out of Islam. If these actions were determined in
themselves as regards the attributes of right and wrong (even granting that [the
agents of] these actions did not know their essence and limits), in this situation it
would be impossible that [the agent of] an action would be ignorant of the right
and wrong in it. [But the agent of an action can be ignorant of the right and
wrong in it.] Therefore the actions of servants are not determined in themselves
to right and wrong.63
Al-Maturidi here considers whether we can talk about an action having right
and wrong in itself. As we saw earlier, we can talk about God having a reason
for the divine commands. But the existence of this right and wrong is not
independent of God. What is right for a thing, he says, is related to the essence
God creates, and no created thing knows its own essence. We will say more, in
Section 6.2.3, about al-Maturidi’s views of the divine decree. But, for now, the
important point is that the rightness and wrongness of an action depend upon
‘the limit and bound set for us’, in al-Ash‘ari’s language, a limit and bound to
which we do not have reliable access, and which is continually maintained by
God’s will. In terms of the distinctions made in Chapter 2, God’s directly
effective commands create and sustain human life and God’s precepts and
prohibitions and permissions and counsels are constrained by fitting this kind
of life (though they cannot be deduced from it).64 In this sense our actions are

63
Tawhid, 358.
64
See Theology ch. 3: ‘Indicators of [God’s] wisdom are in fact found everywhere in the world.
God did not hide His decisions, but actually imparted them in a form understandable to all
 Divine Command in Some Medieval Islamic Thinkers 203
not determined in themselves to right and wrong (if that means, independ-
ently of divine command and prohibition). How much play, so to speak, there
is between the human essence and the divine command, or how much
discretion (arbitrium) God has in the command, is not a question al-Maturidi
addresses here.
The Mu‘tazilites might object that al-Maturidi, by denying that our actions
are right or wrong ‘in themselves’, has denied the objectivity of morality. This
term ‘objectivity’ is anachronistic, not found in any of the authors we are
discussing in this chapter. But we can speculate that he would think it is the
Mu‘tazilites who are most liable to the charge, not the divine command
theorists. The Mu‘tazilites think we know by reason what is right and wrong
in itself, though they do not think we make something right and wrong by our
judgement of it. But it is helpful to remember here the discussion of deducti-
vism in Chapter 4, and in particular the discussion of Robert M. Adams in
Section 4.2. Adams holds that we human beings cannot be deeply wrong about
good and evil, because we fix the reference of the terms by our ordinary use of
them, just as we fix the reference of ‘water’. Like the Mu‘tazilites, he would
deny that we make things good and bad in this way (since, on his theory, they
are made good and bad by the degree of their resemblance to God). But the
reply to Adams is that we should be more modest about our abilities, holding
with al-Maturidi that we have by nature a tendency towards the wrong as well
as a tendency towards the right, and we should not ‘compare God’s actions
with people’s actions’. He also says: ‘The important point is that every human
governor in the perceptible world is a candidate for doing something wrong.’65
The Mu‘tazilites are liable to the same objection as Adams. Holding that what
we judge by reason has the role they assign in justifying a claim that something
is right and wrong denies the full objectivity of morality.66

6.2. HUMAN FREEDOM

We now take up the same three figures, but in relation to the different question
whether human beings have freedom of choice in what they do, or whether

humans. This is evident on numerous levels: in the harmonious direction [tadbir] of the creation;
in the rationality of ethical norms; and even in the way in which God creates harmful life forms
and substances [al-hayyat wa-l-jawahir al-darra) for specific reasons.’
65
Ibid. 348 (emphasis added).
66
The same would be true if we restricted the scope to Muslims, and said, with the Tradition,
‘Whatever believers see as good is good with God’, attributed to al-Shaybani, and discussed by
Hourani, Reason and Tradition in Islamic Ethics, 195–6. The disagreement with the Mu‘tazilites
is not because they held that humans can in fact know the reasons for God’s commands, for al-
Maturidi also thinks we sometimes know this at least to some degree. See Mughni, VI. i. 64.
204 God’s Command
our actions are only the product of divine causation. This question is the
subject of prolonged discussion by all three, but we will focus on material that
has implications for the relation between divine command and human
obligation.

6.2.1. ‘Abd al-Jabbar

‘Abd al-Jabbar starts from the premiss that ‘it is irrational to assign an
obligation to perform an act, unless the addressee is capable, or has the
power to perform it, in order to be considered truly his action’.67 He compares
asking a boy to ascend to the rooftop without giving him a ladder, which is as
unreasonable as asking a bird with broken wings to fly.68 The maxim that it is
bad or irrational to impose unbearable obligations is taken from the Qur’an.69
But now we need to consider what kind of power we are talking about. Two
things are important to say about it: it has to precede the act and it has to be a
power over opposites—that is, ‘a power to perform an action or its opposite’.70
If the power were only at the time of acting, and only a power for the action
performed and not for its opposite, the creator of the power should be
considered the determining cause of the action. ‘Abd al-Jabbar does not see
how we could then avoid the conclusion that, if the action is wrong, the creator
of that power would be the creator of wrong. He is content to say that God
knows of some humans that they will disobey His commands: ‘We have shown
in the book on the will that the Exalted has willed of all who are under
obligation obedience and belief, and that He knows that some of them will
disobey and not do what is willed of them.’71 But he is not content to say that
God creates or wills that wrong, even though He does know about it; rather, it
is the humans who do the wrong that create it. God does have the power to do
wrong, but it is impious to think He does it, and there is no reason to think He
does it.72
‘Abd al-Jabbar uses a distinction here that descends from Aristotle’s dis-
cussion of the ‘mixed’ cases of voluntary action in the Nicomachean Ethics,
which was available in Arabic, though he reaches a slightly different conclu-
sion from Aristotle about praise and blame.73 Aristotle holds that an action is
involuntary (akon) if it is done either by force or by ignorance, and it is done
by force if the origin of the action is outside the agent—for example, ‘if he were
to be carried somewhere by the wind, or by men who had him in their power’.

67 68
Mughni, XI. 371. See al-Attar, Islamic Ethics, 88–9.
69
Mughni, XI. 129, 367, 391. See al-Attar, Islamic Ethics, 89.
70
Mughni, XI. 168. See NE III. 5. 1113b8: ‘For where it is in our power to act, it is also in our
power not to act.’
71 72 73
Mughni, XI. 160. Ibid. VI. i. 177 ff. NE III. 1. 1110a1–29.
 Divine Command in Some Medieval Islamic Thinkers 205
But there are three kinds of mixed cases of ‘force’. One is where the action is
done from fear of greater evils—for example, ‘if a tyrant were to order one to
do something base, having one’s parents and children in his power, and if one
did the action they were to be saved, but otherwise they would be put to death’.
In such cases, one is praised, he says, if one accepts something base or painful
in return for great and noble objects gained. These cases are ‘mixed’ because
they resemble both the voluntary and the involuntary, but Aristotle says they
are more like the voluntary. The second case is where one receives not praise
or blame, but pardon, when ‘one does what he ought not under pressure which
overstrains human nature, and which no one could withstand’. The third case
is where the action is so base that no one could be forced to do it—for example,
matricide. Finally, Aristotle says that we are not forced by what is pleasant and
noble, for then all acts would be forced; but those who act by force act
with pain.
‘Abd al-Jabbar likewise distinguishes causal constraint from force majeure:
‘Something might appear to a person-with-power (al-qadir) that implies that
he has to do a certain action, yet that does not rule out the fact of his being
able, in contrast to the causally constrained [person], to whom something was
done that he cannot avoid.’74 For example, a person might be told that there is
a lion on the road, and he believes it is true because of reliable evidence, and he
is compelled to avoid being killed by choosing a different road.75 The Mu‘ta-
zilite’s conclusion about this kind of mixed case is that a person is not to be
praised (or blamed) for such an action, just as a person is not to be praised or
blamed for cases of causal constraint. He also refuses to allow that we can be
forced to do what is wrong in itself—for example, in the two cases of injustice
and lying. He here extends Aristotle’s treatment of the third kind of case (for
example, matricide, where we cannot be compelled to do it) to cover all actions
wrong in themselves (a category Aristotle does not have). He agrees with
Aristotle’s assessment that in some mixed cases we do not receive praise and
blame, but he says this not about cases of pressure that overstrains human
nature (where Aristotle says we receive pardon), but about all cases where we
are motivated by self-preservation. Again, this is because he has a category
Aristotle does not have, that of actions to benefit others without reference to
oneself, which do deserve praise. Finally, he reflects Aristotle’s point about the
pleasant and the noble (which for Aristotle are ingredients of the agent’s own
eudaimonia), but he says, not that we cannot be compelled by them, but that
we should not be praised for pursuing them as our own advantage. Each of

74
Mughni, VIII. 166.
75
Ibid. XI. 397. See al-Attar, Islamic Ethics, 91–4. However, she points out that ‘Abd al-Jabbar
gives examples only where we are forced or compelled to do acts that are good and beneficial for
ourselves or our kin, and she concludes that this is why he does not consider cases where we are
to be blamed for what we choose to do according to a strong motive.
206 God’s Command
these three changes to Aristotle is highly illuminating about the structure of
the Mu‘tazilite’s thought as a whole, which denies eudaemonism and embraces
the view that we can be moved by what is good in itself, independent of our
own advantage.
‘Abd al-Jabbar has a complex picture of desire, motivation, and will. The
central point for our purposes is that he is concerned to deny that there is any
determining cause of our actions, either external or internal. This is perhaps
because he thinks that, if there were a determining cause, even an internal one
like a strong motive, a person’s action would not be, in the phrase quoted at
the beginning of this subsection, ‘truly considered his action’. He does not
have, just as Aristotle does not have, a Kantian sense of ‘will’, in which it is the
centre of agency. If he had thought in the Kantian way, he might not have been
so reluctant to posit an internal determining cause. But his notion, though
rendered ‘will’, is closer to wanting than what Kant would call ‘willing’. He
allows, for example, that a human being can ‘will’ that another human being
does something, and this is something Kant does not allow. One final point is
that ‘Abd al-Jabbar holds that it is obvious that we have the relevant kind of
power over our actions (a power that precedes the act, and that is a power both
to act and not to act). ‘Any intelligent person does not deny the fact of his
aiming at the act and willing it and choosing. He distinguishes between this
state of his and his rejecting.’76 In this way the Mu‘tazilite resembles Scotus,
and the resemblance is a deep one; the power over opposites is something we
know from ordinary experience.

6.2.2. Al-Ash‘ari

Of al-Ash‘ari’s ten objections to the Mu‘tazilites, listed at the beginning of

Elucidation, three have to do with the matters discussed in the present section
of the chapter: The Mu‘tazilites assert that human beings create evil, they think
that God may wish what is not and that what He does not wish may be, and
they think that they alone, and not their Lord, have power over their works. 77
Al-Ash‘ari responds that ‘he who does not will the existence of anything
except what exists, and nothing exists except what he wills, and nothing is
remote from his will, is the worthier of the attribute of divinity’.78 And he
objects that, ‘if there are under His authority things the existence of which He
disapproves, you cannot deny that there are under His authority things the
existence of which He forbids. . . . And this is an attribute of weakness and
poverty.’

76 77 78
Mughni, VI. Ii .8. Elucidation, 47–8. Ibid. 101–2. See Kitab, 50.
 Divine Command in Some Medieval Islamic Thinkers 207
Perhaps an analogy will be helpful here. Suppose we think of the whole of
the universe under God’s providence as a circle suffused by light, of which our
own lives are a proper, but very small, part. If we were to think of human
freedom as creating evil, this would open up, so to speak, a fissure in the circle.
But then we could not attribute the whole to God’s good care, and this would
mean that we could not trust God to be in control of our own lives, in all their
detail. God would be, in this sense, weak in comparison with the sovereign
omnipotent God of the tradition. To someone who holds such a picture, the
Mu‘tazilite view would be deeply distressing.
It is tempting to think of al-Ash‘ari as privileging God’s omnipotence over
God’s justice, but this is not how he sees it. He is completely convinced of
God’s justice (though he thinks we have to be careful not to think it is the same
thing as human justice), and moreover he is convinced that we are responsible
for our acts, and God rightly holds us responsible. To understand this, we need
to describe his notion of ‘acquisition’. This is the view that a single act can both
be created by God and ‘acquired’ or performed by a human being. Al-Ash‘ari
gives the case of Joseph and Potiphar’s wife. Potiphar willed disobedience (to
God) in imprisoning Joseph, and Joseph also willed to be imprisoned, in
resisting Potiphar’s wife; but Joseph did not thereby will disobedience (to
God).79 Al-Ash‘ari also gives the example of someone who comes to unbelief.
This act has both a producer, who produces it as it really is—namely, vain and
bad—and this is God. But it also has an acquirer, the person who comes to
unbelief, but thinks of this as good, right, and true.80 Al-Ash‘ari helps us
understand this distinction by comparing the case of God making something
move. God does this by giving the thing the power to move, at the moment of
its moving. God makes it as it really is—namely, moving. But this does not
mean that it is God who changes place; that change is made by the thing that
acquires or performs the motion by acquiring the created power for it.81 We
can think of a person who lies in the same way.82 We can distinguish between
the one who lies, who is not the one who makes the act as it really is, and the
one who makes it as it really is (namely, God) who does not lie. We can now
make the distinction familiar from experience between cases of causal con-
straint and cases where we have the power to act, and so responsibility for our
action. Al-Ash‘ari calls these two cases ‘necessary motion’ and ‘acquired
motion’, and he gives the examples of shaking from palsy or shivering from
fever, for the first case, and coming and going or approaching and receding,
for the second.83
One knows how to distinguish between the two states, in himself and in others, by
a necessary knowledge which leaves no room for doubt. So if there be impotence

79 80 81
Elucidation, 104. Kitab, 86. Ibid. 89.
82 83
Ibid. 90. Ibid. 92.
208 God’s Command
in one of the two states, power, which is its contrary, must exist in the other. For if
impotence existed in both states together, the man’s way of acting would be the
same in both. Since this is not so, and since there is power in one of the motions,
this motion must be an acquisition—because the true meaning of acquisition is
that the thing proceeds from its acquirer in virtue of a created power.
We can now ask al-Ash‘ari whether God creates evil (or wrong). The answer
is not straightforward. The interlocutor asks, ‘Has not God, then, created the
injustice of creatures?’, and al-Ash‘ari replies, ‘He created it as their injustice,
not as His.’84 The interlocutor objects, ‘Then why do you deny that he is
unjust?’, and al-Ash‘ari replies, ‘One who is unjust is not unjust because he
makes injustice as another’s injustice and not as his.’ The same reply comes
with the question about whether God creates evil and whether God creates
lying.85 God creates evil for another, and lying for another, but God cannot
Himself do evil, or lie. Does this mean that God has decreed and determined
acts of disobedience?86 Here al-Ash‘ari makes another distinction, between
decreeing and determining in the sense of producing something and decreeing
and determining in the sense of commanding it: ‘We do not say that God has
decreed and determined acts of disobedience in the sense that He has com-
manded them.’ This is the difference we identified in Chapter 2 between two
different kinds of prescriptions—namely, ‘precepts’ (or ‘prohibitions’) and
‘directly effective commands’.
Someone might worry about God’s commanding things when God does not
provide the recipients of the command with the power to carry it out. The
interlocutor asks, ‘Has not God charged the unbeliever with the duty of
believing?’, and al-Ash‘ari answers that He has.87 But this does not mean
that God has given the unbeliever the power of believing, because, if God had
given that power, the unbeliever would believe. It seems to follow that God
enjoins on him an obligation that he cannot fulfil. Here al-Ash‘ari makes
another distinction. Strictly, an inability is an inability both for some act and
for its contrary. A stone has the inability to believe, because this inability is also
an inability to disbelieve. But the unbeliever has the ability to disbelieve, and so
does not strictly have the inability to believe. Al-Ash‘ari considers an objection
to this account of inability—namely, that he has denied that a power is for
an act and its contrary.88 How can he deny this of powers and affirm it of
inabilities? The reason is that, on al-Ash‘ari’s conception of power and inabil-
ity, they are necessarily concurrent with their exercise. The exercise of the
inability both for the act and the contrary (to believe and to disbelieve) makes
sense (as in the case of the stone). But the exercise of the power both for the act
and the contrary does not make sense; it would require a thing to have two
contrary attributes at the same time. It is true that we still need a defence of the

84 85 86
Ibid. 97. Ibid. 107, 119. Ibid. 101.
87 88
Ibid. 135. Ibid. 136.
 Divine Command in Some Medieval Islamic Thinkers 209
account of a power as necessarily concurrent with its exercise. Al-Ash‘ari’s
opponent may reverse the argument, and say that it is obvious we have the
power both to act and not to act, and therefore a power cannot be necessarily
concurrent with its exercise. We will pursue this dialectic further when we take
up al-Maturidi’s contribution to it.

6.2.3. Al-Maturidi

Al-Maturidi’s complex views on these matters are best understood through

three distinctions that he makes, distinctions between two kinds of power, two
kinds of divine attitude, and two kinds of divine decree. Here is a passage
about the two kinds of power:
We hold that . . . power is of two kinds. The first kind concerns the soundness of
purposes and the correctness of instruments that precede actions.89 This power is
not realized in actions, although these actions are not performed except through
this power. It is rather a blessing from God which He has bestowed upon whom
He has wished. Then, because of it, God requires that they praise Him as they
become able to acquire the blessing and to reflect on it. This is a right statement
with respect to rational faculties. That is to say, the receiver of the bestowal should
offer his gratitude to the bestower, should know the reality of the blessing, and
should know how to refrain from denial of the bestower and from ignorance
about the blessing. If this were not so, it would be impossible from the beginning
that there be any command or prohibition. For, without [this power] preceding
[the command and prohibition] in reason, there could be no necessary reason for
gratitude and for abstaining from denial [of the blessing]. The second [kind of
power] is a notion which can only be explained by a definition that makes known
its nature as nothing more than what is assigned to an action. It is not possible
that this power exists in any way other than by performing an action when that
action is actually performed.90
In this passage the second kind of power is relatively straightforward, after we
have understood the description of al-Ash‘ari’s notion of power; this power is
not definable except as the power to perform the act at the time of the act. Al-
Maturidi sees, however, that acknowledging this kind of power is consistent
with acknowledging a different kind of power, the first kind. What is this first
kind? Al-Maturidi is here in the tradition of Abu Hanifah, whom he does not
often cite, but significantly does cite at this point. The first kind of power is the
human capacity to act in two opposite ways, ‘which is the opinion of Abu

89
Kholeif (Tawhid, p. xxxv) translates ‘soundness of organs’ rather than ‘soundness of
purposes’. But he later concedes (ibid., p. xxxvi) that ‘intention and adequacy of means’ (the
same pair) have to go together as a precondition for the second kind of power.
90
Tawhid, 256.
210 God’s Command
Hanifah and his group of followers’.91 God gives us commands and prohib-
itions, and we have the power to receive them and to act upon them. The
commands and prohibitions give us the right purposes and instruments. The
power to receive them precedes the action. However, we also have the power
to deny the bestower of these blessings. Indeed, we have a tendency to do so
(more than a mere power). I have already quoted one of the passages that
make this clear:
Therefore He made people with a nature that enables them to stand things that
are hard for their nature for the sake of the results which give pleasure, and
enables them to bear suffering for the same goal. Because of the fact that human
reason will avoid bearing difficulties, Allah tests responsible people, and for that
reason he encourages people to good action and good morals and he commands
us to prefer legal actions and avoid illegal actions.92
Al-Maturidi continues that God makes us responsible for things that are hard
and easy, steep and level, and gives us principles by which to attain every
virtue. ‘The aim of this is to make clear that the height that people will attain
depends on the degree of their bearing things that human beings by nature do
not like and the souls of human beings hate and it depends on the hard work
of people who use their reason.’ By necessary knowledge, he holds, everyone
knows that he is the one who chooses to do what he is doing, even though the
theological determinists deny this.93
The picture of the two powers we are given is that the first power precedes
the act, and it is a power to choose, and the second power performs the act and
is concurrent with the action. Both powers are the gift of God. The action that
is taken by the second power must be the action that is chosen by the first
power, since al-Maturidi says the action is performed ‘through’ the choice
made by the first power.94
This brings us to the second distinction, between two kinds of divine
attitude. Al-Maturidi quotes the Mu‘tazilite al-Ka‘bi, who raised the problem:
‘If the escaping of a thing from [God’s] knowledge would entail ignorance,
why then should not its escaping from His will be considered a defect, which
means inability?’ Al-Maturidi responds:
[God’s] dislike is with respect to prohibition, while it is compulsion that means
defect. There is also in the Book of God the proof for the distinction between
satisfaction [mahabbah or rida] and will [iradah or mashi‘ah]. For example,
‘[God] is not satisfied with infidelity’ (39: 7), ‘God does not like corruption’
(2: 205), ‘Indeed, God likes those who repent’ (2: 222), and ‘God does not like

91 92 93
Ibid. 263. Ibid. 351 (emphasis added). Ibid. 296–7. See Roots, 211.
94
There is a hard question here whether the second kind of power (that is exercised in the act)
is exercised also in the act of choosing. There is no indication in the text that this is so. But then
there are familiar philosophical and theological difficulties about how to justify such an
exclusion.
 Divine Command in Some Medieval Islamic Thinkers 211
offenders’ (2: 195). And He said regarding will, ‘God lets go astray whom He
wishes, and He lets go on a straight path whom He wishes’ (6: 38). These and
other verses require the distinction between satisfaction in particular and will in
general [iradah or mashi‘ah]. By these last two God is being described and not by
satisfaction, because will assumes power in the sense of its having the decree of
force. . . . The reason for this is that satisfaction [mahabbah] and disliking [sakht]
are two notions that are applied to human actions, while will [mashi‘ah] is not.
This is because of the fact that in human actions there is no such notion [of
mashi‘ah], except in the sense of satisfaction [rida] and wish [tamanni].95
The distinction in divine attitudes al-Maturidi wants here is already suggested
by al-Ash‘ari’s distinction between decreeing and determining in the sense of
producing something and decreeing and determining in the sense of com-
manding it. But al-Ash‘ari resists the implication that there exist things of
which God disapproves, because he does not want to attribute weakness to
God. Faced with the quotes from the Qur’an about God’s not liking offenders,
and so on, al-Ash‘ari does not want to say that these things are outside God’s
will, even though they are contrary to his command. He therefore understands
God’s ‘not liking a thing’ in terms of an external communication in a divine
prohibition of that thing. Al-Maturidi gives us a way to take the distinction
inside God’s will, without losing God’s global providential control.96 This
solution is expressed in the sentence: ‘These and other verses require the
distinction between satisfaction in particular and will in general.’ But this is
not intelligible until we have described the third distinction al-Maturidi
makes, the distinction between two kinds of divine decree.
He says:
The first is the definition with which things come into existence. It has made all
things have whatever goodness or badness, nobility or baseness, wisdom or
foolishness, they possess. It is the explanation [or: interpretation] of wisdom
that everything is made to be as it is [that is, made to possess the qualities that it
does], and that which [or: he who] is most worthy of anything hits upon it exactly.
And in something like this sense God has said: ‘Surely we have created everything
by a decree [qadar].’97
About the second kind of decree [qada], he says (three paragraphs later): ‘Nor
with regard to the second [kind] is it possible for human beings to determine
[taqdir] their actions with respect to time and place, nor does their knowledge
attain this. And so in this respect, too, it is not possible for it to be by them,
such that their actions do not come to be from God.’ Here is a summary of the
whole three-page passage.

95
Tawhid, 296–7.
96
It is also worth comparing the distinction discussed in Chapter 2 between antecedent and
consequent will.
97
Tawhid, 307, translation by Geoffrey Mosely.
212 God’s Command
For, the creation of actions affirms the divine absolute decree [qada] for the
coming of things into existence as well as His detailed decree [qadar] as to their
being good or evil. In fact, al-Maturidi explains, the divine decree [qada] is the
valuation [hukm] of a thing and the definition [qat] of its respective properties.
This definition of the proper properties of an object is good for that object
because it comes from the one who is wise and knowing.98
What is the difference between these two kinds of decree, which we can
refer to following this summary as the ‘absolute decree’ and the ‘detailed
decree’, though the terms are not ideal? It is noteworthy that, whereas the
‘absolute’ decree is an evaluation that is all good for the object because the
decree comes from divine wisdom and knowledge, the ‘detailed decree’ is of
the coming-to-be of good and evil, beauty and ugliness, wisdom and foolish-
ness. When al-Maturidi talks of the distinction between satisfaction and will,
he has in mind (under ‘will’) that everything that is created by the absolute
decree in its final connection with everything else in the history of the universe
is good, and is under God’s working all things together for good; but when
each type of action is put together with its results and circumstances, but still
isolated from the final disposition of the whole universe, it can be good or
evil.99 We can, that is to say, disappoint God in the choice that we make with
our first power, and with the act that God then creates and we acquire by our
second power. God is not satisfied by the good and evil, the beautiful and the
ugly, the wise and the foolish, but only by the good, the beautiful, and the wise.
God chooses to reward in accordance with this ‘detailed decree’ only what
satisfies Him and to punish only those ‘He does not like’. But God by His
absolute decree and in His absolute power turns even the evil that we choose
into good. One way to put this would be to use a distinction al-Maturidi does
not: a murder can still be wrong even though God turns it to good. If this is al-
Maturidi’s picture, he has a way to repair the fissure in the providential circle
we imagined when describing al-Ash‘ari’s views. It will still be the case that we
can attribute the whole final circle to God’s good care.

6.3. REVELATION AND REASON

All three of our authors have an important place for both revelation and
reason, but they describe the relation between the two sources of knowledge
differently. The term ‘revelation’ is a convenience, but is potentially

98
Roots, 213.
99
Cf. Marvin Fox’s account in Interpreting Maimonides: Studies in Methodology, Metaphys-
ics, and Moral Philosophy (Chicago: University of Chicago Press, 1990), 302, of Maimonides’
view that ‘all that God does is only for the good’.
 Divine Command in Some Medieval Islamic Thinkers 213
misleading. It would be better, but cumbersome, to talk about God’s deliver-
ances through the Scriptures and the Traditions.

6.3.1. ‘Abd al-Jabbar

‘Abd al-Jabbar makes a distinction between necessary knowledge and acquired

knowledge.100 Necessary knowledge, unlike acquired knowledge, is known
immediately and is known by all sane adult human beings. The category
includes knowledge from sense perception and rules of logic and knowledge
of one’s own mental states. For our purposes, the most important items
included are certain moral truths and reliable reports. About the first of
these he says: ‘Knowing some aspects (wujuh) that make actions good and
some that make actions bad and some obligations indicates the maturity of the
intellect. Thus an adult with sound mind necessarily knows the evil of
wrongdoing, the evil of being ungrateful to a benefactor, and the evil of
lying if it is not intended to bring about benefit or to repel harm. One also
knows the goodness of compassion and giving. One also knows that thanking
a benefactor and returning a trust when asked for and being just are all
considered obligations.’101 These moral principles are the basis for rational
obligations.102 Knowledge of reliable report is also, for ‘Abd al-Jabbar, neces-
sary knowledge, and is required for religious obligation, which is a part of
obligation not known by reason—for example, the obligation to pray and
fast.103 The question for this section of the chapter, then, is what is the relation
between these two types of necessary knowledge.
One key passage for understanding this relation is the following:
Revelation only uncovers about the character of these acts aspects whose wrong-
ness or rightness we should recognize if we knew them by reason. For if we had
known by reason that prayer is of great benefit to us, leading us to choose our
duty and to earn reward thereby, we should have known that it was obligatory by
reason. Therefore we say that revelation does not necessitate the wrongness or
rightness of anything; it only uncovers the character of the act by way of
indication, just as reason does, and distinguishes between the command of the
Exalted and that of another being by His wisdom, who never commands what is
wrong to be commanded.104
This passage gives a kind of priority to reason over revelation. Neither
revelation nor reason, ‘Abd al-Jabbar says, makes something right or wrong.

100
The use of the term ‘necessary knowledge’ here is misleading to modern ears, because it
makes us assume we are dealing with knowledge of logical or physical necessity as opposed to
contingency.
101 102 103
Mughni, XI. 384. Ibid. XI. 298. Ibid. XI. 385.
104
Ibid. VI. i. 64. See Hourani, Reason and Tradition in Islamic Ethics, 104, 115.
214 God’s Command
But the right that revelation indicates, reason sees is instrumental towards a
right that reason already knows. We know by necessary knowledge that we
should choose our duty, and revelation tells us that prayer is conducive to this
end. There is a difference between intrinsic wrongs and things that are wrong
by relation to their consequences, ‘such as the wrongs of the Law, which are
only wrong inasmuch as they lead to the performance of a rational wrong or
ceasing to perform certain duties’.105 This does not mean, however, that
revelation is redundant. One may not know, before being told, how to achieve
the end in question, and one also may be insufficiently motivated. ‘Knowing
God is considered grace, since one then knows that punishment will be
deserved for wrong doing and reward will be deserved for right deeds. Thus
the person becomes closer to avoiding wrong and pursuing worship.’106
The opponents of the Mu‘tazilites tended to object that the moral principles
that are supposed to be necessarily known, and so known to all sane adults, are
in fact not known by all. There is, in fact, widespread disagreement. Here are
two examples. The nomadic Bedouins of Arabia, ‘Abd al-Jabbar concedes,
approve the practice of plunder, whereas those in settled communities disap-
prove. But this does not mean that there is disagreement here about the
principle that injustice is prohibited; it is just that Bedouins have a different
conception of private property.107 In the same way, the Kharijites approve of
killing anyone who disagrees with them in doctrine. But this is ‘a result of their
believing that their fate is deserved; if they knew that it was not deserved, they
would know the killing to be wrong’.108 There is a problem here we considered
in Section 6.1.1; it is hard to see that the objection from disagreement can be
overcome in this way. To be sure, ‘injustice’ is named together with the wrong,
and so anyone who agrees that some act is unjust is going to agree that it is
wrong; but the relevant disagreement is surely about what kinds of act are
unjust. For example, in the case of lying, the Mu‘tazilite holds that it is in itself
wrong, and there is no compensating right-making property that could prop-
erly override this judgement. But one might object that this is not something
about which all sane adults agree, and indeed many of ‘Abd al-Jabbar’s own
opponents disagreed with him, holding that it would be right to lie to save the
life of a prophet. But the case of lying is anomalous here (as Hourani says it is,
but for different reasons), and it may be that he is using a conception of ‘lying’
that is already evaluatively laden (as suggested in Section 6.1.1). There is
nothing implausible about holding that there are very general principles that
are very widely shared across human cultures, as long as one does not insist
that they generate absolute prohibitions.
According to ‘Abd al-Jabbar, we need to distinguish rational worship and
religious worship. Both kinds involve obligations that are assigned by God. He

105 106 107

Mughni, VI. i. 58. Ibid. XI. 324. Ibid. VI. i. 25.
108
Ibid. VI. i. 20.
 Divine Command in Some Medieval Islamic Thinkers 215
says: ‘Rational worship, in order to be properly performed, does not require
anything except to be performed in the right way. Approaching the One who
has to be worshipped is not a condition for the validity of rational worship, but
it is a condition for religious worship.’109 This seems to imply that we can
worship rationally by obedience to the principles that are necessarily known,
even if we do not know about God, and even if the obedience is not con-
sciously directed towards God. On this view, it is only in relation to religious
worship that God must be ‘described with every and each action’, to use
al-Maturidi’s terms.

6.3.2. Al-Ash‘ari

The difference between al-Ash‘ari and the Mu‘tazilites on the relation between
revelation and reason comes up already in al-Ash‘ari’s list of complaints at the
beginning of Elucidation:
So that they interpret the Qur’an according to their opinions with an interpret-
ation for which God has neither revealed authority nor shown proof, and which
they have not derived from the Apostle of the Lord of the Worlds or from the
ancients of the past; and, as a result, they oppose the traditions of the Compan-
ions, related on the authority of the Prophet of God, concerning God’s visibility to
sight. . . . Also, they maintain the createdness of the Qur’an; thereby approximat-
ing the belief of their brethren among the polytheists, who said, ‘it is merely the
word of a mortal’.110
Al-Ash‘ari is clearly not rejecting the use of reason. His dreams led him to
reject the Mu‘tazilite way, but they did not tell him to stop doing theology
(kalam); rather, he should use it for the defence of a more traditional doctrine.
His work is full of arguments, and is in fact mostly a string of arguments, one
after the other, rather in the manner of the Christian scholastics of the
thirteenth and fourteenth centuries. The parallel is probably not coincidental.
The more that contemporary scholars become conversant in both traditions,
the more they are likely to see influence in both style and content from the
Muslim theologians to the Christian.
Al-Ash‘ari uses reason conspicuously. But the relation between reason and
revelation is approximately the opposite way round from how ‘Abd al-Jabbar
describes it. Al-Ash‘ari operates on the assumption that the Qur’an and the
Traditions are to be interpreted literally wherever this is possible. One place
this comes up is in his acceptance of the literal visibility of God in the next life.
‘We hold that God will be seen in the next world by sight (as the moon is seen
on the night it is full, so shall the faithful see him, as we are told in the

109 110
Ibid. IV. 329. Elucidation, 47.
216 God’s Command
traditions that come down on the authority of God’s Apostle).’111 Al-Ash‘ari
notes that God manifested himself on the mountain to Moses. The same is
true about the words of the Qur’an that God has two hands, and eyes, and sits
on a throne.112 In the first two of these cases, he adds the phrase ‘no explaining
how’ (bila kayfa). He explicitly aligns himself here with ibn Hanbal, ‘because
he is the excellent imam and the perfect leader, through whom God declared
the truth, removed error, manifested the modes of actions, and overcame the
innovations of the innovators, the deviation of the deviators, and the
skepticism of the skeptics.’113 Al-Ash‘ari acknowledges that the Qur’an does
also sometimes speak metaphorically, as in ‘a wall threatening to collapse’.114
The wall has no will, and cannot threaten in the literal sense. But it is not
impossible for the Creator really to will or to speak. Al-Ash‘ari’s principle is
that we should use the literal interpretation unless it is impossible.115
His second major criticism of the Mu‘tazilites is that they hold the Qur’an to
be created, whereas al-Ash‘ari holds that it was recorded in time, but is itself
eternal. Here are two of his many arguments for his position. The first
argument is that the Qur’an is God’s speech, and God creates by speech (by
saying ‘Be’). But, if God’s speech were itself created, it would have to be created
by speech, and we would get an infinite regress of creation.116 The second
argument is that, if God had ever been not-speaking, he would have to be
qualified by one of the contraries of speech, such as silence or some ailment
(like dumbness). But then that attribute would have to be eternal, and then it
would be impossible for it to cease, since what is eternal has no beginning or
ending.117
The most important point for our purposes is that al-Ash‘ari does not think
we are justified in holding revelation to some standard of interpretation
external to it. God gives guidance, he says, to the faithful, and not to the
unfaithful (the infidels). Section 6.1.2 quoted a text from Elucidation where he
discusses this. He considers the objection that the Qur’an also has verses that
teach that the Prophet warns both the one who follows and the one who does
not, the unfaithful one.118 Al-Ash‘ari concludes from this that guidance and
warning are different. The revelation warns both faithful and unfaithful, but
only guides the faithful, and there are some warnings also that are given only
to the faithful. The point is just that Al-Ash‘ari cannot allow what the
Mu‘tazilites assert, that the guidance gives all human beings what they then
recognize as means to what their reason already prescribed for them.

111 112
Ibid. 51–2. Ibid. 50, quoting Qur’an 38: 75, 54: 14, and 20: 4.
113 114
Ibid. 49–50. Ibid. 30, quoting Qur’an 18: 77.
115
The limits of possibility here are wider than one might expect. For example, it is possible
for God to have hands without having a body.
116
Kitab, 27.
117
Ibid. 33. The argument relies on the premiss that all God’s attributes are eternal.
118
Elucidation, 123, quoting Qur’an 41: 12.
 Divine Command in Some Medieval Islamic Thinkers 217
We can relate al-Ash‘ari’s position about the sources of theological know-
ledge to the four traditional sources of Islamic law (the Qur’an, the Traditions,
the consensus of the faithful, and analogical deduction from Scripture).119 Of
these, the first two are given by revelation. For Al-Ash‘ari, the third (consen-
sus), as it applies to theological knowledge, is also given by special divine grace.
According to one of the Traditions: ‘It is related on the authority of the
Prophet that whatever believers see as good is good with God, and whatever
Muslims see as bad is bad with God.’120 But this is not because of a general
truth about communities of religious believers, but because of a special
dispensation given to Muslims. The fourth source, analogy (qiyas), is likewise
strictly restricted in its theological use to what is implied by the revealed texts
themselves. Sometimes (but not always) we can tell from a scriptural prescrip-
tion what God’s reason is for prescribing in this way, and sometimes we can
apply that reason to cases analogous to the original case. For example, wine is
prohibited because it is intoxicating, and we can apply this to other intoxicat-
ing drinks. But the point is that al-Ash‘ari, in accepting these four traditional
sources, is not putting them under two mutually independent headings,
revelation encompassing the first two and reason the second two; rather, the
second two depend for their authority upon special revelation.

6.3.3. Al-Maturidi

We started the discussion of al-Maturidi by quoting from the beginning of

Kitab al-Tawhid, to the effect that in a pluralist society reason is necessary in
defence of the faith. That passage goes on to claim that God gives us, together
with His commands, an understanding of them. Al-Maturidi tells us that this
understanding comes through reasoning (nazar). We can now look at the
relation he sees between this reasoning and revelation. He tells us that
reasoning and theology originally occur in that use of reason through which
we know right and wrong.121 He claims that there are three possible outcomes
of reasoning. The first is that the thinker will be ‘led to the knowledge of his
being created and to see that he has a Creator who will reward him for his
good and punish him for his bad deeds, which, in consequence, will inspire
him to avoid that which angers Him and adopt that which pleases Him’. The
second is that the thinker will deny all this, and indulge himself in all kinds of
pleasure, which will have its consequence in the hereafter. The third is that the

119
See Hourani, Reason and Tradition in Islamic Ethics, 200.
120
Malik-Shaybani, al-Muwatta (Lucknow: 1880, 1888–9), 140, discussed by Hourani, Rea-
son and Tradition in Islamic Ethics, 195.
121
Tawhid, 135–6. Note the primacy of practice here, as in Scotus’s account of theology as a
practical science in Lectura prol., pars 4, qq 1 and 2.
218 God’s Command
thinker will be led ‘to the realization of the incomprehensibility of knowledge
and its reality which inspired him to search, but, then, his heart will rest and
the pain will disappear which afflicts him when he tries to think’.122 Of these
three possible outcomes, al-Maturidi claims that the second will not in fact
occur, because he has the optimistic confidence that reflection has its own
momentum towards belief in God; in fact he mounts an elaborate form of the
cosmological argument. But this means that reasoning, whether on the first
outcome or the third, will be ‘a gain to the thinker in all of its aspects’.
Al-Maturidi argues that, if God did not give us understanding together with
the command, we would be like a slave who could say to the master ‘if I had
known that my action would anger you, I would have avoided it’.123 As it is,
however, God has both given us ‘a sign by way of which [man] can know the
command, and He has stirred his mind to thought and reminded him of the
various consequences [of his actions]’. This means that, if he then disobeys
God, ‘that will be only because of his abandoning the pursuit of reasoning, and
that is his own fault. Therefore, he will be argued against on the Day of
Judgement by the very thing he could be excused by [if there were no sign
or stirring towards reflection]. This is a result of his own act.’ Ignorance is here
no excuse, and that is because God has given both the sign and the prompting,
which are in happy harmony; reflection will lead to belief in the very God who
gives the command. Al-Maturidi also holds that the Qur’an itself suggests that
we engage in speculative thought; it ‘also has informed [us] that this [reason-
ing] will successfully lead them to the truth and will show them the path’.124
The important point here is al-Maturidi’s use of the language of ‘both/and’,
in what he calls ‘two methods’. The beginning of the following passage was
quoted in Section 6.2.3, but the end is more relevant to our present concerns:
God created the kinds of things we are responsible for in two categories, namely
hard and easy, level and steep. God arranged principles for this system and by
these principles people can reach the main principle that makes it possible to
grasp every virtue. The main principle is to reach knowledge with the following
two methods: the clear method that is obvious and the hidden method that is
covered. The aim of this is to make clear that the height that people will attain
depends on the degree of their bearing things that human beings by nature do not
like and the souls of human beings hate and depends on the hard work of people
who use their reason. Because of that, God determines two ways for that know-
ledge. The first one is theoretical intuition, which is the noblest of the sources of
knowledge and it does not contain any ignorance. Therefore it makes the ground

122
Compare the first sentence of the first edition of Kant’s first Critique (KrV A vii): ‘Human
reason has the peculiar fate in one species of its cognitions that it is burdened with questions
which it cannot dismiss, since they are given to it as problems by the nature of reason itself, but
which it also cannot answer, since they transcend every capacity of human reason.’ For Kant the
realization of incomprehensibility allows faith.
123 124
Tawhid, 137. Ibid. 10.
 Divine Command in Some Medieval Islamic Thinkers 219
for the rest of knowledge that is hidden. The second one is religious sources
[naql], and it is possible to understand within the domain of perceptible things
whether [this second] is true or not.125
A first reading of this passage might suggest to readers who are familiar with
Western thought since Descartes that the ‘clear’ way is the way of reason, and
the ‘hidden’ way is the way of revelation, but actually al-Maturidi means the
two to be the other way around. The way of revelation is clear, accessible
within the domain of perceptible things. The way of speculative thought is
hidden. It may start from something like theoretical intuition, but it requires
difficult reflection about things that are beyond the reach of the senses. He
means to make these two ways, in Arabic ‘aql (sources in reason) and naql
(religious sources), partners, but not equal partners. Revelation may tell us
something that we do not understand—for example, that ‘The Merciful was
seated Himself on His throne’ (Qu’ran 20: 5), when it is rationally demon-
strable that God exists in eternity without place. If this happens, we should
accept the saying and refrain from making a definitive interpretation of it. We
believe in whatever God meant by it.126
Al-Maturidi also gives a role to reason in checking the reliability of reports.
We need speculative thinking not merely to reflect about that which is beyond
the reach of the senses, but also to check the kind of reports ‘which may or
may not be erroneous’.127 This includes reports about the Prophet by others.
‘These reports which come down to us from the Prophets, come down from
mouth to mouth and are, therefore, liable to error and lying. For [the trans-
mitters] do not possess the proof of their truthfulness nor do they have the
demonstrative proof of their infallibility. In that case, therefore, this kind of
report needs to be examined.’ Al-Maturidi accepts the principle of credulity,
that human beings have to rely on the reports of others and should therefore
give initial credence to what someone tells them, as well as to what they receive
through the senses and through reason. This is true for sceptics and agnostics
as much as religious believers.128 But what is received by report may be true or
false, and needs to be tested by a form of knowledge that can discriminate
between reliable and unreliable testimony. Al-Maturidi holds that the divine
report (the Qur’an) and the Prophet’s personal reports pass this test, and are
supported by the consensus of the faithful and by clear miraculous signs. But
historical reports in general, and some of the traditions about the Prophet, do
not have this degree of reliability.
Reason also has the function of showing that the universe has a purpose,
being made by a rational Creator, for whom ‘to act unwisely is a bad thing’,
who ‘combines that which is [properly] combined and divides that which is

125 126
Ibid. 282. Ibid. 69–74. See Kholeif (ibid., p. xxxii).
127 128
Ibid. 9. Ibid. 4.
220 God’s Command
[properly] divided’, and who directs human beings in their ‘different desires,
divergent natures, and those passions that are inlaid in [them] for the most
part’ so that they do not come to nothing ‘by mutual destruction and corrup-
tion’.129 He does, however, acknowledge that our reason has its own proper
limits. Here is one of them. He knows the work of Aristotle, in some areas in
significant detail.130 But he thinks Aristotle is misled by too ambitious an
account of analogy (qiyas). Analogy can take us from what is present to what is
absent. But, if we think, as al-Maturidi claims Aristotle did, that we can move
from the present to the absent with a proportionality of the equal to the equal,
this will give us the eternity of the universe, a doctrine that al-Maturidi rejects
but thinks rightly that Aristotle accepted (though this is not in fact Aristotle’s
argument for this doctrine). Al-Maturidi objects that, if analogy really worked
like this, ‘it would be necessary to imagine that anyone who sees himself, sees
the whole world to be like himself ’.131
By way of conclusion, we might say that al-Maturidi gives the place of a
junior partner to reason in relation to divine command. It is not, as in ‘Abd
al-Jabbar, that revelation merely gives us instruments to what are already
known as ends by reason. And it is not, as in al-Ash‘ari, that reason simply
works out the implications of what is already given by revelation. We could
put the matter this way. For both ‘Abd al-Jabbar and al-Ash‘ari there is only
one final place for access to our proper ends; for the Mu‘tazilite it is reason,
and for al-Ash‘ari it is revelation. But for al-Maturidi there are two, and they
are mutually reinforcing. This is not to say that they are equal in status, for our
human ‘rational faculties were originated as finite and therefore are short of
grasping the absolute reality of things. . . . This is because the rational faculties
are parts of the world which is in its entirety finite.’132 The Qur’an is,
al-Maturidi and al-Ash‘ari agree, God’s own eternal word, received by the
Prophet. But, in the view of al-Maturidi, God has stirred our minds to be
receptive to another source of value, the reason that God himself has for his
command, and, though we are divided in our nature, and our access to this
source of value is not reliable, we have been given difficult and partial access if
we do the necessary hard work.
What can a divine command theorist learn from al-Maturidi? Many things,
but here are three, one each from Sections 6.1.3, 6.2.3, and 6.3.3. First, it is
consistent to hold both that God makes the divine command intelligible to us,
even sometimes giving us access to the divine reason for the command, and to
say that our access is only partial and difficult. The combination here comes

129
Ibid. 4–5. If we were left alone to our natures, al-Maturidi says, we would destroy each
other and ourselves.
130
For example, he gives a detailed account of the Categories: ibid. 147.
131
Ibid. 28.
132
This is a quotation from al-Maturidi’s exegetical work, Tawilat Ahl al-Sunnam, ed.
Ibrahim Awadain and al-Sayyid Awadain (Cairo, 1971). See Roots, 90–1.
 Divine Command in Some Medieval Islamic Thinkers 221

from the fact that our nature is divided. Because we have a tendency towards
what has bad results along with a reason that is attracted by what is good, our
inclinations are not a reliable guide. Therefore we cannot derive rightness
from our inclinations. But the rightness and wrongness derive from our
relation to God, which is what determines whether some initial harm is in
the end a benefit, and, since God knows this relation much more perfectly than
we do, God’s commands are more helpful even than our knowledge of
ourselves. Second, it is consistent to hold both that we have the power to act
in opposite ways, and that what we do is determined by the divine decree. This
decree needs to be distinguished into what God reveals to us as the divine
preference, which we can disappoint, and God’s final effective command,
which brings always overall good. The three linked distinctions, between two
kinds of power, two kinds of divine attitude, and two kinds of divine decree,
start to give us a way to hold together God’s sovereignty with our freedom.
Finally, we can see in al-Maturidi an acknowledgement of the authority of both
reason and revelation. He refuses to reduce the final authority of revelation to
that of reason or vice versa. The metaphor of reason being a ‘junior partner’ is
not his language. But the idea that we need both, and not merely instrumen-
tally, is important for anyone living as he did, and we do, in a pluralistic
culture. To some limited extent (because reason is the junior partner) we can
rely on what is common between traditions to adjudicate disagreements
between them.
In all three of these ways it is instructive to compare al-Maturidi with
Scotus. There is no suggestion in his texts that Scotus had any knowledge of
al-Maturidi, but the two play some of the same mediating roles in the debates
within their own communities. We can see this similarity in relation to each of
the three questions that have structured the present chapter. First, like al-
Maturidi, Scotus is hesitant to allow a deduction from our nature to the moral
law. Like al-Maturidi, he thinks there is a consonance or fittingness of the
commandments with our essence, and that our essence is to be pilgrims on
the way to a certain relation to God. But our composite nature makes the
deduction problematic, even though we can see the fittingness with our
reason. Second, Scotus holds that we have the power of opposites, which is
like al-Maturidi’s first kind of power. But God’s generosity leads to our good
(as in al-Maturidi’s second kind of divine decree), despite our tendency to
move towards what is not good. This generosity is consistent with the divine
justice that punishes us when we fail to do what God commands by the
revealed divine will (as in al-Maturidi’s first kind of divine decree). Finally,
Scotus has the same combination of trust in human reason and emphasis upon
its limitations. He gives what is probably the most complex rational argument
for the existence of God in the whole of Christian scholastic philosophy, and
he is insistent on the need and capability of ‘right reason’ to work out how we
ought to live. On the other hand, he thinks we do not know our own
222 God’s Command
(individual) essence, or the essence of God, or who the people are that God has
elected for salvation, and he is hesitant about saying that we know by reason
what God must do.
Al-Maturidi gives us useful middle ground. His voice needs to be heard in
the current debate about divine command in ethical theory, together with the
voice of other late Ash‘arites such as al-Baqillani and al-Juwayni. The sooner
we can get a good English translation of his major work with commentary to
explain the points of obscurity, the better positioned the international com-
munity of scholars will be. But it is also true that more of us need to learn
Arabic and read the texts in the original.
 7

Divine Command in Some Recent

Jewish Thinkers

I N T R O D U C TI O N

This chapter will be, like Chapter 6, a discussion of three figures who occupy
respectively a divine command position, a natural law position, and a position
that retains a strong emphasis on divine command but avoids the errors of the
first position. These are three Jewish philosophers writing in the last hundred
years, and the chapter will focus, especially for the first two, on their response
to Maimonides’ Guide of the Perplexed.1 This focus has the advantage of a
more manageable scope, but also of connecting the present chapter naturally
with the previous one. Maimonides wrote in Arabic, and was familiar with
many of the ideas discussed in Chapter 6 (though there is no sign that he knew
the work of al-Maturidi).2 This chapter differs from Chapter 6, however,
because it does not split each figure’s views under three distinct questions
that frame the chapter as a whole. Rather, it treats each as generating his own
questions, though the common themes between them should be readily
apparent. This chapter is also like Chapter 6 in that it trespasses on territory
that requires a lifetime or several lifetimes for mastery, and it will undoubtedly
make errors of interpretation.3 It is important to my project, however, to
illustrate the commonalities between the three Abrahamic faiths. My thesis is

1
I will use the translation by Shlomo Pines: Maimonides, The Guide of the Perplexed
(Chicago: University of Chicago Press, 1963) (henceforth Guide) and I will give the page
numbers of this translation. But I will occasionally refer to the translation by M. Friedländer:
Maimonides, The Guide for the Perplexed (1940; 2nd edn, New York: Dover, 1956), and I will
note where I do so.
2
Maimonides discusses the Mu‘tazilites and the Ash‘arites in Guide, i. 71, and says that some
of ‘our’ authors followed the theory and the method of the first-named sect, though this was
because they encountered them first, adopted them, and ‘treated them as demonstrated truth’.
3
I have been helped very significantly by Paul Franks in this chapter, who suggested my
starting from Marvin Fox and David Novak, and endorsed the attempt to understand Franz
Rosenzweig. He should not be held responsible, however, for my errors in the execution of
this plan.
224 God’s Command
that the debate between divine command and natural law is at home within
each of them. I am defending a version of divine command theory, and it helps
my case to see that all three of these faiths have a central concern with divine
command. They also, all three, wrestle with the question of how divine
command relates to human nature. I am not, by saying this, trying to suggest
that the three faiths are just the same in the way they deal with this set of
issues. In the present chapter, I will occasionally comment on how Judaism is
different from Christianity in the way the issues present themselves within it.

7.1. MARVIN FOX ON MAIMONIDES

7.1.1. An Esoteric Text

One disadvantage of focusing on the Guide is that it is an esoteric text, and just
what this means for our understanding is itself hotly debated between
scholars. Maimonides says that ‘in speaking about very obscure matters it is
necessary to conceal some parts and to disclose others’.4 He goes on to say
that, sometimes, ‘this necessity requires that the discussion proceed on the
basis of a certain premise, whereas in another place necessity requires that the
discussion proceed on the basis of another premise contradicting the first one’.
And he warns that ‘the vulgar must in no way be aware of the contradiction;
the author accordingly uses some device to conceal it by all means’. But what
does this mean for our interpretation? The ‘perplexed’ that Maimonides is
trying to guide are those who do not see how to put together their Jewish faith
with the teachings of the philosophers. He wants to give them peace in their
perplexity without undermining the faith of the much greater number of less-
educated people who are not troubled by this question. Hence the esoteric
method. But, again, what does this mean for how we are to understand him?
Here are three possible interpretations, of which the best is the third. One
possible rule, suggested by Leo Strauss (who wrote the introduction to Pines’s
translation), is as follows. ‘We may therefore establish the rule that of two
contradictory statements in the Guide or in any work of Maimonides that
statement which occurs least frequently, or even which occurs only once, was
considered by him to be true.’5 He gives as an example Maimonides’ reference
to a single passage in the Book of Daniel, which Maimonides takes to establish
that we are resurrected after death, even though there are many other biblical
passages indicating the contrary. But one example does not license a rule; and

4
Guide, i, introduction, 18.
5
Leo Strauss, ‘The Literary Character of The Guide for the Perplexed’, in Persecution and the
Art of Writing (Glencoe, IL: Free Press, 1952), 74. This rule is discussed by Marvin Fox in IM 70.
 Divine Command in Some Recent Jewish Thinkers 225
in the case of this particular rule, its general application to the texts of
Maimonides would have highly counter-intuitive results.6 I am not going to
apply Strauss’s rule in what follows. But in any case the goal of my project is
not to determine what Maimonides actually meant. I am interested, rather, in
the different views of divine command that different Jewish interpreters derive
from his text.
A second interpretation is that in Marvin Fox’s Interpreting Maimonides.
He has his own way to take the esoteric character of the Guide, and it seems
problematic, just as Strauss’s does. Fox thinks that Maimonides ‘deliberately
takes the position that opposed views may each have so much to recommend
them that we must commit ourselves to both and hold them in a balanced
dialectical tension’.7 On this view, Maimonides excels in the art of preserving
this balance. The phrase ‘opposed views’ here is important, since Maimonides
does not tell us to embrace strict logical contradictories; and it is good that he
does not, because our use of logic requires that this be prohibited. Rather,
Maimonides introduces the term ‘divergences’, and says that their inclusion in
his book is due to the need to accommodate the student’s level of preparation
and to start from different premisses on different occasions.8 This is the
passage just quoted in the first paragraph. But surely it is problematic to
recommend finding and embracing dialectical tension as a general policy; we
end up in a general fog. If we find ourselves drawn to two opposing views, we
should do our utmost to discover which of them we can finally reject, or to find
a synthesis between them.
A third interpretation is that the Guide does not suggest dialectical tension
as the goal, though admitting it is sometimes the best we can do for now.
Surprisingly often, what looks like an impasse yields to a set of distinctions
that allows us to go forward. We start from a premiss on one occasion that
seems right in its given context. On another occasion we start from an
opposing premiss that also seems right, in a different context. We then do
our utmost to find either a ranking, or a set of qualifications that enables us
either to hold both or to reject one. If we do not succeed, that is not because
dialectical tension is our proper destination, but because we have not yet
found the right distinctions or because human embodied minds are not
adequate to the task.9 This unfortunate failure is what needs to be kept
obscure, so as not to disturb the many.

6
For example, Maimonides says once (Guide, ii. 24, p. 327): ‘And even the general conclusion
that may be drawn from [the heavens], namely that they prove the existence of the Mover, is a
matter the knowledge of which cannot be reached by human intellects.’ But he repeatedly relies
upon a proof of the existence of God drawn, like Aristotle’s, from the movements of the heavenly
bodies.
7 8
IM 23. Guide, i, introduction, 19–20.
9
The second of these two possibilities is emphasized by Josef Stern, in The Matter and Form
of Maimonides’ Guide (Cambridge, MA: Harvard University Press, 2013).
226 God’s Command

7.1.2. Maimonides on the Mean

Fox’s account of how Maimonides sees the relation between obligation and
nature depends on his account of how Maimonides relates to Aristotle’s
doctrine of the mean. Fox thinks of Aristotle as defending the view that virtue
lies in a mean because virtue is natural. If Maimonides were to follow Aristotle
in this, Fox says, he would be locating the ground of our morality in our
nature. This takes us back to the topic of the so-called Aristotelian categoricals
discussed in Chapter 4, and Foot’s delight in Geach’s dictum that humans
need the virtues just as bees need stings. Fox denies that Maimonides follows
Aristotle in this, and it is worth looking at the details, because they show just
how Fox wants to separate Maimonides from the tradition of natural law.10
This subsection describes how Fox sees Aristotle here and suggests that he
misconstrues Aristotle in two key areas. But then it concludes that he is
nonetheless right to see a fundamental difference between the two philo-
sophers on the doctrine of the mean.
For Aristotle, as Fox sees him, ‘moral virtue, like any virtue (arete), is
concerned with the proper excellence of its subject, in this case man. Man’s
proper excellence is determined by man’s proper end as it is given in nature.’11
In this way virtue is like physical health, which is ‘the end given in and defined
by the physical nature of the human patient’.12 Since we are both animal and
rational, Fox says, our goal is to subject our animal nature to the rule of reason.
He then tries to show that for Aristotle ‘the nature of all things is to find their
proper completion, qua natural, in the mean or the middle way’ (to meson),
and that therefore ‘this is also the way that reason requires us to choose in
order to achieve moral virtue’.13 Fox cites numerous texts from Aristotle’s
biological and practical works in order to make this point. One example is ‘of
ears, some are smooth, some are shaggy, and some are of medium texture; the
last kind are best for hearing’.14 Fox thinks that the Aristotelian model of
deliberation is an appeal that reason makes to nature’s middle way. He
concedes that the appeal to nature is not the whole story in Aristotle’s account

10
Fox is here following Hermann Cohen, Charakhteristik der Ethik Maimunis, in Jüdische Schriften,
iii (Berlin: B. Strauss, 1924).
11
IM 100.
12
IM 98. Fox is here following Werner Jaeger, ‘Aristotle’s Use of Medicine as a Model of
Method in his Ethics’, Journal of Hellenic Studies, 77/1 (1957), 57.
13
IM 101.
14
Aristotle, Historia Animalium, 1. 10. 492a32–3. These quotations from Aristotle use the
translation in The Complete Works of Aristotle, ed. Jonathan Barnes (Princeton: Princeton
University Press, 1984). Fox also cites Aristotle’s view that the heart is in the middle (to
meson) (De Partibus Animalium, III. 4. 666a15), that the temperature of an animal should be
brought to the mean (ibid. II. 7. 652b17–20), and that there should be a mean in quickness of
speech (De Rhetorica, III. 15. 1416b30–5), and in the structure of the state’s distribution of the
goods of fortune (Politica IV. 11. 1295b3–7).
 Divine Command in Some Recent Jewish Thinkers 227
of deliberation, because there is need also for the exercise of judgement in
particular cases (akin to perception), and for appeal to one’s society and its
conventions (different from one society to another). But Fox makes the point,
like Foot’s argument to a similar effect, that it is our nature that gives us
sociality, and so there is no final dichotomy between the two.
The main point in this subsection of the chapter is to show how Fox
distinguishes Maimonides from Aristotle, even though both of them uphold
a doctrine of the mean. But, before that, it is worth mentioning two ways in
which Fox gets Aristotle wrong. The first of these is that Fox has not appre-
ciated how Aristotle in the Nicomachean Ethics separates the evaluative from
the descriptive components of the mean.15 We can see emotions like fear and
the appetite for risk as lying on a continuum, and the brave or courageous
person as feeling a mean amount of them, which is between excess and defect.
But Aristotle also says that the mean is ‘a peak’ with regard to what is best and
right, for neither every action nor every passion admits of a mean: ‘some have
names that already imply badness, e.g. spite, shamelessness, envy, and in the
case of actions adultery, theft, murder.’16 For example, the best amount of
calories to eat in a day (the peak) is going to vary between ways of life (the
boxer’s and the accountant’s), and is not in any case going to be descriptively
in the very middle. With some kinds of action and passion there is, indeed, no
best descriptive middle at all, and the best state is zero.
The second place where Fox gets Aristotle wrong is that in his discussion of
the individual moral virtues Aristotle does not in fact even once mention the
kind of argument from nature that Fox supposes he has in mind.17 It is true
that he mentions the human function when he connects happiness with
virtue.18 And it is true that he therefore thinks of the moral virtues and the
intellectual virtue of practical wisdom as subordinate in some way (hotly
disputed between scholars) to the intellectual virtue of theoretical wisdom.19
But neither of these points alters the fact that, in his actual treatment of the
moral virtues, one by one, he seems indifferent to an argument from natural
law. This is not surprising because he does not have the notion of ‘natural law’
that would be required here to make this argument.20 He does not argue that

15
I have described this point in greater detail in my God and Morality, ch. 1, pp. 51–65.
16
NE II. 6. 1107a8–10.
17
See John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994),
79, who calls the attribution to Aristotle of an attempt to provide a naturalistic foundation for
ethics ‘a historical monstrosity’. He is criticizing Alasdair MacIntyre.
18
NE I. 7. 1097b1 ff.
19
NE X. 7. 1177a11 ff. I have discussed some of this dispute in my God and Morality, 43–51.
20
See Broadie, Ethics with Aristotle, 130–1: ‘In Aristotle’s universe there is no set of laws to
which all natural things are subject. At no level, for instance, is it true that all bodies in space
behave in the same way under the same conditions. The kinds of things are many and irreducibly
different, in the following sense: the generic characteristics and analogical resemblances which
we cannot fail to notice, and which science should not ignore, are consequential on the distinct
228 God’s Command
bravery, for example, is in a mean because it is a law of nature that substances
aim at the mean. He does have a source of authority to appeal to in his
treatment of the individual moral virtues, and it is not nature, but Homer.
Thirteen times in this part of the Ethics he appeals to Homer in order to
characterize the virtues. And he appeals to notions of ‘the noble’ and ‘the
divine’ that are full of Homeric resonance (the gods have the noblest kind of
life, with the greatest ease and freedom and control of others through patron-
age). Fox admits that Aristotle is aware that societies differ in their customs
and value patterns, but nonetheless Fox’s account of Aristotle on the mean
does not mention Homer at all. This is important because we can see in Fox’s
Maimonides a tendency we saw in the previous chapter in al-Ash‘ari, a
tendency to over-emphasize the disagreement with Aristotle, even though he
is familiar with and depends upon many of the details of Aristotle’s analysis.
Fox is nonetheless right to say that Maimonides is fundamentally different
from Aristotle (the real Aristotle and not just Fox’s misreading of Aristotle),
even though both of them have a doctrine of the mean. The difference has two
parts. The first point is that Maimonides is quite different in his use of sources.
This is not because Aristotle does not have sources; he has Homer and he has
the views of ‘the many and the wise’, by which he means ordinary common
sense and the opinions of his philosophical predecessors. But Aristotle takes
his sources from his own tradition. Maimonides self-consciously argues for
the legitimacy of taking sources from outside his tradition. Thus he says that
what is of value in his own treatise (in this case The Eight Chapters) has been
gleaned ‘from the words of the wise occurring in the Midrashim, in the
Talmud, and in other works, as well as from the words of the philosophers,
ancient and recent, and also from the works of various authors. . . . One should
accept the truth from whatever source it proceeds.’21 Maimonides is willing to
make this use of external sources, and to justify it as required for intellectual
virtue. There are similar points made by the Mu‘tazilites discussed in
Chapter 6. In Aristotle there is no such use of sources seen as external to his
tradition and therefore no such need for justification.
This brings us to the second fundamental difference. Their use of internal
sources is quite different. Aristotle does not have a God who commands. He
says, in the Eudemian Ethics: ‘For God is not a ruler in the sense of issuing
commands, but is the End as a means to which wisdom gives commands.’22

forms of the different natural kinds. . . . Such common features cannot explain the occurrence of
distinctive kinds of behaviour, since common features occur only as properties of so to speak
already subsistent distinctly natured substances.’ Fox makes Aristotle say that aiming at the
mean is a law of nature, and therefore reason prescribes it to us.
21
The Eight Chapters of Maimonides on Ethics, ed. and trans. Joseph I. Gorfinkle (New York:
Columbia University Press, 1966), 35–6.
22
EE VIII. 3. 1249b14–15. I have used the translation by H. Rackham in Aristotle XX:
Athenian Constitution, Eudemian Ethics, Virtues and Vices (Cambridge, MA: Harvard
 Divine Command in Some Recent Jewish Thinkers 229
But Maimonides always makes final appeal, in his discussion of how we should
live, to God’s commandments. Even where he appeals to philosophical doc-
trine (such as the doctrine of the mean), it is always with an eye to the
interpretation of the commandments. Thus he says about the Law given by
Moses:
For it says: ‘Just statutes and judgments’ [Deuteronomy 4: 8]; now you know that
the meaning of ‘just’ is equibalanced. For these are manners of worship in which
there is no burden and excess . . . nor a deficiency. . . . When we shall speak in this
treatise about the reasons accounting for the commandments, their equibalance
and wisdom will be made clear to you insofar as this is necessary. For this reason
it is said with reference to them: ‘The Law of the Lord is perfect’.23
In this passage and in others like it the doctrine of the mean is functioning as a
confirmation of the perfection of what is already revealed by command. The
content of the Law is nowhere derived from any doctrine of the mean. But Fox
locates Maimonides’ departure from Aristotle in the wrong place. Aristotle
does have an authoritative source (Homer) but he treats him as corroborating
Aristotle’s own worked-out theory. Homer’s poems are not treated like the
Torah as having the authority of divine commands.
One consequence is that Maimonides thinks these commandments of the
Lord are not ‘something that varies’.24 Fox is right that there is an important
difference here from Aristotle. Both authors think that a ruler prescribes for
the sake of the people’s virtue, which lies in a mean, and the ruler is thus in
some ways like a physician. But Maimonides’ conclusion is to distinguish law
and medicine. He says: ‘I therefore say that the Law, although it is not natural,
enters into what is natural. It is a part of the wisdom of the deity with regard to
the permanence of this species of which He willed the existence, that He put it
into its nature that individuals belonging to it should have the faculty of
ruling.’25 But he also stresses that this divine law must be ‘always practiced
in the same way’. This means that the medical analogy fails. The [divine] law
is not
dependent on changes in the circumstances of the individuals and of the times, as
is the case with regard to medical treatment, which is particularized for every
individual in conformity with his present temperament. On the contrary, gov-
ernance of the Law ought to be absolute and universal, including everyone, even if
it is suitable only for certain individuals and not suitable for others; for if it were

University Press, 1986). One might argue that Plato has the conception of a god who commands.
See Evans, God and Moral Obligation, 17–19, who points to Socrates’ various claims in the
Apology to be following divine orders (e.g. Apology 37e6). But this is fundamentally different
from Maimonides, who has God’s commands in a holy text given to a community.
23
Guide, ii. 39, 380. The quotation at the end of this passage is from Psalm 19: 7.
24 25
Guide, iii. 34, 535. Guide, ii. 40, 382.
230 God’s Command
made to fit individuals, the whole world would be corrupted ‘and you would make
out of it something that varies’.26

7.1.3. Maimonides on the Reasons for the Commandments

Maimonides thinks God has reasons for the commandments that God gives
through Moses. A large part of the Guide is devoted to the attempt to say what
these reasons are. In this way, he is like the Mu‘tazilites and like al-Maturidi.
But in this section I want to stress the ways in which he is nonetheless unlike
the Mu‘tazilites in his approach to these reasons. We saw in Chapter 6 that the
Mu‘tazilites held that the rightness/wrongness or goodness/badness that lies
behind God’s commands is known by human reason. God could, according to
‘Abd al-Jabbar, command what would be wrong, but he does not, and, if God
did, we would know it was wrong. This knowledge by reason is primary in
justification. We can therefore use these standards as a measuring rod to
determine what God is and is not commanding us to do. There is a difficulty,
as we saw, in the translation of the key normative terms (hasan and qabih),
which have a large range of meaning, and can simply mean ‘beautiful’ and
‘ugly’. There is the same difficulty with Maimonides, and we will come back to
this later in this subsection.27 For now, the important point is that Maimoni-
des holds that the authority relation is reversed from the Mu‘tazilite position.
We do indeed have some access just as human beings to the rightness or
wrongness of what is done, an access that is not restricted to the special
revelation to God’s people. But what renders conforming to these standards
obligatory is that God does reveal them in special revelation. In the same way
for a Christian who is a divine command theorist, the Golden Rule gives us
obligation not because we know the reason for it in our reason, but because
Jesus commanded it in the Sermon on the Mount.
The traditional 613 commandments of the Torah are addressed specifically
to the Jews. Some of them (the seven Noahide laws) are also addressed to all
human beings. We will come back to these in a moment, and they are a key to
understanding the difference between Marvin Fox and David Novak. Novak
says baldly: ‘Natural Law is Noahide law.’28 For now, it is sufficient to note that

26
Guide, iii. 34, 534–5.
27
Fox translates the key normative terms in Guide, i. 2, 24, as ‘beautiful and ugly’. Pines
translates as ‘fine and bad’. Friedländer translates as ‘right’ and ‘wrong’, as I suggested we should
do with the Mu‘tazilites. As noted in Chapter 6, there is the same range of meanings for the
Greek terms kalon and aischron.
28
Novak, Natural Law in Judaism (henceforth Natural Law), 191. Christine Hayes, in What’s
Divine about Divine Law? Early Perspectives (Princeton: Princeton University Press, 2015),
supports my negative view about Novak’s treatment of the Noahide laws. In particular, she
emphasizes (pp. 357–8) that these laws cannot be natural law in the classical sense because they
set up differences in the way Israelites and Gentiles are to be treated. Our earliest version states:
 Divine Command in Some Recent Jewish Thinkers 231
Maimonides nowhere suggests that there is something called ‘nature’ that we
can look at in order to justify the commandments, in the sense of rendering
them obligatory. They are obligatory whether we understand the reasons for
them or not. But sometimes we do see, after being given the obligation, how it
is beneficial to us to obey (in a way particular to that obligation, and not
derivative from the general benefit of obedience to God). Sometimes, espe-
cially with the ritual commandments, we do not see this, though Maimonides
is remarkably fertile in suggesting reasons even for these.29 It is true that there
is a traditional rabbinic comment on Leviticus 18: 4: ‘“Mine ordinances shall
ye do”: Such commandments which, if they were not written [in Scripture],
they should by right have been written, and these are they: [the laws concern-
ing] idolatry, sexual immorality, bloodshed, robbery, and blasphemy.’30 Con-
trary to the apparent meaning, Maimonides interprets this to mean that we
can see, after being given the prohibitions, that they are useful and beneficial.
Fox points out that the passage goes on to contrast these commandments with
ritual commandments whose reasons are not so easily seen, and that both
types obligate the Jews ‘because they come from God’ whether we see the
reasons or not. The key philosophical point here is to distinguish explanation
from justification. These reasons for the commandments, though they may
explain in some cases why God’s beneficence gives them to us, do not in any
case justify them in the sense of rendering them obligatory. Goodness is not
enough for justification in this sense.
To go back to the Noahide laws, there are seven of these. The following list
gives them in the order found in Maimonides: (1) the prohibition of idolatry;
(2) the prohibition of blasphemy; (3) the prohibition of wanton destruction of
human life; (4) the prohibition of adultery, incest, homosexuality, and besti-
ality; (5) the prohibition of robbery; (6) the requirement to establish a judicial
system in society; (7) the prohibition of eating a limb torn from a living
animal.31 Of these seven, one tradition says that the first six were already

‘One who violates a sexual prohibition according to Israel is judged by the laws of Israel and one
who violates a sexual prohibition according to the Noahides is judged by the laws of the
Noahides and the only [difference] is the case of [sexual relations with] a betrothed young
woman.’ Also, a Jew is not liable for certain penalties when he murders or robs a non-Jew, though
a non-Jew who murders or robs a Jew is liable. She says that the Rabbis formulated a natural law
position as an alternative, so they were aware of the possibility; but they rejected it (p. 368).
29
But see Guide, iii. 49, 606: ‘However, our intellects are incapable of apprehending the
perfection of everything that He has made and the justice of everything He has commanded.
We only apprehend the justice of some of His commandments just as we only apprehend some
of the marvels in the things He has made, in the parts of the body of animals and in the motions
of the spheres. What is hidden from us in both these classes of things is much more considerable
than what is manifest.’
30
Yoma, 67b, discussed by Fox, IM 127.
31
The commands are here given in the abbreviated terms that Novak uses (Natural Law,
149). It is worth noting that Novak restores the order of the commands that we know from our
earliest source, putting the requirement to establish a judicial system in society first. For
232 God’s Command
given to Adam, and handed on by him to his descendants, and the seventh (in
addition to the others) was given to Noah (being derived from Genesis 9: 4,
‘You must not, however, eat flesh with its life-blood in it’). Maimonides tells us
that
any man [that is, any gentile] who accepts the seven commandments and is
meticulous in observing them is thereby one of the righteous of the nations of the
world, and he has a portion in the world to come. This is only the case if he
accepts them and observes them because God commanded them in the Torah,
and taught us through our teacher, Moses, that the children of Noah had been
commanded to observe them even before the Torah was given. But if he observes
them because of his own conclusions based on reason, then he is not a resident-
alien and is not one of the righteous of the nations of the world, nor is he one of
their wise men.32
The key point here is the order of justification. Maimonides is treating the
gentiles who are under the Noahide laws as resident aliens (ger). It is irrelevant
to the exegesis of Maimonides here whether there was in fact a resident-alien
status at the time the Law was written down.33 The question is, rather, what
Maimonides meant by his description of the status of the Noahide law. The
passage just quoted suggests that he meant that those who are under this law
are gentiles under Jewish jurisdiction, and the justification for requiring their
obedience is that this law is given in the Torah.
So far Fox seems on solid ground with respect to the Noahide laws; at least
he seems so to an observer external to the tradition. However, he now makes a
mis-step. He says that Maimonides ‘considers all moral statements to be non-
cognitive. As such they are neither true nor false.’34 The mistake here is rather
similar to the mistake made by Hourani discussed in the previous chapter.
Both historians take a twentieth-century philosophical position and force it on
their respective authors. Hourani makes ‘Abd al-Jabbar sound like Ross, by
using Ross’s notion of ‘ground’ to explain the Mu‘tazilite’s notion of aspect
(wujuh). Hourani is then puzzled by how his author can treat injustice as a
value-neutral ground. Fox in a similar way wants to take the non-cognitivists
(like A. J. Ayer) as a model for understanding Maimonides. But non-
cognitivism, as discussed in connection with RMH in Section 4.3.2, does not
fit Maimonides at all well. A judgement can both be truth-apt and express

Maimonides, the priority is the foundation in God’s command. For Novak, the priority is the
commonality of all human beings in their need for justice. His list also omits the details about the
differences in the treatment of Jews and non-Jews mentioned in n. 28.
32
Mishneh Torah, H. Melakhim, 8: 11. There is a textual dispute about the final phrase here,
which might contain a reading better translated ‘but he is one of their wise men’, See IM 132.
33
See Natural Law, 150: ‘The biblical institution of the ger, meaning a resident-alien, had long
since ceased to operate when the doctrine of the Noahide Law was formulated.’ Novak does not
try to justify this claim.
34
IM 150–1.
 Divine Command in Some Recent Jewish Thinkers 233
some desire or emotion or state of will. So it will not follow from the fact that a
judgement expresses an emotion or feeling or act of will that it cannot be true
or false.
The central text for understanding Maimonides on this question comes at
the beginning of the Guide, where he is dealing with the punishment of Adam
and Eve after their sin in the Garden of Eden, and with what is meant by
‘knowing good and evil’. He explains that the true perfection of a man is in his
intellect, and the possession of intellect is what makes the human ‘in the image
of God’. Moreover, only those with intellect (unlike non-human animals) can
properly be commanded. He then continues, in Fox’s translation: ‘Through
the intellect one distinguishes between truth and falsehood, and that was
found in Adam in its perfection and integrity. Beautiful and ugly, on the
other hand, belong to the things generally accepted as known [that is, con-
ventions], not to those cognized by the intellect.’35 Maimonides gives as an
example of the kind of thing that is true the judgement that the heavens are
spherical, and as an example of the kind of thing generally accepted as known
Adam’s judgement that nakedness is shameful (whereas before the sin, with
his intellect unimpaired, he was naked and not ashamed). Fox quotes, in
support of his interpretation, a remark of Moses Mendelssohn, in his gloss
on the term ‘conventions’ in Maimonides’ early Treatise on Logic: ‘These are
matters which are incapable of being either true or false, but are only ugly or
beautiful.’36 Mendelssohn says that we judge something ugly or beautiful
depending on whether it is socially approved or condemned.
Maimonides’ meaning is obscure here, but Fox is probably wrong about it.
This is because the sphere of convention is still a sphere of cognition. The
rendering of the central normative terms as ‘beautiful’ and ‘ugly’ has a
tendency to reduce value judgement, as when God sees that what he has
created is good, to the aesthetic. But, even if we accept that reduction, it is a
different claim, and probably a false one, that aesthetic judgements are non-
cognitive. In Kant’s view, just to take one example of a philosophical view on
this question, when we say that something is beautiful we are saying that it
makes possible the free play of the imagination and the understanding insofar
as they refer a given presentation to cognition in general, and we claim that
everyone should so judge.37 It can be true (or false) that an object deserves to
cause this kind of aesthetic response.

35
Guide, i. 2, 24. This is discussed by Fox (IM 136–7). Friedländer translates differently the
central passage that Fox renders as: ‘Beautiful and ugly, on the other hand, belong to the things
generally accepted as known, not to those cognized by intellect.’ He translates: ‘The right and the
wrong are terms employed in the science of apparent truths (morals), not in that of necessary
truths.’
36
Maimonides, Be’ur Millot ha-Higgayon, with commentary by Moses Mendelssohn (Berlin:
B. Cohen, 1927), 38b. This is discussed by Fox (IM 134).
37
KU v. 217.
234 God’s Command
Perhaps we should interpret Maimonides as making a distinction. There are
two ways of something being good or bad, and two ways of something being
true or false. There is a similar distinction in al-Maturidi.38 One way is what
Maimonides calls ‘necessary’, the way in which the heavens are spherical, and
God’s commandments are good. A second way is what Plato in the Republic
describes as rolling around between being and not-being. Plato says: ‘We
would seem to have found, then, that the many conventions [nomima] of
the many about the fair and honorable and other things are tumbled about in
the mid-region between that which is not and that which is in the true and
absolute sense.’39 The judgement that nakedness is shameful is a good example
of something that changes in this way, but so would be the judgement that
there are twenty-five ants in the kitchen. Perhaps most (but not all) of our
judgements about good and bad or right and wrong fall in the sphere of the
second kind of cognition, whereas, at least for the person who ‘possesses
[intellect] perfectly and completely’, more of his judgements (including his
value judgements) fall in the first sphere, as for ‘Abd al-Jabbar. Moving from
the first kind of cognition to being at home only in the second kind is indeed a
deterioration, in Maimonides’ view, and he thinks that it accompanies the
expulsion from Eden. He can very naturally, in the context of discussing this
deterioration, talk as if it is a move from the true to the good. But this will be,
on the present reading, an over-simplification. In any case it does not mean
moving from the cognitive to the non-cognitive. Perhaps Maimonides means
that most of our judgements about the good are not necessary, because they do
not follow from our nature but are ‘something that varies’ in the way that
convention varies from one society to another. In this way our morality is
different from God’s law.
We do not want a non-cognitivist reading of value judgement for Mai-
monides because he gives us reasons for the commandments that show God’s
goodness or beneficence, and the praise of that goodness should not be held to
be merely an expression of feeling or convention, rather than being true.
Section 4.3.2 defended the view that, when I praise God by saying ‘God is
good’, I do not merely express my pro-attitude, but I claim that there is a
goodness there whether I have a pro-attitude towards it or not. So on the better
reading Maimonides thinks that the judgements of the nations about, for
example, adultery or robbery, while they are truth-apt, are only secured (tied
to the ‘necessary’) by God’s revelation to Israel. This is why Adam and his
descendants even after leaving Eden knew these commandments, but only on
the basis of convention. They are not demonstrable from our nature, though
after we have them we can see how very well they fit our nature. Maimonides

38
Tawhid, 296–7.
39
Plato, Republic V. 479d, in the translation in Plato: Collected Dialogues, ed. Edith Hamilton
and Huntington Cairns (Princeton: Princeton University Press, 1961).
 Divine Command in Some Recent Jewish Thinkers 235
says: ‘The Law, although it is not natural, enters into what is natural.’40 Fox
comments: ‘The law takes account of natural qualities and is designed to cope
with them. It does not demand what human beings are incapable of doing, and
its commands are framed with an eye to what human nature requires for its
own perfection.’41 He gives the example of Maimonides’ account of God’s
institution of animal sacrifices.42 God was merciful to us, because to discon-
tinue all forms of sacrifice would have been contrary to human nature, ‘which
always likes that to which it is accustomed’.43 God therefore ‘transferred them
from created or imaginary and unreal things to His own name, may He be
exalted, commanding us to practice them with regard to Him, may He be
exalted.’ Here is an excellent example of the Law ‘entering into what is
natural’. A custom of worship ‘which was in those days general among all
men’ was made, by God’s command, an expression of the necessary truth of
the need to worship the One God. God’s Law here enters into our nature
because it fits our natural tendency to stick with what we are used to. But it
also gives us something radically new, a way to use our worship in the service
of the Lord. The trouble with Fox’s error about non-cognitivism is that it has
the same tendency as his error about Aristotle on the mean, the tendency to
take Maimonides in the direction of a view (like that of al-Ash‘ari) that denies
there is human cognitive and rational access to the good, independent of
special revelation. But, to the contrary, sticking with our present example,
there is something cognitive and truth-apt about the judgement that the law
on sacrifices fits human nature. Two claims are consistent, as argued by
al-Maturidi. We can hold both that special revelation is senior partner to the
general revelation to reason of the goodness of the commandments, and that
there is some access to the true and the good independent of special revelation.

7.2. DAVID NOVAK

The primary text under discussion in this section of the chapter is David
Novak’s Natural Law in Judaism.44 This is not a work primarily about

40 41
Guide, ii. 40, 382. IM 142.
42
See John Hare, ‘Animal Sacrifices’, in Michael Bergmann, Michael J. Murray, and Michael
C. Rea (eds), Divine Evil? The Moral Character of the God of Abraham (Oxford: Oxford
University Press, 2011), 121–49, for an ethical reflection upon this institution that makes use
of this insight of Maimonides.
43
Guide, iii. 32, 526.
44
I will also refer to David Novak, Natural Law and Revealed Torah, ed. Hava Tirosh-
Samuelson and Aaron W. Hughes (Leiden: Brill, 2014) (henceforth NLRT), and Anver Emon,
Matthew Levering, and David Novak, Natural Law: A Jewish, Christian, and Muslim Trialogue
(Oxford: Oxford University Press, 2014) (henceforth Trialogue).
236 God’s Command
Maimonides (though Maimonides is the single philosopher most discussed).
This section will emphasize the places where Novak is talking about Mai-
monides, but it will have to put these into the context of the larger argument of
the book in order to make clear what Novak is saying about him. Like all the
main figures discussed in this chapter and the last one, Novak is trying to steer
a middle path: ‘One can either affirm natural law and seemingly allow Judaism
to be swallowed up by something greater and more universal than itself ’ (as he
thinks Hermann Cohen did), ‘or one can deny natural law and leave all issues
of the relations of Jews with the non-Jewish world on the level of pure power
politics’.45 The middle path he chooses involves the assertion that, ‘in the case
of Judaism, the tradition of natural law is, of course, expressed in and through
the Noahide laws, which express the minimal standards necessary for Juda-
ism’s moral justifiability’.46 The burden of this section of the chapter is to ask
whether this middle way works, and especially its claim that the moral
justifiability of Judaism depends on the natural law; the conclusion will be
that it does not work.

7.2.1. The Current Situation of Judaism

Novak starts from three ways in which Jews are now in a situation different
from the one they lived in during the time of Maimonides. The three ways are
(1) the acquisition of citizenship by Jews as individuals in modern secular
nation states, (2) the destruction of one third of Jewry in the Holocaust, and
(3) the establishment of the State of Israel.47 By starting here, Novak makes it
plain that his central interest is the relations between Jews and non-Jews. He
thinks that these relations now require something not essentially internal to
Judaism to which both Jews and non-Jews can appeal. A central figure in this
discussion is Richard Rorty, and his notion that anyone who invokes God’s
will in a democratic conversation is unavoidably a ‘conversation-stopper’.48
Novak both agrees and disagrees with Rorty. The focus of his disagreement is
that he thinks that democratic society as a whole does not need to be one’s
primary community. Instead, a community within society, and in particular a
religiously based community, can be ‘the primary locus of our human social-
ity’.49 But he thinks ‘Richard Rorty is right about the invocation of God’s will
being a “conversation-stopper” in a conversation conducted within the con-
fines of a democratic society and observant of its criteria of discourse’.50 The

45 46 47
Natural Law, 11. NLRT 7. Natural Law, 1–9.
48
Richard Rorty, ‘Religion as Conversation-Stopper’, Common Knowledge, 3 (1994), 1–6. We
will return to this, and to Rorty’s later concessions about this point, at the end of Chapter 8, in
Section 8.4.2.
49 50
Natural Law, 19. Ibid. 16.
 Divine Command in Some Recent Jewish Thinkers 237
way to allow God into the conversation, he thinks, is to distinguish as
Maimonides does between God’s will (Novak gives the example of the com-
mand to avoid pork) and God’s wisdom (Novak gives the example of the
command to avoid murder).51 Novak says that the difference between the two
is that humans have access to the second independently of special revelation,
but not to the first. The second, he says, is formulated out of a theology of
creation, and the first out of a theology of revelation. He allows us to bring the
second, but not the first, into the public sphere. One problem here is that two
different distinctions are conflated. The first is the distinction between ethical
and ritual law. This separation is initially useful but in the end, unsustainable.
Purity laws are full of ethical content, and vice versa.52 The second is the
distinction between commandments whose goodness anyone can see when
they are revealed, and commandments whose goodness is not accessible in
that way. There is no need for these distinctions to coincide. The main
problem, however, is that the rule that we can introduce God’s wisdom into
the public forum but not God’s will accepts Rorty’s principle that what is
admissible is only what is accessible to the general public. This principle is
wrong, and Novak concedes too much here to Rorty.
A good account of why Rorty is wrong (though from a Christian viewpoint)
can be found in Miroslav Volf ’s brief but pithy book, A Public Faith.53 He says:
‘I want to make Christian communities more comfortable with being just one
of many players, so that from whatever place they find themselves—on the
margins, at the center, or anywhere in between—they can promote human
flourishing and the common good.’54 Since there is no general agreement on
what constitutes human flourishing, Christianity offers its own distinct vision,
this being ‘the most important contribution of the Christian faith to the
common good’.55 There are constraints, indeed, on this sharing. Christians
have to be willing to listen as well as to speak: ‘As givers, we respect receivers
by seeing ourselves as potential receivers too.’56 The respect here includes
forgoing the goal of taking over the society root and branch, and forgoing
coercive means of influence (and Volf concedes that Christians have often
wrongly used coercion). Finally, the sharing of the words needs to be accom-
panied by engaging in social and political action for the common good. But,
with these constraints in place, there is nothing wrong, indeed there is much
right, with the sharing of distinctive faith perspectives in the public square. To
say this is not to minimize the difficulty of implementing these constraints in

51
Guide, iii. 26.
52
This is one of the lessons of the work of Mary Douglas in Purity and Danger: An Analysis of
the Concepts of Pollution and Taboo (London: Routledge & Kegan Paul, 1966.)
53
Miroslav Volf, A Public Faith: How Followers of Christ Should Serve the Common Good
(Grand Rapids, MI: Brazos Press, 2011). See also the discussion of this book by a number of
authors in Political Theology, 14/6 (2013), 721–834.
54 55 56
Volf, A Public Faith, 79. Ibid. 63. Ibid. 111.
238 God’s Command
countries where Christianity or Islam or Judaism is a majority religion. But the
individual religions need to speak out (to witness to) their distinctive perspec-
tives in public because there is not enough ground sufficiently neutral to be
accessible to all citizens regardless of their religious affiliations or lack thereof;
not enough, that is, for the construction of a good society on such a basis. The
distinctive revelation-based witness of the different faiths can often be, as Volf
says, a contribution to public discourse, not an impediment to it. In the civil
rights movement, for example, Jews and Christians found common cause
because of their faith, not in spite of it.
Novak supports his view that Judaism needs justification through the
Noahide laws by giving a narrative of God’s dealings with the patriarchs
before Sinai. ‘The most unambiguous example’, he says, ‘of a natural law
type of position in Scripture emerges from the first unmediated covenant
between persons presented there, namely, the covenant between God and
Abraham’.57 He thinks the story shows that ‘Israel must have come to the
Torah with some degree of wisdom and understanding already in hand to be
able to accept its teaching in practice’. And he thinks ‘all of this comes out
most clearly in the dialogue between God and Abraham over the unjust
situation in the cities of Sodom and Gomorrah’. It is worth spending some
time, then, on this dialogue, and its context in Genesis. A cry comes up from
these two cities, like Abel’s blood crying from the ground. God shares with
Abraham the divine plan and explains the divine reason by saying (in Novak’s
translation): ‘Abraham will surely become a great and important nation, and
because of you will all of the nations of the earth be blessed. For I know him,
and it follows that he will command his children and household after him to
keep the way of the Lord to do righteousness and justice [tsedeqah u-mishpat]’
(Genesis 18: 18–19). Novak thinks this means that Abraham already knew
about righteousness, and that the object of this knowledge is ‘the first collective
obligation of the gentiles (and the Jews, a fortiori) to maintain courts for the
administration of justice’, which is the first (on Novak’s listing) of the
Noahide laws.
Novak does not quote, however, the key verse in this context, which is
Genesis 15: 6: ‘And [Abram] believed in the Lord; and he counted it to him for
righteousness.’ This is the only verse before the dialogue about Sodom and
Gomorrah in which Abraham’s righteousness is mentioned. Surely the key in
Genesis 18 is that God talks of knowing Abraham, and knows that he will
command his children and his household to keep the way of the Lord. In
other words, it is not some Noahide law that lies behind what God knows
Abraham knows, but it is Abraham’s faith, his believing in and his relationship
with the Lord. Novak defines the natural law as the set of ‘norms of human

57
Natural Law, 39.
 Divine Command in Some Recent Jewish Thinkers 239
conduct that are universally valid and discernible by all rational persons’.58
But surely God is pointing in Genesis 18 to something different about Abra-
ham, something that is not already common to all the nations of the earth,
because it is through Abraham and his seed that the nations of the earth will
reach a destination they have not yet reached. Novak might say that what is
different is that Abraham will keep the common norms. But the story at the
beginning of Genesis 18 that precedes the dialogue about Sodom and Gom-
orrah shows that Abraham’s goodness is something beyond these norms. The
chapter starts with a story about Abraham’s hospitality to ‘three men’ who are
presented as an appearance of God, and this is an act of hesed (loving-
kindness) that is not prescribed by one of the Noahide seven. The text
deliberately links the two stories by saying that the three men ‘rose up from
thence, and looked toward Sodom: and Abraham went with them to bring
them on the way’ (Genesis 18: 16).
Novak is too quick to say what God must do. A Scotist modesty about
the modality here is more appropriate. Novak imagines Abraham as saying
to God:
there are prior conditions, first and foremost being that the covenant must
include a morally intelligible form of polity, both in terms of the internal
relationships of the members of the covenanted community one with the other,
and their relationships with others outside their community. The minimum of
that involvement in justice is that judicial decision be made fairly (what we now
call “due process of law”). . . . For us [humans] justice is a natural necessity.59
Novak thinks this conception of justice is one that Jews now need in their new
situation, and he is right that due process of law is an excellent thing. But he
should not rely for a proof of this need on the Rortean premiss that appeal to
God’s will is a conversation-stopper, or on the details of the story of Abra-
ham’s dialogue with God over Sodom and Gomorrah.

7.2.2. Novak on Maimonides

What does any of this have to do with Maimonides? Novak, like Fox, quotes
with approval Maimonides’ dictum: ‘Therefore I say that the Law, although it
is not natural, enters into what is natural.’60 But the two exegetes interpret the
dictum differently. For Fox, we (who are natural) are the recipients of the Law
that God (who is beyond nature) speaks to us by command, and what God

58 59
Ibid. 1. Ibid. 46 (emphases added).
60
Ibid. 29, where Novak concedes that the way he appropriates this dictum is probably not
what Maimonides actually meant.
240 God’s Command
speaks fits what we are but is not deducible from it. For Novak, we can only
receive this Law (in the sense of accepting it as a law for us) if it prescribes
what is already natural to us. Justice is, as quoted from him earlier, a natural
necessity, and this means that our obligation to be just can be deduced from
our nature itself. Novak says that our reception of the Torah ‘is only possible
when [it] lends itself to a rational understanding in the world inhabited by its
human recipients’.61 Again, he appeals to Maimonides. He wants to establish
that human reason is the presupposition of revelation itself, and it is what
makes the revelation of God’s word possible in the world. He says: ‘For
Maimonides, this is so because both the Torah and the world are creations
of the same divine wisdom. That is why the science [madda] of the Torah and
the scientia of the world can employ the same methods. Both are the result of a
creation word [dibbur].’62 Maimonides here, he says, sides with Plato of the
Timaeus against Aristotle. Plato speaks of a ‘demiurge’ who ‘brings order out
of disorder’, and
there is always a residue of disorder requiring, as it were, subsequent creative acts
to bring order to it. To be sure, Plato believed that God engaged in some kind of
internal discourse in the sense that the cosmos is made ‘according to the word’.
But one cannot say that the word itself was creative. For Maimonides, though,
God’s creative act is a word [dibbur] because creation and prophetic communi-
cation are in the essence the same divine act.63
To support this claim, Novak quotes from the Guide: ‘Similarly . . . it has been
said that the world derives from the overflow [al-faiz] of God and that He has
caused to overflow to it everything in it that is produced in time. In the same
way it is said He caused His knowledge to overflow to the prophets.’64
When we turn to the quoted text, however, we find that Maimonides does
not say that creation and revelation are the same divine act, but that both are
kinds of emanation or ‘influence’ from an incorporeal being. To understand
this, we have to understand something of the doctrine of the overflow. The
Aristotelian background is the treatment of the Active or Agent Intellect in De
Anima III. 5. This is one of the most confounding chapters in Aristotle, and we
will not dwell on it long. In the Arabic sources of Maimonides there are
various interpretations, but all of them suppose that the active intellect is
separate from our bodily existence, and is the recipient and thus the mediator
of divine influence on the sublunary world, including human life.65 Maimoni-
des says:

61 62
Ibid. 30. Ibid. 30.
63
Ibid. 117 (emphasis added). See Plato, Timaeus 30a, 51a and 31b. Novak translates ho theos
as ‘God’ here, but the demiurge is not to be identified with the God of the monotheisms.
Moreover, the demiurge does not go on with his activity after the first ordering.
64
Guide, ii. 12, 279.
65
See Ogden, ‘Receiving and Making’.
 Divine Command in Some Recent Jewish Thinkers 241
Similarly with regard to the Creator, may His name be sublime; inasmuch as it
had been demonstrated that He is not a body and had been established that the
universe is an act of His and that He is its efficient cause—as we have explained
and shall explain—it has been said that the world derives from the overflow of
God and that He has caused to overflow to it everything in it that is produced in
time. In the same way it is said that He caused His knowledge to overflow to the
prophets. The meaning of all this is that these actions are the action of one who is
not a body. And it is His action that is called overflow. This term, I mean
‘overflow’, is sometimes also applied in Hebrew to God, may He be exalted,
with a view of likening Him to an overflowing spring of water, as we have
mentioned. For nothing is more fitting as a simile to the action of one that is
separate from matter than this expression, I mean ‘overflow’. For we are not
capable of finding the true reality of a term that would correspond to the true
reality of the notion. For the mental representation of the action of one who is
separate from matter is very difficult, in a way similar to the difficulty of the
mental representation of the existence of one who is separate from matter.66
This passage shows us that God’s creation and God’s revelation are only ‘the
same divine act’ in the sense that they are both divine acts. All acts upon the
physical world by incorporeal beings have this same mysterious emanative
character. We can only call the creation and the revelation the same act if we
are willing to say that all God’s acts are the same act. But, even if we think (in
accordance with the doctrine of divine simplicity) that all God’s essential
attributes are the same, it seems particularly unhelpful to say that all God’s
acts are the same.
Novak says that he does not want to reduce revelation to creation, and he is
aware of Rosenzweig’s critique of any such reduction, which will be the main
topic of the final section of this chapter. Nonetheless he does end up reducing
in this way, despite his protestations. Novak’s strategy for denying the reduc-
tion is drawn from Maimonides: Even if we can answer the ‘why’ question
about a Law, it does not follow that we can answer the ‘how’ question, or the
question how the details of the law fit the purpose assigned to it. Novak quotes
Maimonides saying: ‘Those who imagine that a cause may be found for
suchlike things are as far from the truth as those who imagine that the
generalities of a commandment are not designed with a view to some real
utility.’67 Maimonides thus accomplishes the middle path, says Novak, since
‘he saves revelation from being reduced to reason, and he saves the law from

66
Guide, ii. 12, 279.
67
Guide, iii. 26, 506–7 (emphasis added). Maimonides also says in this chapter: ‘What
everyone endowed with a sound intellect ought to believe on this subject is what I shall set
forth to you. The generalities of the commandments necessarily have a cause and have been
given because of a certain utility; their details are that in regard to which it was said of the
commandments that they were given merely for the sake of commanding something’
(emphases added).
242 God’s Command
being reduced to divine caprice’.68 Novak goes on, however, to concede that
‘when it comes to the moral law, what for us is the locus of natural law,
Maimonides seems to have been able to explain just about every detail quite
rationally’.69 It is true that with ritual law, in particular, the connection of
these purposes with all the details may not be clear. After he has gone through
all fourteen classes of the Law, Maimonides concedes that there are a few
statutes the reason for which is unknown to us, most of them serving as a fence
against idolatry. But he says that implicitly he has given the reason even of
these, and every intelligent reader will easily find it.70 The point is this: if the
strategy for denying the reduction of revelation to creation is to deny that we
know why the laws fit God’s purposes, then Maimonides is in practice guilty of
the reduction in most cases of his treatment of the Law.
The moral of this subsection of the chapter is that we should deny the
premiss here. Novak has been misled by his view (which he attributes to
Maimonides) that creation and revelation are ‘the same divine act’. The
problem comes in the mediating term ‘rationality’. If the rationality of God’s
revealed commandments derives from the benefit they give to beings of our
nature, then it is easy to think that the rationality of our following these
commandments derives from those same benefits. Chapters 3 and 4 discussed
at length the problems with eudaemonism and deductivism, respectively the
view that we should do everything that we do for the sake of our own
happiness and the view that we can deduce the moral law from our human
nature. If we start from eudaemonist and deductivist assumptions, we will end
up with a reduction of the kind Novak says he wants to avoid. Appealing to the
opacity of some of the details of some of the ritual laws is not going to help.
The key in avoiding the reduction is the distinction in Section 7.1.3 between
justification and explanation. God does indeed fit the commandments to our
nature, so that we can flourish by keeping them. But this is not the justification
of the obligation. We have the obligation just because they are commanded.71
The fittingness with our nature is not a justification but a partial explanation
(in Aristotelian terms a ‘cause’, which we can partially see retrospectively after
being given the commandments) of God’s giving them to us in the first place.
Novak is not saying that God’s role is redundant, or that all we need is
nature. This is especially clear in his trialogue with Anver Emon and Matthew
Levering.72 He accepts a version of the view that obligation necessarily has a
social character, which we discussed in Section 2.2. He thinks that the

68 69 70
Natural Law, 97. Ibid. 135. See Guide, iii. 49.
71
See ibid. iii. 24, 501, Maimonides’ discussion of the binding of Isaac, in which Abraham is
said to have been obedient ‘solely because of what is incumbent upon the Adamites—namely, to
love Him and fear Him, may He be exalted—and not . . . for any hope of a reward or for fear of
punishment’.
72
Trialogue, 20–7, esp. 2: ‘So too does the recognition of the universal morality that is natural
law depend on the recognition of the Sovereign of the universe who creates the subjects of that
 Divine Command in Some Recent Jewish Thinkers 243
recognition of natural law depends on the recognition of the Sovereign of the
universe who created us. But he thinks that the authority of special revelation
requires that we already accept the authority of natural law. This is what
Maimonides would not accept, and we should not accept it either.

7.2.3. Novak on the Noahide Laws

Finally, we can return to Novak’s treatment of the Noahide laws. He makes an

important distinction between what he calls the ‘comparative’ or ‘co-Judaic’
conception of these laws and the ‘ontological’ or ‘pre-Judaic’.73 Under the first,
‘co-Judaic’, conception, the ‘Noahide’ is the non-Jew standing before Jews here
and now, ‘the “other” with whom Jews desire to discover some significant
commonality without, however, sacrificing the singularity which is theirs
because of the covenant’. This is the emphasis of Marvin Fox, and Novak
says (without evidence) that there was probably no resident-alien status at the
time the Law was written down. But we argued in Section 7.1.3 that this does
not show that Maimonides was not using a co-Judaic conception. To under-
stand the second member of the pair, we need to use another distinction by
Novak, between ‘ontological’ and ‘metaphysical’ (where Novak identifies
metaphysics and teleology). On the second, ‘pre-Judaic’ conception the Noa-
hide is the non-Jew ‘situated in the natural order of creation itself, who the
Jews themselves had been before being chosen by God and given his Torah’.
But Novak says that we need to understand ‘the natural order of creation’ here
(using the Greek terms) not as a telos (a goal or end) but as a peras (a limit).74
Understanding our nature as a telos was the mistake of Maimonides, and
belongs, in Novak’s distinction, to a teleological metaphysics.75 Understanding
our nature as a peras is still ontological, since it starts from what we are, but it
does not derive our nature from the goal or end of our life and action. Instead,
it places our nature as a limit on what God can command.
The three teleologies that Novak rejects are those he attributes to Saadiah,
Maimonides, and Kant. The first purported teleological error is that of Saa-
diah, probably Maimonides’ most important predecessor in Jewish philoso-
phy.76 On this view, humanity is the telos of creation, and we know what this

law and inspires them to learn that law when they are convinced of God’s wisdom and
beneficence.’
73 74
Natural Law, 156–8. Ibid. 164.
75
Novak is inconsistent in his terminology here, but the problem is merely terminological. At
Natural Law, 158, he calls his preferred view ‘not teleological’, but on p. 126 he says that his view
is, along with the three he rejects, a teleological option.
76
When Maimonides objects to those who talk about ‘rational commandments’, it is prob-
ably Saadiah that he has in mind.
244 God’s Command
end is by looking at the goods that we naturally pursue. This is the error,
Novak says, of thinking that our relation to God is only through creation. He
reports Saadiah as holding that ‘humans receive everything from God by
means of the world. . . . Thus the human relationship with God is not only in
the world, a point common to any theology of creation, but it is always through
the world as well.’77 The important thing for present purposes is not to
determine whether Novak is accurate in his portrayal of Saadiah, but to see
the role of Saadiah in Novak’s dialectic. The first purported error is giving the
world-as-a-teleological-system an exclusive mediating role in our relation
to God.
The purported error of Maimonides is to identify the human end too
exclusively with contemplation (in Greek theoria). Maimonides inherited
this error, Novak says, from Plato and Aristotle. It is true that the conclusion
of the Guide is puzzling if we interpret Maimonides this way. Maimonides
writes: ‘The way of life of [a perfect] individual, after he has achieved his
apprehension [of God], will always have in view loving-kindness [hesed],
righteousness [tsedaqah] and judgment [mishpat], through assimilation to
His actions, may He be exalted.’78 Hermann Cohen takes the emphatic
placement of this declaration, at the very end of the whole work, to mean
that Maimonides is taking what Cohen thinks is a fundamentally Kantian line
about the relation between theoretical and practical reason.79 We do not need
to settle this here. There is a very similar problem in understanding Aristotle,
but it is the other way round. In Aristotle much of the text in the Nicomachean
Ethics teaches the primacy of practice in the best human life, but the conclu-
sion teaches the primacy of contemplation.80 In Maimonides’ Guide much of
the text teaches the primacy of contemplation, but the work ends with
practice. Novak agrees with those who find the emphasis of the text is on
the activity of intellect, and he thinks this emphasis is too rationalistic.81 He
also holds that Maimonides’ teleology requires a cosmic teleology that con-
temporary science has made incredible.

77 78
Natural Law, 130. Guide, iii. 54, 638.
79
See Hermann Cohen, Religion of Reason out of the Sources of Judaism, trans. S. Kaplan
(New York: Ungar, 1972), 240–1.
80
I have discussed this matter, and the relation to the Eudemian Ethics, in God and Morality,
ch. 1.
81
Novak also argues that Maimonides gives us ‘a uniquely Jewish idea’ that the human
person ‘participates in divine creativity first by deriving his or her wisdom from that creativity
through contemplation of its effects, and then creatively applying it in establishing the city of
God on earth’ (Natural Law, 120). It is not clear to me how this praise of Maimonides fits with
Novak’s condemnation of his metaphysical teleology, according to which ‘the operation of
[human] intelligence in its fullest and highest sense intends a vision of God not mediated by
the world or anything in it’ (ibid. 133). The charitable interpretation here is that the relation
between these two sentences has not yet been worked out, because they occur in different
chapters (4 and 5) that have their origin in different periods of Novak’s authorship.
 Divine Command in Some Recent Jewish Thinkers 245
The third purported teleological error is that of Kant, or perhaps we should
say of Maimonides again, but this time in Cohen’s reading of him. The error is
the opposite of the second error, and identifies the human end too centrally
with practice. Novak’s Kant (like Fox’s Kant) is very different from the Kant
whose voice has been so strong in this book. For Novak, Kant teaches that
human autonomy is a substitution of our will for God’s and that the human
will is essentially good.82 Both of these claims about Kant seem wrong. We
have to recognize our duties, Kant says repeatedly, as God’s commands, and
we have to recognize also that the evil maxim that subordinates duty to
happiness is innate (as is the predisposition to good).83 What matters for the
present chapter, however, is not the demonstrable errors in Novak’s and Fox’s
portrayal of Kant, but the role he is made to play in their dialectic. For Novak,
Kant represents the error of making justice (and moral value more generally)
stand over God, an error that he thinks is typical of those who propose a ‘God
of the philosophers’.84 Novak wants to insist, in contrast, that ‘God is subject
to nothing higher than his own will’.85
Novak rejects all three teleological views, and he thinks all teleology is going
to fall into one of the errors he has just listed.86 He proposes that, instead, we
should think of human nature as giving us a peras (limit) rather than a telos,
and then he suggests that we can find this limit in the Noahide laws. What
kind of limit would this be? He wants to use the notion of respect for human
dignity as a constraint on anything we can properly take to be a command of
God. In that sense, it is a ‘precondition’. Novak claims that the sense of
‘precondition’ here is similar to what Kant means by the term, when he says
that space and time are ‘conditions’ (Bedingungen) for sensory intuition.87 But
this cannot be right, because space and time are precisely not components of
intuition, nor are they abstracted from it. The Noahide laws are, in a sense,

82
Ibid. 45, 31, 166.
83
I have discussed these views of Kant in God and Morality, ch. 3.
84
An important point of view on this question is that of L. E. Goodman, in The God of
Abraham (Oxford: Oxford University Press, 1996), p. vii, who disagrees with Pascal’s opposition
of the God of Abraham, Isaac, and Jacob to the God of the philosophers. Goodman says: ‘In
Jewish tradition, the God of Abraham is the God of the philosophers.’ He endorses the view of
Maimonides ‘that Abraham is the prototype of the natural theologian, whose idea of God is
cemented by the inner affinity of the human mind with God himself ’. Goodman and I were
fellow Gifford lecturers in 2005, and in Love Thy Neighbor as Thyself (Oxford: Oxford University
Press, 2008) he gives a generous response to my views. He objects to what he takes to be my view
and the view of Kant, that ‘balks at ascribing [good human] potentials to our nature while we
remain unredeemed’, and ‘packages only weaknesses in our idea of human nature, reserving all
the strengths to the work of grace’. But it is not my view, or Kant’s, that our nature is only evil.
Our nature is a mixture, though the evil maxim is innate. We are born both with the predis-
position to good and the propensity to evil, and only the first of these is essential to us.
85
Natural Law, 43.
86
He does not defend this premiss, and a defence would require looking at the contemporary
recovery of teleology, especially Rosenzweig’s teleology, to see if it fails in one of these three ways.
87
Natural Law, 186, citing Kant, KrV B72.
246 God’s Command
components of the Torah. They are not included as such, being formulated by
the Rabbis. But item by item (at least in the reduced form that Novak lists
them) they can be found there. In this sense they are a precondition for the
Torah, but only in the very limited sense in which a component is a precon-
dition for the whole: for the set of letters of the alphabet a–z, the labial
consonants are a precondition. Novak wants them as a precondition in a
much stronger sense. He wants their commonality to be ‘immediately evident’
to all the other nations of the world, but most importantly he wants them to
serve as ‘minimal standards for Judaism’s moral justifiability’.88
There are two problems with this proposal. The first is that the Noahide
laws give us a great deal more than merely human dignity, and the second is
that they also give us less. They give us a great deal more. They give us
institutions like private property (presupposed in the prohibition on robbery)
and marriage (presupposed in the prohibition on adultery) and a judiciary
(presupposed in the requirement of a judicial system). The burden of
Chapter 4 was an attack on various forms of deductivism, and an argument
that the idea of human nature does not itself license a deduction of such
institutions. To put this eschatologically, we are fully human in heaven, and we
do not know that any of those institutions will be present there.
But the Noahide laws also give us less than human dignity. If we are to
identify them, as Novak does, with natural law, and if we are to identify natural
law with ‘norms of human conduct that are . . . discernible by all rational
persons’, this presents us with a dilemma. When we look at certain parts of
our human past, we do not see a respect for human dignity in place. Human
dignity turns out to be, in Novak’s conception, a highly demanding norm. He
takes a basically Kantian view of it, with a transformation suggested by
Emmanuel Levinas: I am to treat the other person as ‘essentially an analogue
of my own fully conscious moral personhood’ so that we become both
together members of a ‘kingdom of ends’.89 The contribution of Levinas is
to see that ‘the object of my moral concern presents himself or herself to me
before I have constituted myself as a moral subject’. But hunter-gatherer
societies, which comprise most of human history, do not seem (as far as we
can tell from present examples, which is not very far) to recognize the equal
fundamental value of all human beings, even though they have (like some
non-human primates) some sense of fairness in distribution within the group.
They distinguish between the value of the members of their own tribe and the
value of everyone else. The same is true of the relations of Greeks and
barbarians in ancient Greece. It is important, then, that the Noahide laws in
their original form have different treatments prescribed for Jews and non-
Jews. We will return to this in Chapter 8. It is open to Novak to say that not all

88 89
Natural Law, 163. NLRT 7. Natural Law, 166.
 Divine Command in Some Recent Jewish Thinkers 247
humans are in his sense ‘rational’, and so they do not all have access to natural
law, but this goes against the spirit of his proposal.
To sum up this subsection of the chapter, Novak tries to make the Noahide
laws do work for which they are unsuited. They are too determinate to be
deduced from human nature and to serve as a precondition for taking any
commandment as a commandment of God, except in the very limited sense in
which any proposed divine commandment has to be consistent with what we
already take to be divine commandments as those have been revealed to us.
The question here is not whether there is general revelation by God to human
beings. There is, though the Noahide laws are too determinate to qualify. The
prohibition on theft, to take Scotus’s favourite example, presupposes the
institution of private property, but there are now and there probably always
have been human societies without this institution. There are human societies
in which there is no king and so no judge, but people do that which is right in
their own eyes (Judges 21: 25). Private property and judicial process have to be
learnt. But the key question is about authority. Do we take some set of natural
laws known from ‘natural necessity’ as the final justification, the standard
against which any purported divine revelation has to be measured? Should we
then obey the commandment because it is consistent with this set? For
someone who believes that God has given us a special revelation as well as a
general revelation, the relation is better the other way round. We should trust
the natural laws unreservedly because they are validated by special revelation.

7.3. FRANZ ROSENZWEIG

7.3.1. Introduction

Franz Rosenzweig’s The Star of Redemption is a deeply difficult text.90 This is

partly because it is frequently elliptical. The ambition of the author is extra-
ordinary in scope, but this means that he moves swiftly, and the price is that
fairly often the reader finds herself asking: ‘Why on earth does he say that?’
Nonetheless the text is rewarding, and I hope to give a sense of this in what
follows, even though my account will be even more telescopic than his text is.
I also hope to show the continuity of The Star with a tradition of reflection on
the topics it addresses, and I will do this by naming and quoting some of the
figures in this history (usually in the footnotes). Often the ellipsis results from
the work’s failure to name the people and the views to which it is responding.

90
Franz Rosenzweig, The Star of Redemption, trans. William H. Hallo (Notre Dame, IN:
Notre Dame Press, 1970) (henceforth Star).
248 God’s Command
Rosenzweig comes out of the same philosophical milieu as Barth, and was
strongly influenced (as was Barth) by Hermann Cohen. Barth studied with
him at Marburg, and then Cohen moved to Berlin in 1912, where Rosenzweig
was a student. Unlike Barth’s work, however, The Star of Redemption shows
the influence of Maimonides pervasively, even in its structure, while it does
not mention Maimonides very often. Like David Novak, Rosenzweig is in part
mounting a critique of Maimonides, but, whereas Novak failed in his stated
purpose to avoid a reduction of revelation to creation, Rosenzweig succeeds.
One central theme comes in the following quotation:
Only revelation [and not creation] has the knowledge—and it is the primary
knowledge of revelation—that love is as strong as death. . . . Revelation wakens
something in creation that is as strong as death and sets it up against death and
against all of creation. The new creation of revelation is the soul, which is the
unearthly in earthly life.91
This quotation is not self-explanatory. The context is that Rosenzweig is
opposing Maimonides’ view on emanation or ‘influence’. God does not always
act by ‘overflow’, but God acts in love, and that is something different; our love
is then a response to God’s love. The goal of this section of the chapter is to
unpack this portmanteau of ideas.
Rosenzweig’s relations to Christianity and to Islam are important in
understanding his book. According to N. N. Glazer, in his foreword,
Rosenzweig almost became a Christian, and had decided to be baptized,
though he insisted that he ‘could turn Christian only qua Jew’.92 But, on the
Day of Atonement service in Berlin in 1913, he decided not to go ahead.
The quotation in the previous paragraph comes in his description, to me the
most moving part of the book, of the ceremonies and clothing for the Day of
Atonement. The book is full of remarkable insight about the relation of
Judaism and Christianity. On the other hand, Rosenzweig’s attitude to Islam
is one of hostility and contempt. The passages about Islam in part two are an
embarrassment. We can learn, though, about what he finds good in Judaism
and Christianity from his defamation of Islam and the role he gives to that
faith in his overall structure.
This chapter will not comment on the whole of Star. It will omit part one,
which is the most obscure of the three parts, encouraged by Rosenzweig’s own
remark: ‘He who understands Part One does not need the rest—and vice
versa.’93 Most of the commentary will be about part two. The overall scheme of

91
Ibid. 326. Rosenzweig is quoting ‘Love is as strong as death’ from Song of Songs 8: 6.
92
Star, p. xi. But this story, though commonly told, has been recently put into doubt. See
Benjamin Pollock, Franz Rosenzweig’s Conversions: World Denial and World Redemption
(Bloomington, IN: Indiana University Press, 2014). Rosenzweig does not mention the climactic
event in any of his abundant letters of the period.
93
Star, p. xiv.
 Divine Command in Some Recent Jewish Thinkers 249
the work is that part one (figured as an equilateral triangle with its base
horizontal) distinguishes God, Man and World. Part two (figured as an
equilateral triangle with its top horizontal) distinguishes creation, revelation,
and redemption. Part three then superimposes the two triangles to form a star,
thus giving the work its title. The three following subsections discuss the three
doctrines distinguished in part two (creation, revelation, and redemption), one
in each subsection.

7.3.2. Creation: The Disappearance of God

‘God’s creating is the beginning of his self-expression,’ Rosenzweig says at the

beginning of part two.94 This is not going to be true for a Christian, for whom
the second person of the Trinity is already logos (word) before creation.95 But
Rosenzweig wants to distinguish between the concealed God and the manifest
God, where it is creation that makes God manifest, and expression must be to
the outside of the One.96 He is here in conversation with Maimonides, and he
both agrees and disagrees with him. He agrees with him that creation should
not be assimilated to divine caprice: ‘And so it was precisely in this point that
Maimonides, the great theoretician of revelation, diverged from Arabic Scho-
lasticism and, with utmost decisiveness, asserted God’s creativity as his essen-
tial attribute.’97 But Rosenzweig insists that we still need the idea of divine
caprice in order to save God’s inner freedom, which is threatened by a
Maimonidean doctrine of emanation or ‘influence’. The solution to this
difficulty is to see that there is a way to avoid making creation either a need
of God (what Rosenzweig calls ‘yearning love’) or a necessity for God (what he
calls ‘overflowing love’). The solution is that ‘there is caprice, not in the
creator’s act of creation, but prior to it in the self-configuration of God
which precedes his act of creation. . . . But as “manifest” God he cannot do
otherwise than to create. Those who ascribed inner, substantive necessity to
the divine creative act are right as against those who assert its capricious-
ness.’98 Rosenzweig endorses the rabbinic suggestion that God created the
world in righteousness rather than in love, because the power that actualizes

94
Ibid. 113.
95
There are, to be sure, the ten ‘utterances’ from the Genesis account of creation that are held
by the Rabbis to precede the creation, and could be called the logos by which creation occurs. This
is probably what Rosenzweig means by talking of God’s ‘self-configuration’. But for Christians
communication precedes creation.
96
Star, 159.
97
Ibid. 115. Note that he here takes just one strand in medieval Islam, the Ash‘arite strand, to
stand for the whole.
98
Ibid. 116.
250 God’s Command
creation ‘provides its mettle precisely in the generation and execution of
justice. Caprice is the explicit antithesis of such power.’99
What kind of world does God create? Rosenzweig emphasizes, in a way
reminiscent of Barth’s focus on the particular (discussed in Section 5.1.1), the
distinctiveness of each created thing. Creation, he says, is ‘Existence in contrast
to Being’.100 These are terms from Rosenzweig’s philosophical milieu. One is
reminded also of Heidegger’s attack on metaphysics. The postulated opponent
of all these thinkers is the idealist who wants to impose ‘sameness’ or ‘totality’
on experience because he thinks he is licensed by the identity of being and
thought. Idealistic metaphysics tries to turn the ‘chaos of the distinctive’ into
cosmos by thought.101 Applied to ethics, the result is that our moral lives are
seen as ‘submission to ever higher communities, ever more inclusive general-
ities of life’.102 What happens when this idealist project is undertaken is that
God disappears into creation, where ‘he could modestly hide behind eternal
laws’.103 We are left not with a person who speaks to us, but a mere ‘thing-in-
itself ’ or ‘Absolute Spirit’. But ‘absolute personality’, Rosenzweig says, ‘is a
contradiction’.104
This critique of idealism can be generalized to any attempt to locate the
ground of the moral law in creation. The tendency of any such attempt is to
lose God as a person who speaks to us. One example is the ‘new natural law’ of
John Finnis. Finnis claims to provide a ‘purely philosophical’ version of
natural law, one that does not depend upon theological premisses, yet also
one that gives us ‘exceptionless specific norms’ binding on every human
being.105 It is not that God is absent from Finnis’s system. In Natural Law
and Natural Rights ‘religion’ comes in as the seventh of the basic values that lie
behind rational choice, and at the end of the book he uses an argument from
natural theology for the existence of an uncaused cause, which he calls ‘D’;
one could pursue friendship with D, if D turns out ‘to favour the well-being of
everyman, for no other reason than [its] own goodness’.106 But it is hard to

99
Ibid. 117. Most Christians will not agree that God created in righteousness rather than
love. See, for just one example, Jean-Luc Marion, L’Idole et la distance: Cinq études (Paris:
Éditions Grasset and Fasquelle, 1977), who argues that all that is—and the fact that anything is—
comes from the Gift of a Lover. Nonetheless it is perfectly possible for Christians to suppose both
that creation is an act of love and that revelation is a different act of love, different in kind
because the recipient is already existing and is put by the revelation into relation with God.
100 101
Star, 120. Ibid. 141.
102
Ibid. 142. See the discussion in Section 3.3.3 of Jean Porter’s account of a nested hierarchy,
though she is not an Idealist.
103 104
Ibid. 162. Rosenzweig is quoting Schiller. See also Star, 220, 258. Ibid. 144.
105
John Finnis, Aquinas: Moral, Political and Legal Theory (New York: Oxford University
Press, 1998), pp. vii, 94, 164, 294. I am grateful to Neil Arner for this reference, and the reference
to John Rist. Another work of this kind, under Finnis’s influence, is Mark Murphy, Natural Law
and Practical Rationality (Cambridge: Cambridge University Press, 2001). This characterization
is not true of his subsequent work.
106
John Finnis, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980), 387, 406.
 Divine Command in Some Recent Jewish Thinkers 251
resist the impression that this reference to God is paradoxically both vestigial
in the explicit theory and doing a great deal of work as an unstated assump-
tion.107 In the work of Philippa Foot and Rosalind Hursthouse discussed in
Section 4.4.3, there is an argument from nature to morality that does not have
a place for God at all. Not all natural law theories fall prey to this temptation,
but it is a real temptation of this kind of theory. If we can get straight to the
moral law from human nature, it can seem that it is optional whether we
should go on to enquire what gives us human nature or nature as a whole.
Rosenzweig was prescient about this. He attached the critique to German
idealism, but it works as a criticism of any system that focuses on the
immanent relation of our obligations with our world, and does not emphasize
at the same time their transcendent relation with the person who finally makes
something obligatory by commanding it.

7.3.3. Revelation: Initiative and Response

Here we come back to love strong as death. Rosenzweig starts from the fact
that all created physical things die. This is different from the standard Chris-
tian account that death came into the world through sin, but there have been
Christian voices even from the patristic period saying that death came with
creation.108 It is certainly hard to see how our present understanding of
ecology can allow predators without prey, for example, and the death of
animals is massively implicated in the biological world as a whole. The
important thing for present purposes is that God introduced something new
into this cycle of birth and death, whether we think of this cycle as originally
created or the result of the Fall. Rosenzweig suggests that God, who could have
been mere origin and bare creative power, chooses to be unconcealed in a new
way. Chapter 8 will retell the Genesis story with this in mind. It will suggest
that we can imagine humans, with cognitive equipment newly differentiated
from their primate ancestors, now able to receive God’s self-revelation. What
will this revelation be like? Rosenzweig’s suggestion is that we will find
ourselves loved by God, and this frees us from our state of merely being
created.109 As Barth emphasized particularity, so Rosenzweig stresses that
God’s love is an event (not, like God’s creativity, an attribute), a particular
disclosure at a particular time to a particular person. It ‘transfixes’ or ‘lights up’
the individual in whatever that individual’s current state might be, and is not

107
This is the complaint of John Rist in Real Ethics: Reconsidering the Foundations of
Morality (New York: Cambridge University Press, 2002), 260.
108
Augustine, The City of God against the Pagans, trans. R. W. Dyson (New York: Cambridge
University Press, 1998), xii. 4.
109
Star, 161.
252 God’s Command
conditional on the individual’s merit. Here again Islam is the foil. ‘The idea that
man’s shortcomings are more powerful to arouse divine love than his merits—
the conceptual nucleus of belief—is an inconceivable paradox for Islam.’110 But
there is an implicit correction here also of Maimonides. Rosenzweig thinks of
Maimonides as having a God too much like Aristotle’s God, a God who (or
which) is indeed loved (who moves everything, Aristotle says, ‘by being loved’)
but who (or which) does not himself (or itself) love.111 It is not clear how there
could be a covenant, a mutual and reciprocal relation, with such a God.
The awareness of being loved in this way produces a change in us. It is the
essential part of what Rosenzweig calls a change from ‘self ’ to ‘soul’. He
describes it in four stages. The initial stage is one of self-enclosure. We then
become aware of being loved by God, and the first response is one of defiance.
Freedom wants to be its own master. But the third stage is one that could be
called both a form of humility and a form of pride.112 It is humility because the
self is ‘conscious of being what it is by the grace of a superior’. But it is also
pride because the self is aware of being sheltered by this love, so that no power
can destroy it. At the fourth stage, it chooses to allow itself to be loved, and this
becomes an essential attribute, the attribute of faithfulness. How can it be an
essential attribute of the beloved, when it seems to be merely passive, merely
being loved by the lover? Rosenzweig answers by saying that the second stage,
the defiance, persists, but it is converted, or turned to the other side. ‘Defiance
is the arch-evil in man, bubbling up darkly; it is the subterranean root whence
the juices of faithfulness rise into the soul beloved of God.’113 The very energy
that had opposed God initially is now turned to trusting him, and the
faithfulness is the other side of the same coin as the defiance. In this way the
loving that we do is not merely passive, a receiving of love, but an active
reciprocation. This idea of an initial fight with God resonates strongly with
some of the traditions from the Rabbis quoted by Christine Hayes in What’s
Divine about Divine Law?.114 There is a Christian parallel in some of the
poetry of George Herbert—for example, ‘Bittersweet’:115
Ah my dear angry Lord,
Since thou dost love, yet strike;
Cast down, yet help afford;
Sure I will do the like.

110 111
Ibid. 166. Aristotle, Metaphysics XII. 7. 1072b3. 112
Star, 168.
113
Ibid. 170. There is a Christian version of this suggestion in Stephen Verney, Water into
Wine (London: Fount Paperbacks, 1985), who gives a reading of John 10: 17–18 in terms of
leaving an ego-self by command and receiving it back transformed as a divine human self.
114
Hayes, What’s Divine about Divine Law?, 320–3.
115
See John Drury, Music at Midnight: The Life and Poetry of George Herbert (Chicago:
University of Chicago Press, 2014), 360.
 Divine Command in Some Recent Jewish Thinkers 253
I will complain, yet praise;
I will bewail, approve;
And all my sour-sweet days
I will lament, and love.
Rosenzweig again has a target in describing the development in this order.
Hermann Cohen asks: ‘Does God first love man, or does man first love the
unique God?’116 Rosenzweig has just told us firmly that God first loves us. But
Cohen answers his own question obliquely in the following way: ‘Only now,
after [nachdem] man has learned to love man as fellowman [Mitmensch], is his
thought turned to God, and only now [jetzt erst] does he understand that God
loves man.’ He here says that our love of God comes after our love of human
beings, and is not a response to the prior experience of being loved by God.
Cohen is here giving what he thinks is Kant’s view, and is also his own, and, as
discussed earlier, he attributes a similar view about the priority of the practical
to Maimonides (based on the end of the Guide). A great deal hinges on what
kind of priority we are talking about. Kant does not mean, in the view of him
presented in my book, to deny the need for God’s prevenient grace. When he
says ‘the right way to advance is not from grace to virtue but rather from virtue
to grace’, he is talking not about the order of the development of true virtue in
us, but the order of practical faith.117 He thinks we should start from the fact of
reason, that we are under the moral law, and move to the postulation of
freedom, God and immortality. What he is against is sitting around waiting for
grace; we should, rather, get moving in doing our duty, and have the faith that
God will have already supplied the means both internal and external. Cohen,
by contrast, is talking about the development of virtue in us, and saying that,
once virtue is in place (loving man as fellowman), we are then ready for God’s
assistance. Rosenzweig is here also talking about the order of development, but
he claims that in this order we have to assume that God’s loving and our
experience of it come first.
This order of development is strikingly illustrated in a reading of Genesis
3: 9: ‘And the Lord God called unto Adam, and said unto him, “Where art
Thou?”’ Rosenzweig assumes that the calling to Adam is a calling by name, so
that what God actually says is: ‘Adam, where art thou?’ On this reading Adam
discovers an ‘I’ by being addressed as a ‘Thou’. Rosenzweig points out that this
would be the first time that Adam’s name is used in an address. The emphasis
here on the name can remind us of a theme in Chapter 1 from Scotus and in
Chapter 5 from Barth, and the emphasis on the ‘Thou’ is reminiscent of
Darwall on the second person, discussed in Chapter 2. The self-enclosed self
is silent and unnamed. In the Genesis story Adam and Eve try to hide

116
Cohen, Religion of Reason out of the Sources of Judaism, 169–71. See Natural Law, 84.
117
Rel. vi. 202.
254 God’s Command
themselves. When they cannot succeed in this, the answer they give to God is
all in third-person terms: ‘The woman did it’ or ‘The snake did it’. But by the
direct divine address, ‘man is cut off from every retreat into hypostatization.
The general concept of man can take refuge behind the woman or the serpent.
Instead of this the call goes out to what cannot flee, to the utterly particu-
lar . . . to the proper name.’118 Rosenzweig imagines Adam finally responding,
‘Here I am’, as Abraham responds to God’s command to go up Mount
Moriah.119
What God now gives to the subject who is ‘as yet wholly receptive, as yet
only unlocked, only empty’ is a commandment—namely, the commandment
‘Thou shalt love the Lord thy God with all thy heart and with all thy soul and
with all thy might’.120 ‘But in what sense can love be commanded?’, Rosenz-
weig asks (like Kierkegaard, and like Kant).121 His answer is that the com-
mandment is ‘none other than the voice of love itself ’, since love by its nature
calls for a response, and indeed an immediate response.122 This immediacy of
the requirement means that the love commandment is not a law, as Rosen-
zweig conceives of law, and certainly not a natural law. Laws, he thinks, have
reference to an ‘ever’, as in: ‘Don’t ever spit out of the window.’ The love
commandment requires a response now. Rosenzweig refers to Psalm 95: 7,
‘Today if ye will hear my voice, harden not your hearts.’ In the language of
Section 5.1.3, this means that the term in the action position of the imperative
sentence is not universal, since it is temporally indexed. This seems too
restrictive. God, we can say, addresses us now and requires a response now,
but this is consistent with the same response being now required for future
occasions of the same kind. Perhaps the issue here is not a serious one, since
Rosenzweig thinks God’s love is present at each moment, and indeed ‘love
loves the beloved each day a little more. This constant augmentation is the
form which steadfastness takes in love, although—and because—it is instabil-
ity itself.’123 We need a way to say both that the divine love is new every
morning and that it is constant, and Rosenzweig has a way to say both. The
initial response of the subject to the command is shame, Rosenzweig says, and
the sense of a past that has failed. But, as with the discussion of humility
earlier, there is a kind of pride associated with this sense of shame, and that is
the pride in being loved.

118
Star, 175–6. See John 10: 3: ‘He calls his own sheep by name.’ Note that Rosenzweig’s
account of the Fall does not have any reference to any ‘original righteousness’ in Adam and Eve.
119 120
Genesis 22: 1. Deuteronomy 6: 5.
121
See Kierkegaard, Works of Love, 24: ‘this apparent contradiction: to love is a duty’. See
Kant, KpV v. 83: ‘[Pathological love] cannot be commanded, for it is not possible for a person to
love someone merely on command.’ See John E. Hare, ‘Kant on Practical and Pathological Love’,
in Frederick Simmons (ed.), Love and Christian Ethics (Georgetown: Georgetown University
Press, 2015).
122 123
Star, 176–7. Ibid. 163.
 Divine Command in Some Recent Jewish Thinkers 255
We are now ready for the second part of the commandment, the require-
ment to love the neighbour as the self, and this takes us to the territory
Rosenzweig allots to redemption, and so to book three of part two. Merely
to acknowledge oneself as loved and to reciprocate with love for the lover is, he
says, to be ‘waiting, not walking’.124 He is repeating here a critique of mysti-
cism that is strong in Kierkegaard and Kant as well, that it carries with it the
danger of staying to bask in the experience of intimacy.125 God’s loving
requires not just the response of loving God, though it starts with this, but
also the response of loving the neighbour. How does Rosenzweig show this?
He argues that the neighbour is in the image of God, and, if we love God, then
we will love what is in the image.126 Love of neighbour ‘surmounts’ mere
dedication to God, while at the same time always presupposing it.127 More-
over, if we love God we will love what God commands us to love.128 Again,
though, as we have seen, Rosenzweig says that this love of the neighbour as
obedience is possible only after we have been the recipients of God’s love to us,
and reciprocated it.129
Finally, Rosenzweig emphasizes that the object of our love, the neighbour, is
the nighest neighbor, and we have to be careful not to skip over this one in
order to get to the neighbour we prefer.130 He puts this in terms of the
‘indefinite’, because it is anyone whom I meet, but he says that this indefinite
is also definite, with a definiteness that is ‘superior, not subordinate to it’.131 To
understand this, we have to see that love of the neighbour is the sign of the
kingdom towards which we are headed. It is only in the kingdom, only when
the ‘all’ is in accordance with God’s commandments, that the individual can be
completely definite. This is because the definiteness in question comes from a
relation to the world as a whole in all its interconnectedness. Only if all
are good can the individual be completely good.132 There is a difficulty here.
For it is also true that only if the individual is good can the world be good.

124
Ibid. 208.
125
See ibid. 207: mysticism ‘becomes a cloak of invisibility for the mystic’. See also Kierke-
gaard, Works of Love, 15, ‘that love is such a hidden feeling that it is too exalted to bear fruit’, and
see pp. 47, 97–9. See also Kant, Rel. vi. 201: ‘If the delusion of this supposed favorite of heaven
reaches heights of enthusiasm, to the point of imagining that he feels the special effects of faith
within him (or even has the impertinence of trusting in a supposed hidden familiarity with God),
virtue finally becomes loathsome to him.’
126 127
Star, 259. Ibid. 214.
128
Ibid. 214. Kierkegaard puts this by saying that the love of God has a forwarding address:
Works of Love, 161: ‘But as you bring [your love] to Him you immediately receive, if I may put
this way, a notice designating where it should be delivered further.’
129
Star, 215. See Kierkegaard, Works of Love, 58: ‘Only by loving God above all else can one
love the neighbor in the other human being.’
130
Star, 235. Rosenzweig makes the point from the Hebrew and the Greek, and the Latin
proximus is a superlative (not just ‘near’ but ‘nearest’). See Kierkegaard, Works of Love, 51: ‘the
very first person you meet is the neighbor.’
131 132
Star, 236. Ibid. 228.
256 God’s Command
Rosenzweig solves the difficulty in the same way as Kant. What Kant calls an
‘ethical state’ or ‘a kingdom of virtue’ seems to have ‘an objective reality in
human reason (in the duty to join such a state), even though we cannot
subjectively ever hope of the good will of human beings that these will work
harmoniously toward this end’.133 The solution to this difficulty is to see that
we need ‘the presupposition of another idea, namely a higher moral being
through whose universal organization the forces of single individuals, insuf-
ficient on their own, are united for a common effect’.134 For Rosenzweig:
‘[Humans] cannot deliver themselves by themselves from each other; they can
only be delivered together with each other—delivered by a third one, deliver-
ing one on the other, one by means of the other. Besides man and the world,
there is but One who is third; only One can become their deliverer.’135

7.3.4. Redemption: Revelation and Creation Seen Backwards

The final subsection of this chapter makes two points. The first is that we get a
proper understanding of creation and its law only backwards, through re-
demption and revelation. This will help us grasp the proper relation between
love of the neighbour and natural law. The second is that we can revisit the
three arguments from Chapter 1 (the argument from providence, the argu-
ment from grace, and the argument from justification), and see them afresh by
rephrasing them in Rosenzweig’s language. The chapter will end with a brief
summary of what a divine command theorist can learn from Rosenzweig, just
as the previous chapter ended with a brief summary of what the divine
command theorist can learn from al-Maturidi.
The first point is that the created world is, contra Novak, best understood
teleologically, and so is only known as what it is by what it is at the end.
Rosenzweig says that the world is facing the opposite way from God and from
man. ‘Thus God “first” created and “then” he revealed himself; man “first”
received the revelation and “thereupon” prepared himself for the world act
[redemption].’ Both for God and for humanity, that is, the once-and-for-all
occurrence preceded that which happens moment by moment. The idea is that
God is what God is in God’s self-configuration before creation, and human
beings are what they are as loved by God and reciprocators of that love before
redemption. (This has to be distinguished from what human beings are as
created, which is, like the created being of the world, to be understood
teleologically, in terms of what it is headed towards.) The ‘once-and-for-all
occurrence’ for God is creation, though to put it this way is misleading, since it
implies that there is no continuous creation. Rosenzweig does not mean to

133 134 135

Rel. vi. 95. Ibid. vi. 98. Star, 228.
 Divine Command in Some Recent Jewish Thinkers 257
deny that God is continuously active in creation, and indeed he needs this
premiss, as we will see when looking at his version of the argument from
providence. But he is talking in the present passage about the statement ‘In the
beginning God created’ in the first verse of the Bible. Creation is then followed
by the moment-by-moment divine loving that Rosenzweig associates with
revelation. The ‘once-and-for-all occurrence’ for human beings is what Chris-
tians (or at least Lutherans) would call ‘justification’.136 Rosenzweig calls it
being ‘lighted up’ or ‘transfixed’ or ‘vitalized’.137 Justification, on the Lutheran
picture, is then followed by sanctification, which is a moment-by-moment
working-out of this fundamental change of status in the successive choices we
make during our daily lives.
For the world, on the other hand, this temporal relationship is ‘reversed’.138
The decisive change is not a working-out of what is already implicit internally.
The state of moment-by-moment change, the ‘chaos of the distinctive’, is the
initial condition: ‘While God and man are older in essence than as phenom-
ena, the world is created as phenomenon long before it is redeemed for its
essence.’139 The error of the idealists is to suppose that they can turn this chaos
into a cosmos by thought. Actually it is only God who brings the kingdom: ‘He
consummates. He does it. He is the Redeemer.’140 The world as created, as
‘nature’, is mysterious, because its end and so its essence are still hidden.
Rosenzweig says that this mystery is what we mean by saying that the world is
‘enchanted’ or that it is ‘uncanny’ (unheimlich). As created, the world has not
yet found what he calls its ‘life’ or ‘eternal life’ or ‘soul’ or ‘immortality’.141 The
tendency of modern science, he says, is to reduce the world to creation, but
without this sense of mystery, and so to ‘disenchant’ it. A disenchanted world
is one without soul, and so without ‘the seeds of name, of animated individu-
ality, of immortality’.142
What does all this tell us about the relation between revelation and natural
law? To suppose that we can tell how we ought to live just by looking at what
we can observe of nature would miss the heart or the life that is the purpose of
creation itself. This is not to say that nature is irrelevant to how we ought to
live. Creation gives us the structures that are then redeemed, or the raw
material that is then configured. ‘Thus blood kinship, brotherhood, nation-
hood, marriage, in sum all human relationships are established in creation.
Nothing exists but it is at bottom primeval. All have their prototypes in the
animal kingdom, and through the rebirth of the soul in revelation, all are first
animated with a soul in redemption.’143 One way to explain this is that
Rosenzweig is here supporting Fox and not Novak in the understanding of

136
Kant translates justification into a ‘revolution’ of the will, which is noumenal, and
sanctification into ‘reform’, which is in time and subject to experience. See Rel. vi. 47.
137 138 139
Star, 162, 164. Ibid. 218. Ibid. 219.
140 141 142 143
Ibid. 238. Ibid. 222. Ibid. 241. Ibid. 241.
258 God’s Command
the dictum of Maimonides, ‘Therefore I say that the Law, although it is not
natural, enters into what is natural.’144 For Novak, the Law prescribes what is
already natural to us. For Fox, the Law fits what we are, but it is also something
radically new, something not deducible from what we already see of what we
are like. One way to put this point is that Rosenzweig’s account gives an
eschatological teleology to the world, but it is not one that faces any of Novak’s
three objections to teleology. It does not collapse the God-relation into the
world-relation (the purported error of Saadiah), or over-emphasize either
theory (the purported error of Maimonides) or practice (the purported error
of Kant).
If we return now (the second main point of this subsection) to the three
arguments of Chapter 1 (the argument from providence, the argument from
grace, and the argument from justification), we can use concepts and vocabulary
from Rosenzweig to make the same three points. It is important to the project of
this book that the same basic questions arise in each of the Abrahamic faiths,
even though the working-out of answers may differ. In every case the quoted
passage from Rosenzweig will have the tell-tale sign of the language of ‘must’ or
‘can only take place if ’, and in every case a ‘transcendental’ argument is being
given, an argument from the conditions of possibility for some state of affairs
antecedently recognized.
First, the argument from providence is that our obligation already presup-
poses that duty and happiness can in the end be conjoined, and only God can
accomplish this. We have already seen one passage from Rosenzweig that
makes this point, when he says ‘only One can become their deliverer’.145 Here
is another passage, from the very end of part two: ‘Thus the world appears to
direct its glance now to this now to that side, now to seek refuge in the eternal
arms of the Creator, now to expect everything from the earthy lord of creation
[the human being].’146 The problem is that redemption lies on neither side;
not from God as creator, because nature does not yet give us the command-
ment to love, and not from human beings, because civilization, though it
knows the commandment, does not know how to carry it out or produce the
kingdom which is its promised fruit.
The world appears fated to remain in this eternal contradiction. But is it? We
know that, for the world, the contradiction will vanish in the eternity of the day of
the Lord, in the redemptive advent of the kingdom. Thus it is not an eternal
contradiction. But how is this identification of human act and divine labor to take
place unless the human act itself also derives, as act, from God, and unless God’s
creative labor augments and fulfills itself in awakening man?
Note that in order to avoid the contradiction we have to postulate God’s
activity in two roles. God has to be acting through human action, in order to

144 145 146

Guide, ii. 40, 382. Star, 228. Ibid. 260.
 Divine Command in Some Recent Jewish Thinkers 259
produce the kingdom, so that when we act there is an ‘identification’ with
God’s act. And God has to be acting through nature, in creative labour, to fulfil
that creation in those actions that humans take. Rosenzweig thinks of God as
pushing the world forwards into growth, so that it ‘moves towards the act of
love’, even though we only see this in retrospect.147 Here we have the dual
structure of the argument from providence, where God has to be both author
of the law and ruler of nature, playing both legislative and executive roles in
bringing about the kingdom.
Second, the argument from grace is that we start with a propensity to evil, in
Kant’s terms in subjection to the evil maxim, even though we also start with
the predisposition to good, and we cannot by our own devices help ourselves
to escape from this condition. We have already seen part of a passage from
Rosenzweig that makes this point, but here is the whole of it: ‘Only the soul
beloved of God can receive the commandment to love its neighbor and fulfill
it. Ere man can turn himself over to God’s will, God must first have turned to
man.’148 The argument from grace in Rosenzweig’s language is that we start
self-enclosed, and it is only the experience of being loved by God that releases
us, first to defiance and then to acceptance and faithfulness. The doctrines of
the Fall are different here between the Lutheranism that Kant is translating
‘within the boundaries of mere reason’ and Rosenzweig’s Judaism. But the
need for God’s initiative is the same, since we cannot escape our initial
condition by our own devices. Perhaps it is worth mentioning that Rosenzweig
separates himself from Kant, saying that a Kantian autonomy gives us ethics,
but does not give us the commandment to love.149 This is because he takes his
Kant from Cohen, just as Fox and Novak do, and this Kant is not the Kant of
my book. But it is true that, even if we take ‘autonomy’ to mean ‘repeating in
our will God’s will for our willing’, we go beyond Kant if we stress the direct
experience of God’s love.
Finally, the argument from justification asks the normative question ‘Why
should I try to be morally good?’ and answers that the quest for justification
terminates in something that does not itself require justification, the truth that
God is to be loved. Again, we have seen part of a passage from Star that makes
this point, but the whole of it is helpful:
The commandment to love can only proceed from the mouth of the lover. Only
the lover can and does say: love me!—and he really does so. In his mouth the
commandment to love is not a strange commandment; it is none other than the
voice of love itself. The love of the lover has, in fact, no word to express itself other
than the commandment.150
Rosenweig is not particularly interested in asking the normative question. But
he does in effect give us a justification from something that does not itself need

147 148 149 150

Ibid. 240. Ibid. 215. Ibid. 214. Ibid. 176.
260 God’s Command
justification, a truth about the nature of love itself. When we find ourselves
loved, we find ourselves also with a call to love back. The love is not complete
until it is returned. But this is not merely like a questionnaire that we might
receive in the mail, which is also not complete until it is returned. Rather, the
divine love, even if it first occasions defiance, has an internal power (though
not an irresistible power) to make us want to return it. This is not always true
of human love, at least human love of some kinds, which can be merely
irritating and which does not have this internal power. But Rosenzweig’s
claim, and it is also the Christian claim, is that God’s love has the internal
power to bring out something in us, a love in response.151
In Chapter 6, we ended with a brief summary of what a divine command
theorist could learn from al-Maturidi. To end the present chapter, we can say
by way of parallel that there are three things a divine command theorist can
learn from Rosenzweig, one from the material covered by each of the subsec-
tions 7.3.2–4. First, Rosenzweig shows us the danger of the disappearance of
God into nature if we try to derive the moral law from nature or creation.
What we tend to lose is the personal divine lover whose command establishes a
relation of communication between us. And we can see in various natural law
accounts that this danger has come to pass. Second, Rosenzweig gives us a new
account of the relation between command and love. Love, he says, calls out a
response from the beloved, and in this sense we are commanded by God to
love God. As described just now in terms of the argument from justification,
the divine love, revealed in the command, has the power (but not the irresist-
ible power) to make us want to return it in kind. Third, we find in Rosenzweig
the view that creation, or nature as created, faces in the opposite direction
from God and human beings. Creation is not to be understood until its end, but
God and human beings are already intelligible in terms of the relation of
revelation between them. Rosenzweig is not here saying that we can understand
morality without the doctrine of creation, as though we could get it from
revelation alone. For him, and this is new in Star from his earlier view that he
called ‘gnosticism’, the categories of family and nation have to be introduced in
addition to personal redemption, and he thinks this is a characteristic Jewish
corrective to the characteristic Christian neglect of these categories. But, that
said, creation cannot give us the law by itself without the command, because it is
not yet sufficiently determinate. It is ‘uncanny’ until it reaches its destination.

151
Whether we have the power also to refuse to endorse this response is a different question,
and one debated between Christians. I gave my own positive answer to this question in
Section 1.3.
 8

Divine Command and Evolutionary

Psychology

I N T R O D U C TI O N

This book has been a defence of the claim that what makes something morally
obligatory is that God commands it, and it has added the claim that God’s
commands fit our human nature, even though they cannot be deduced from it.
The present chapter is about the implications of our evolution for this view of
the relation between God and morality. It distinguishes four possible ways to
think about these implications, and each will have its own section and cover its
own set of thinkers, though some of them belong in more than one of these
sections. The goal is to give a map of the conceptual territory. The chapter
rejects the first three ways and endorses the fourth. The first proposal is that
evolution has given us a set of moral capacities that are limited to parochial
attachment, and that any morality (like Kant’s) that goes beyond these limits
to a universal morality is incoherent. Since I have claimed that God commands
a universal morality (though morality is also particular), I have to reject this
proposal. The second proposal is anti-realist about value and about God. It
suggests that we can learn from our status as evolved that there are no
objectively prescriptive moral demands upon us, and that the sense of any
such demand is an illusion, though perhaps useful for our survival and
reproduction. Moreover, on this view, evolution suggests that belief in God
is illusory in the same way. The third proposal is that there is indeed a gap
between what morality now demands and how we originally evolved, but that
our original evolutionary situation itself gives us by extrapolation a method for
how to bridge this gap, and we do not need (and evolutionary psychology
shows us that we should not look for) an appeal to God’s command. The
fourth proposal is that bringing in our relation to God gives us a good
explanation for how we are morally obligated and how we might have
developed to appreciate such obligation. God commands us in a way that
fits the capacities that we have evolved with, though we cannot deduce these
commands from these capacities. On this fourth proposal, the main reason for
262 God’s Command
this combination of fit with non-deducibility is that evolution has given us a
mixture of tendencies, some of which produce what is good when we follow
them and some of which produce what is bad. God’s commands allow us to
determine which we are to follow.
After telling a story in the first section, the chapter devotes the next four
sections to looking at these four options for how we might construe the
relation between our morality and our evolutionary history. The second
section, looking at a reduction of the moral demand, examines three figures,
Herbert Spencer, Larry Arnhart, and Jonathan Haidt. The third section dis-
cusses arguments for an anti-realist position in ethics and theology, especially
the arguments of John Mackie, Michael Ruse, Sharon Street, and Paul Bloom.
The fourth section looks at suggestions for how to go beyond our evolutionary
history without God, and examines the views of Joshua Greene and Philip
Kitcher. Finally, the fifth section returns to the story in Section 1 and suggests
that we see God’s command as a gift enabling us to transcend the limits of our
evolutionary history.
A great deal of the literature discussed in this chapter is in the genre of the
popularization of science. Works in this genre have a short shelf life. We
should focus on the general principles to which these authors appeal, which
have a much longer life. It is typical for such works to treat a number of
different disciplines. This chapter is entitled ‘Divine Command and Evolu-
tionary Psychology’, but it refers also to anthropology and primatology and
game theory and sociology. The scope is presumptuous. But I am encouraged
in undertaking so ambitious a project by observing that the professionals in
those disciplines think it quite appropriate to write about philosophy.
The first section tells a story about the origins of our morality. The story is
just a story, not history or science. The story is not, however, merely fiction.
The aim is to embed elements of the essential structure of the story at the
beginning of Genesis about the Garden of Eden in an account whose details
are mostly drawn from contemporary (non-theological) anthropology. It is
still a story or myth, telescoping what a scientific account would spread over
hundreds of thousands of years. It also abstracts from the role of inter-group
relations in the development of morality. If we ask what is the point of such a
story, the answer is that we think many of our most important thoughts in
stories, and probably always have. The story does not mention God. But the
fifth section of this chapter suggests that a storyteller who did mention God
would provide a satisfying addition from an explanatory point of view. We can
see the story as one that an anthropologist might tell her children, or as a
Kant-like translation of the biblical story ‘within the boundaries of mere
reason’.1

1
Kant himself tells a similar story in Speculative Beginning of Human History (viii. 109–13),
based on the structure of Genesis 3, though he treats the command not to eat the fruit as a
 Divine Command and Evolutionary Psychology 263

8.1. THE S TORY

Once upon a time there lived in Central Africa a group of apes. They were
different from the groups of apes who lived around them, and they recognized
this difference. For one thing, they seemed to be able to think of themselves as
a group, and to think of what helped them as a group and what harmed them
as a group. They would regularly meet together, and they sometimes had a
kind of experience together when they met that also separated them from the
other apes. They had an experience of everything belonging together, not just
their own group, but everything; the water in the river and the trees and the
other animals and birds and the stars in the night sky. And it all seemed to
them good and beautiful. These assemblies gave them great joy and also a
sense of awe, and they came to organize their lives together around them. They
were able at these times to forget what kept them apart from each other, and to
rejoice in what kept them together. Because of their new kind of unity, they
were able to invent new cooperative ways to find food, and find new places to
live that could sustain their form of life.
There arose among them a symbol for this goodness and beauty they had
discovered, and a symbol of how the enjoyment of it distinguished them from
the other apes in the old lands. They found themselves refraining from a
particular kind of fruit, and this restraint was connected with their distinctive
new form of life. Eating this fruit had been typical of the old way, the way of
their ancestors, and they now needed to separate off their new way, connected
with their new capacities and their new assemblies. One way they could tell
their own members, even though they looked in most ways like the other apes,
was that they would not eat of this fruit. They came to think of the fruit as
forbidden by their common life, even though there was no reason (other than
the symbolic connection) for refraining.
Some times were good and plentiful for food, and some times were bad and
food was scarce. One day, when food was scarce, the elders of the group saw
other animals eating the forbidden fruit, and they felt weariness with the
restriction and a desire to go back to the old ways. They wanted to decide
for themselves what they should eat and what they should not. They decided
to eat the fruit themselves. This was a decision different in principle from
eating the fruit in the old life, even though it was a decision to eat the same
food, because it was now a decision against the authority of the common
standard for their lives that they had accepted.
When they had made this decision, they found consequences that were
natural but unexpected. One was that they lost the joy in their assemblies

command of nature or instinct (since our ancestors did not naturally eat that kind of fruit),
which reason (in conjunction with imagination) then violates. He describes the story as ‘an
exercise of the imagination in the company of reason’ (viii. 109).
264 God’s Command
together. They also found that their sexual lives changed. Before the decision,
they had been so conscious of what held them together as a group that they
had not needed to protect themselves from each other, though they protected
themselves and each other against common enemies. Now, they found them-
selves hiding from each other or fighting each other. The power of their
common life waned, and competition increased for what each controlled
individually. That included their food, but also their own bodies. They started
to hide their bodies from each other by covering them, and to feel a new
emotion of shame when they were uncovered.
Finally, the fighting and the competition between them got so bad that they
were not able any longer to trust each other in the ways required for the
cooperation in finding food that they had discovered in their new place.
Without this cooperation their lives in the new place became unsustainable,
and they were forced to leave. However, they kept with them the memory of
how it had been, and the aspiration to return to it. They became in this way
divided, each internally in their hearts, between the desire to protect what
belonged to the individual and the desire for the common good that had been
shared between them.

8.2. EVOLUTION AND REDUCING THE

MO RAL DEMAND

The first way of thinking about the relation between evolution and morality is
that evolution shows the idea of impartial benevolence to be utopian.
Section 8.2.1 covers the views of Herbert Spencer and Larry Arnhart, and
Section 8.2.2 the work of Jonathan Haidt.

8.2.1. Herbert Spencer and Larry Arnhart

In this subsection we look at two attempts to oppose a Kantian or universal

morality on the basis that it is unrealistic for our present condition, given our
evolutionary endowment. I start with Herbert Spencer because he states a
classic version of the view, even though no one now wants to be associated
with him. Then I look at Larry Arnhart, though I do not want to suggest by
putting the two side by side that they are the same other than in their common
opposition to Kant.
Spencer is now deeply unpopular because of the use that was made of his
eugenic ideas in the twentieth century. For Spencer, as Michael Ruse puts it,
‘what holds as a matter of fact among organisms holds as a matter of
 Divine Command and Evolutionary Psychology 265
obligation among humans’.2 The relevant fact about organisms is the struggle
for existence, and the consequent weeding-out of the less fit. Spencer says:
Blind to the fact that under the natural order of things, society is constantly
excreting its unhealthy, imbecile, slow, vacillating, faithless members, these un-
thinking, though well-meaning, men advocate an interference which not only
stops the purifying process but even increases the vitiation—absolutely encour-
ages the multiplication of the reckless and incompetent by offering them an
unfailing provision, and discourages the multiplication of the competent and
provident by heightening the prospective difficulty of maintaining a family.3
The ‘unthinking though well-meaning men’ he is talking about are those who
were advocating in the name of a universal humanitarianism some interven-
tion by the state to counteract the effects of the unregulated market in
nineteenth-century Britain. In Germany, this idea of the law of struggle was
taken up, notoriously, by Hitler in Mein Kampf: ‘He who wants to live must
fight, and he who does not want to fight in this world where eternal struggle is
the law of life has no right to exist.’4 National Socialism took up also the idea of
encouraging the natural order by which imbecile and unfit parts of the
population are eliminated, and the highest form of life flourishes. Spencer
did not, however, think that this natural order of struggle was permanent. He
was a Lamarckian, not a Darwinian, and he thought that there would be
human progress through the inheritance of acquired characteristics, so that
the lower forms of human life most given to violence would decline, and we
would end with universal peace. Still, in our current condition, he thought that
we should let the order of nature weed out the unfit also in human society,
since we are part of nature.
The particular application to eugenics and laissez-faire economics is not the
important thing for our present purposes, but the general principle that we
should follow our biological nature. Chapter 4 argued against what it called
‘deductivism’, the principle that we can deduce our moral obligations from
human nature. The present principle is a species of deductivism, telling us that
we can tell how we ought to live by looking at the nature of organisms in
general, since we are organisms. The trouble with this principle is that the
nature of organisms in general, and human nature in particular, contains
characteristics that, when promoted in human society, produce evil as well
as good by Kantian and utilitarian standards. To say this is not so much to

2
Michael Ruse, Can a Darwinian Be a Christian? (Cambridge: Cambridge University Press,
2001), 170.
3
Herbert Spencer, Social Statics; or the Conditions Essential to Human Happiness Specified
and the First of Them Developed (London: J. Chapman, 1851), 323–4.
4
Adolf Hitler, Mein Kampf (My Struggle), trans. J. Murphy (London: Hurst and Blakett,
1939), 242. Ruse associates Spencer with Hitler, and qualifies the association, in Can a Darwinian
Be a Christian?, 173.
266 God’s Command
argue against Spencer as to display some of the consequences of his view, and
the same is true with Arnhart. Both of these authors are perfectly well aware
that their position is inconsistent with utilitarian and Kantian systems.
This deductivism is clearly displayed in Larry Arnhart’s Darwinian Natural
Right: The Biological Ethics of Human Nature, a work we earlier compared
with Philippa Foot’s Natural Goodness in Section 4.4.3.5 The governing
principle of Arnhart’s book is that the definition of the good as the desirable
(as in Aquinas) means that the good is what is generally desired, or what most
people in every society throughout our time on earth have in fact desired.
Arnhart claims that evolution has given us these desires because of their
adaptive value, and he lists twenty of them. The claim is not that these desires
are universal, because there can be defective individuals who lack them. But
the principle of his book is that only if a desire is general in the above sense, or
is a specification or application of such a desire, is its fulfilment good. The
normative theory that results is one, he claims, that enables us to ‘understand
human nature within the natural order of the whole’.6 He intends a contrast
here with Christianity, which invokes ‘the supernatural’ in explaining how we
should live. And he faults Darwin for having been misled by the prevailing
‘universal humanitarianism’ of his time into a ‘utopian yearning for an ideal
moral realm that transcends nature’, a yearning ‘which contradicts Darwin’s
general claim that human beings are fully contained within the natural order’.7
Arnhart does not deny that humans have a natural sympathy for others, but,
‘although sympathy can be expanded to embrace ever-larger groups based on
some sense of shared interests, this will always rest on loving one’s own group
as opposed to other groups’. Arnhartian morality will always be, in the
language of Chapter 3, self-indexed.
The important point for our present purposes is that the list of twenty natural
desires does not include disinterested benevolence or the love of the enemy, and
therefore the theory cannot say that the fulfilment of such desires or preferences
is good. It is significant that Aristotle is Arnhart’s philosophical hero, to whom
he continually appeals (though he is blind to what I call the ‘vertical dimension’
of Aristotle’s thought). Section 4.2 argued that Aristotle thinks, at least most of
the time, that an admirable human life requires wealth and power and high
status, and in general the competitive goods. He may by and large be right about
the desires we are born with, which come to give a dominant place to
these goods, but it does not follow that he is right in his inference that the
fulfilment of this ranking is good. The thesis of my book has been that

5
Arnhart himself is no longer committed to all the doctrine in Darwinian Natural Right. See
the discussion of our ‘moral sense’ in Darwinian Conservatism (Exeter: Imprint Academic Press,
2005). But he still distinguishes between the ‘utopian vision’ and the ‘realist vision’, and holds to
the list of twenty natural desires.
6 7
Arnhart, Darwinian Natural Right, 275. Ibid. 146.
 Divine Command and Evolutionary Psychology 267
‘following nature’ in this way is not a good alternative to following Kantian or
Christian morality.

8.2.2. Jonathan Haidt

There is much to commend in Jonathan Haidt’s The Righteous Mind: Why

Good People Are Divided by Politics and Religion.8 But for the present section
the key point is that Haidt defends the view we saw in Arnhart that evolution
has given us a ‘groupish’ attachment, one that is designed ‘to make [groups]
more effective at competing with other groups’.9 So far there is no objection.
But Haidt goes on immediately to ask: ‘But is that really such a bad thing
overall, given how shallow our care for strangers is in the first place? Might the
world be a better place if we could greatly increase the care people get within
their existing groups and nations while slightly decreasing the care they get
from strangers in other groups and nations?’ He suggests comparing two
nations—one nation full of small-scale groups with internal cohesion but in
competition with other such groups, and the other nation with no such groups
at all. Which nation, he asks, will score higher on measures of social capital,
mental health, and happiness? Clearly the first. His conclusion is that ‘it would
be nice to believe that we humans were designed to love everyone uncondi-
tionally. Nice, but rather unlikely from an evolutionary perspective. Parochial
love—love within groups—amplified by similarity, a sense of shared fate, and
the suppression of free riders, may be the most we can accomplish.’10 Religion
is, he thinks, the crucial social practice that enables group formation. ‘But
should we really expect religion to turn people into unconditional altruists,
ready to help strangers under any circumstances? Whatever Christ said about
the Good Samaritan who helped an injured Jew, if religion is a group-level
adaptation, then it should produce parochial altruism.’11
How does Haidt get to this conclusion? He is aware of the difference
between a descriptive and a normative account, and he says that he has been
‘entirely descriptive’ until the very end of the penultimate chapter.12 But in fact
he slips back and forth between the normative and the descriptive all the time,
and the passages quoted in the previous paragraph are one example. What
enables the shift in this case is the premiss that evolution has set limits to how
we can live together, and (though he does not say so) ‘ought implies can’. Our
genes under the prompting of religion give us parochial altruism, but not
disinterested benevolence, or the kind of care that the Good Samaritan gave to
the injured Jew. What is strikingly absent in this account is any exploration of
the universalizing tendency of some religion. Religion is treated throughout as

8
Jonathan Haidt, The Righteous Mind (New York: Vintage Books, 2012) (henceforth RM).
9 10 11 12
RM 281. RM 284. RM 308. RM 315.
268 God’s Command
a ‘hive switch’, a group-level adaptation that gives us cohesion within the
group together with competition against those outside it. But one theme of my
book has been that we can find within the Abrahamic faiths not only tribal
loyalty but divine commands that tell us to love or show mercy to the enemy
and stranger and give us (because of the nature of the commander) resources
for doing so.13 In Chapter 1 I gave three arguments, directly or indirectly from
Kant, who refers back continually to the Sermon on the Mount (his favourite
passage in Scripture) and takes it that the moral law telling us to share the
morally permitted ends of all those affected by our actions is itself a divine
command. This command, by the argument from providence, can be followed
consistently with our happiness, even though we do not see how we can secure
such consistency ourselves. And this command, by the argument from grace,
is what God assists us to follow, even where we are not inclined by our own
internal resources to do so. And this command, by the argument from
justification, is what makes something morally obligatory for us, even if we
do not perceive that it is God’s command. The present point is that these three
arguments reveal an internal structure to this form of religion. If we are going
to talk about the contribution of religion to morality, we need to take these
features into account.
To understand why Haidt thinks of religion and morality the way he does,
we need to go further back within his system. In 2001 he published an
influential article entitled ‘The Emotional Dog and its Rational Tail: A Social
Intuitionist Approach to Moral Judgment’.14 The burden of the article was
that psychology has been mistaken in following the lead of Lawrence Kohlberg
(and behind him, Kant, and behind Kant, Plato) in valorizing reason as the
source of moral judgement. Rather, to use a different metaphor that is central
in The Righteous Mind, we should think of emotion as the elephant and reason
as a rider who is controlled by the elephant. The contrast is with Plato in the
Phaedrus (246a), who thinks of reason as the charioteer, controlling the two
horses of ambition and passion. On Haidt’s picture there is nothing control-
ling emotion except other emotions.
Some clarification of terms here will help. In The Righteous Mind Haidt
acknowledges that the use of the term ‘emotion’ draws the wrong contrast,
because it suggests that what is in control is something non-cognitive, and that
cognition is what is controlled. He now concedes that emotions are in fact
‘filled with cognition’, and he moves to saying that the contrast is between two
forms of cognition, which he now calls ‘intuition’ and ‘reasoning’.15 But this is

13
I do not deny that we can find tribal loyalty, e.g. in 1 Samuel 18: 24, and Deuteronomy 13: 6–10.
14
Jonathan Haidt, ‘The Emotional Dog and its Rational Tail’, Psychological Review, 108
(2001), 814–34.
15
RM 52–3. He was influenced in his initial terminology by Edward O. Wilson, in Sociobiol-
ogy: The New Synthesis (Cambridge, MA: Harvard University Press, 1975), 563: ‘Ethical philo-
sophers intuit the deontological canons of morality by consulting the emotive centers of their
 Divine Command and Evolutionary Psychology 269
still somewhat confusing terminology because intuition, in philosophical
history, is a kind of reason. Thus Aristotle distinguishes between nous (intui-
tive reason) and dianoia (discursive reason). Discursive reason takes time,
since it is a process of reaching a conclusion from premisses, whereas intuitive
reason perceives at a moment. But this temporal difference has no implication
that one faculty is rational and the other is not. Section 4.3.3 discussed RMH’s
distinction between what he called ‘critical thinking’ and ‘intuitive thinking’,
two different levels of moral thought. Most of the time, on his view, we operate
at the level of intuitive thinking, applying principles that we ideally endorse at
the critical level.
Haidt has misunderstood Plato here. He thinks that Plato is telling us in the
Republic that ‘passions are and ought only to be the servants of reason, to
reverse Hume’s formulation’, so that the philosophers are kings.16 But Plato
does not say that philosophers are kings, or that passions are the servants of
reason; he says that they should be. Much of the Republic is a description of
states or cities in which there is no rule by reason. The fact that we are actually
ruled often by something non-rational does not show that Hume is right and
Plato is wrong.
Haidt is wrong also about Kant. He is not a philosopher, and it may seem
small-minded to object to mischaracterizations of philosophical sources by a
psychologist. But these references to tradition are supposed to carry weight.
Hume’s victory over Kant is repeatedly trumpeted. So what is the picture of
Kant here? He was ‘rather low on empathizing’, though not as low as Bentham,
who probably had Asperger’s syndrome.17 And what is the evidence for this?
Haidt suggests that Kant provided an abstract rule, the Categorical Imperative,
which is based in logic, and in particular in the law of non-contradiction. But
Haidt does not seem to know the formula of the end-in-itself, as explained in
Section 1.1. According to this formulation of the Categorical Imperative, we
have to share as far as possible the ends of all those we affect by our actions,
and we have to make those ends our own ends. This requires us, Kant says, to
sympathize.18 Haidt is trading in caricature.
To return to the main point about religion and morality, Haidt’s view is that
we should not think of God as giving us a command to universal morality,
because there is no rational moral compass that could receive such a com-
mand, and no ‘inner scientist’ trying to find the truth about how to live. The
key here is not the meaning of ‘reason’ or ‘rational’, but whether we have the

own hypothalamic-limbic system. This is also true of the developmentalists [such as Kohlberg],
even when they are being their most severely objective. Only by interpreting the activity of the
emotive centers as a biological adaptation can the meaning of the canons be deciphered.’
16 17
RM 33 (emphasis added). RM 140.
18
MdS vi. 457, where ‘sympathizing’ is said to be a duty, and we should not avoid places like
sickrooms or debtors’ prisons, ‘for [sympathy] is still one of the impulses that nature has
implanted in us to do what the representation of duty alone might not accomplish’.
270 God’s Command
ability to submit our initial inputs (call them ‘emotions’ or ‘intuitions’) to
scrutiny, in order to determine which to accept and which to reject. In terms of
his picture, does the elephant have a rider who directs it? To the layman, it
seems clear that we do have this ability at least some of the time. If we do, then
the question is whether we are to be held to universal morality as the screen
through which this scrutiny should pass.
Haidt has three kinds of evidence for the hypothesis that ‘the intuitive dog
wags the rational tail’. The evidence is from what he calls ‘dumbfounding’ and
‘post hoc fabrication’, from psychopathy, and from the bias towards the self
that is pervasive in moral justification. To obtain the first kind of evidence
Haidt tells his subjects stories that involve what he calls ‘harmless taboo
violations’ and that he contrasts with ‘harm-based’ stories like the one Kohl-
berg used to tell his subjects about Hans stealing a drug to save his wife. Here is
one example of a ‘harmless taboo violation’: ‘Jennifer works in a hospital
pathology lab. She’s a vegetarian for moral reasons—she thinks it’s wrong to
kill animals. But one night she has to incinerate a fresh human cadaver, and
she thinks it’s a waste to throw away perfectly edible flesh. So she cuts off a
piece of flesh and takes it home. Then she cooks it and eats it.’19 The subjects
presented with this vignette experienced a predictable flash of disgust. Only 13
per cent said that what Jennifer did was all right. But when asked to say what
was wrong with what she did, the subjects seemed at a loss. ‘They seemed to be
flailing around’, Haidt says, ‘throwing out reason after reason, and rarely
changing their minds when [the experimenter] proved that their latest reason
was not relevant. . . . People were making a moral judgment immediately and
emotionally. Reasoning was merely the servant of the passions, and when the
servant failed to find any good arguments, the master did not change his
mind.’20
At least to the layman, dumbfounding does not seem to show what Haidt
wants. If we return to RMH’s picture of the two levels of moral thinking, the
critical and the intuitive, we will find a spectrum of cases where the connection
between the levels is more and less articulated. We might call this ‘the
articulateness spectrum’. Reaction to harm-based stories like the ones Kohl-
berg told his subjects apparently tends to be higher on this spectrum. We can
compare Aristotle on this question of articulateness. He has the picture
discussed in Section 3.2.1, of our lives having a vision of the good, but he
does not say that this vision has to be explicit or articulated.21 For most of us it
is not. Even without this articulation, it can shape the lives we try to lead.
Suppose that what used to be pervasive in society was a justification of the
prohibition of cannibalism or incest (Haidt’s other main example) in terms of
divine command, that these were against the order that God had established.

19 20 21
RM 45. RM 47. NE I, 4, 1095a32–b13.
 Divine Command and Evolutionary Psychology 271
But suppose this kind of justification has become less socially prevalent. We
would expect people to become less articulate in their discursive reasoning.
Dumbfounding may well be culturally relative, so that cultures that stress what
Haidt calls ‘the ethic of divinity’ are not dumbfounded by just the same
stories.22 But from this cultural relativity it would not follow that the intuitions
of people in those cultures were not tracking something actually bad, or that
they did not have a conscience or rational moral compass whose job it was to
do this tracking.
The data here are important, because they show that we are less good at
explicit discursive reasoning than we tend to think we are. But the data do not
establish the conclusion that Haidt wants, that ‘the rider’s job is to serve the
elephant, not to act as a moral compass’.23 Again, we have here the slip
between the descriptive and the normative. This last quote comes in the
context of Haidt’s discussion of psychopathy, which is his second kind of
evidence. The whole passage is revealing.
The serial killer Ted Bundy, for example, was a psychology major in college,
where he volunteered on a crisis hotline. On those phone calls he learned how to
speak to women and gain their trust. Then he raped, mutilated, and murdered at
least thirty young women before being captured in 1978. Psychopathy does not
appear to be caused by poor mothering or early trauma, or to have any other
nurture-based explanation. It’s a genetically heritable condition that creates
brains that are unmoved by the needs, suffering, or dignity of others. The
elephant doesn’t respond with the slightest lean to the gravest injustice. The
rider is perfectly normal—he does strategic reasoning quite well. But the rider’s
job is to serve the elephant, not to act as a moral compass.
Haidt is saying here that we do not have a faculty, which Kant would call a
rational will or conscience, whose job it is to act as moral compass. There is
only strategic or executive reason, Ted Bundy’s reason, the ‘perfectly normal’
rider serving Bundy’s elephant. But how could this conclusion be established
from the data of psychopathy? Even if there is a genetic base for psychopathy,
nothing follows about whether people without this condition have a faculty of
reason that can guide them in more than strategic planning. Haidt has reduced
reason to what Aristotle calls ‘cleverness’, which works out the means to any
end presented. Aristotle says: ‘Both practically wise and villainous people are
called clever.’24 But the evidence of our failures of practical wisdom does not
show that we do not have the faculties that would make such wisdom possible,
only that we do not exercise them reliably.
The third kind of evidence Haidt uses is from the bias towards the self that is
pervasive in moral justification. He tries to show that ‘reason is not fit to rule; it

22 23 24
RM 123–4. RM 73. NE VI, 12, 1144a23–28.
272 God’s Command
was designed to seek justification, not truth’.25 What his data show, however,
is something else, something he says in the very next sentence: ‘People care a
great deal more about appearance and reputation than about reality.’ There is a
key difference between these claims. The second is perfectly consistent with,
and indeed supports, the Kantian view that we start off under the propensity
to evil that overrides the equally innate but (unlike the propensity to evil)
essential human predisposition to good. But the first denies this view, because
it denies Kant’s account of the predisposition, which is that we are the sorts of
creature who respond with a certain kind of feeling (Achtung, usually trans-
lated ‘respect’) to recognition in the reason that something is required or
forbidden by the moral law. Haidt details evidence that ‘people invest their IQ
in buttressing their own case rather than in exploring the entire issue more
fully and evenhandedly’.26 Our reason is seen here as like an inner lawyer or
inner press secretary, producing arguments for outside consumption, and not
an inner scientist trying to discover the truth about what we ought to do. He
shows also evidence from the brain-scanning of partisans of one political party
who were presented with reports of their favourite candidates apparently
behaving hypocritically.27 He asks:
Would subjects reveal . . . that the head (the reasoning parts of the brain) processes
information about contradictions equally for all targets, but then gets overruled by
a stronger response from the heart (the emotion areas)? Or does the partisan brain
work as Hume says, with emotional and intuitive processes running the show and
only putting in a call to reasoning when its services are needed to justify a desired
conclusion? The data came out strongly supporting Hume.
But the fact that we pay attention to and delight disproportionately in thinking
about what suits our own inclinations does not show that when we do so we
are thinking properly, or that our reason is doing its ‘job’. Rather, it shows that
we are not doing our job as rational animals at all well.
To conclude this subsection, a divine command theorist should take cog-
nizance of the evidence of all three types (dumbfounding and psychopathy
and bias), and should be chastened by it because of what it shows about our
lack of intellectual virtue and some people’s lack of conscience altogether. But
this should not make her abandon her theory. What she holds possible and
what she holds obligatory depend upon her theological premisses, and what
she thinks in particular about the three arguments presented in the first
chapter. Evidence about our various forms of cognitive failure does not
show that we do not have the ability to screen our initial inputs given the
available assistance, or that universal morality is not an appropriate screen. If
this is right, then this evidence does not show us that ‘parochial altruism is the
most we can accomplish’.

25 26 27
RM 86 (emphasis added in the following sentence). RM 95. RM 101–2.
 Divine Command and Evolutionary Psychology 273

8 . 3 . E V O L U T I O N AN D A N T I - R E A L I S M

In this section the topic is whether evolutionary psychology gives us a reason

to be anti-realists, either about value or about God. The first of these forms of
anti-realism rejects the view described in Section 4.3.2 as ‘prescriptive realism’.
According to prescriptive realism, when we make moral judgements we are
both expressing some attitude of the will or desire and claiming that evaluative
reality is a certain way independently of our judgement, so that our judgement
is appropriate to it. The second part of this, the realism, is what is at stake in
the present context. Section 8.3.1 is about John Mackie, Section 8.3.2 about
Michael Ruse, and Section 8.3.3 about Sharon Street. The second form of anti-
realism is about God, and since most of the figures discussed in this chapter
are atheists, they are anti-realist in this sense. But Section 8.3.4 is the only
subsection focused this topic, and is about Paul Bloom.

8.3.1. John Mackie

We begin with John Mackie’s argument in Ethics: Inventing Right and

Wrong.28 Mackie’s first sentence is: ‘There are no objective values.’ He was a
Humean like Haidt, and he thought that our tendency to believe in objective
value results from what Hume called the mind’s ‘propensity to spread itself on
external objects’ together with the pressure of our sociality.29 Mackie proposed
an error theory, ‘that although most people in making moral judgments
implicitly claim, among other things, to be pointing to something objectively
prescriptive, these claims are all false’.30 In other words, Mackie conceded that
realists are right about what moral language means, but he held that nonethe-
less what people mean when they make moral judgements is always false. He
conceded also that, if divine command theory were true, then moral judge-
ments that claim objective prescriptivity would also be true. But Mackie was
an atheist, and accordingly he thought divine command theory false. Mackie
was also opposed, like Haidt, to Kant’s universalism, and behind this to the
biblical commandment ‘You shall love your neighbour as yourself ’.31 This is
simply impracticable, he said, and it is inconsistent with human nature. This is
because ‘a large element of selfishness—or, in an older terminology, self-love—
is a quite ineradicable part of human nature’, and it is doubtful that any agency

28
J. L. Mackie, Ethics: Inventing Right and Wrong (Harmondsworth: Penguin, 1977) (hence-
forth E.).
29 30
E. 42–3. E. 35.
31
E. 130. I recall a seminar at Oxford in the early 1970s, taught by John Mackie, RMH, and
John McDowell. One of the persisting topics was Mackie’s resistance to what he called the ‘third
stage’ of universalization, which ‘approximates to the giving of equal weight to all real interests’
and which ‘is plainly not characteristic of moral thought in general’ (E. 97).
274 God’s Command
could effect the fundamental changes that would be needed to make practic-
able a morality of universal concern.32 If we ask why human nature is this way,
this is ‘a sociological and biological question to be answered by an evolutionary
explanation’.33
Mackie had two arguments against realism, which he called ‘the argument
from relativity’ and ‘the argument from queerness’. The first argument is that
humans’ moral views are too diverse for us to suppose plausibly that we are all
receptors of the same objectively prescriptive values beaming down to us. Our
moral views seem to reflect, rather, participation in different ways of life. But
for a divine command theory of the type I am defending, it is not at all
surprising to find substantial variation in the reception of divine commands.
First, in Kant’s language, we are born under the evil maxim, so that we have, in
addition to the predisposition to good, the propensity to evil. The closer a
faculty is to our heart or will, the more likely the faculty is to be distorted in its
perceptions by the preference for our own happiness over what is good in
itself, independently of its relation to ourselves. There are manifold ways in
which it is possible to get value perceptions wrong, and so there is manifold
variety in moral views. The contrast with colour perception is interesting here.
Though there are marginal differences in how different people split up the
spectrum, there is large-scale agreement. Second, what God commands one set
of people, or one person within a group, may be different from what God
commands another. But a third point is that Mackie may have been wrong
about the amount of variety. The pendulum seems to have swung back within
evolutionary psychology to the acknowledgement of ‘human universals’.34 We
will return to this in discussing Michael Ruse in the next subsection. It is
surprising in fact how much agreement there seems to be on basic principles
between cultures, though the details and application of these principles vary
substantially.
The argument from queerness is that the objectively prescriptive values that
realism proposes and their effects on us are very strange things, not easily
related to any kind of causation we know about within science. ‘How much
simpler and more comprehensible the situation would be’, Mackie said, ‘if we
could replace the moral quality with some sort of subjective response which
could be causally related to the detection of the natural features on which the
supposed quality is said to be consequential’.35 He held that ultimately we
must turn to science to explain the existence of things in the world and their
causal relations. Evolutionary science might be able to explain why we have

32 33
E. 132–3. E. 192 (emphasis added).
34
There is a good discussion of Mackie on this point in Ritchie, From Morality to Metaphys-
ics, 19–20. See also Leonard D. Katz (ed.), Evolutionary Origins of Morality: Cross-Disciplinary
Perspectives (Exeter: Imprint Academic, 2000).
35
E. 41. He thought that science could at least in principle explain why we are caused to feel a
certain way by certain natural features of the world.
 Divine Command and Evolutionary Psychology 275
moral sentiments for adaptive advantage, though the details still need to be
worked out. But the notion of something objective in the world like rightness
and wrongness is, in Mackie’s term, ‘queer’, by which he meant inexplicable by
scientific theory. He accepted that it might make sense if we believed in a God
who was prescribing, but science acknowledges, in his view, no such thing. He
was right to point out that a theist has less reason than an atheist to be an anti-
realist about value. A divine command theorist already believes in a divine
spiritual person outside normal science. She will still have valid questions
about how a spiritual being communicates with material beings like us, but she
will be less inclined to think such communication is impossible. A divine
command theorist still needs to worry, however, about Mackie’s various
arguments that there is no God.36 We return to this in Section 8.3.4. The
main point is that it is not the business of science to pass judgement on
whether there is or is not a God.

8.3.2. Michael Ruse

Michael Ruse is also an anti-realist, in many respects a follower of Mackie, but

he has departed from Mackie in some significant ways that make his anti-
realism unstable. ‘Ethics is an illusion,’ he says, ‘put in place by natural
selection to make us good cooperators’.37 Ruse is a moral sceptic.38 He does
not think the sense of right and wrong has a justification at all. He thinks it is
an illusion foisted on us by our genes, in the same way we might have a visual
illusion, a mirage of an oasis in the desert; we really have the sensation, but the
oasis is not really there.
Ruse is strikingly optimistic that our moral lives will not be affected by the
kind of scepticism he endorses. But surely we need some kind of justification
for morality, some answer to what was called in Chapter 1 ‘the normative
question’, because not everybody is consistently moved by the forces of natural
selection to cooperate in the way morality requires. Moral obedience is fragile.
We do find precursors of the moral sentiments in our non-human ancestors,
but we also find defection, and we have inherited both of these tendencies. We
are by nature, in this sense, a mixture. But this means we need support from
our cultural sources not only for our beliefs about what morality requires, but
for our beliefs about why we should comply with it, or endorse it; why I should
take the moral demand as a valid demand upon me. Even if we are committed

36
J. L. Mackie gives the arguments in The Miracle of Theism (Oxford: Clarendon Press, 1982).
37
Michael Ruse, ‘Is Darwinian Metaethics Possible (And If It Is, Is It Well Taken)?’, in
Giovanni Boniolo and Gabriele De Anna (eds), Evolutionary Ethics and Contemporary Biology
(Cambridge: Cambridge University Press, 2006) (henceforth DM), 13. Ruse is quoting from
Michael Ruse and E. O. Wilson, ‘The Evolution of Morality’, New Scientist, 1478 (1985), 108–28.
38
Ruse appeals explicitly to Mackie as his source at DM 21.
276 God’s Command
to the moral law, we need something to say to ourselves when that commit-
ment becomes difficult to honour. There is evidence in the psychological
literature that the force of the moral demand can be undermined by teaching,
as Ruse does, that objective morality is an illusion.39 For example, a before-
and-after study was done on students enrolled in two introductory economics
courses and an introductory astronomy course. The students were asked at the
beginning and at the end of each course what they would do if they found
an addressed envelope with $100 in it. While the students scored the same in
the economics and astronomy courses at the beginning of the semester, the
economics students were significantly more willing to keep the money at the
end. The difference probably lay in the content of the courses, in which they
were taught in line with old-style neoclassical theory that people act in their
own enlightened self-interest. Now Ruse is not arguing that we are motivated
only by our own interests; he wants to say that morality is genuine, in the sense
of sincere, even though it is not genuine in the sense of non-illusory. He thinks
we really do have the moral sentiments. But the point is that saying that ethics
is an illusion put in place by natural selection to make us good cooperators is
likely to have the same undercutting effect as the economic theory just
described.
Here is a thought experiment to make this vivid. The ancient Epicureans
used to believe that the gods had no care for us, and they set us up and watched
us for their own entertainment, rather in the same way as we now watch soap
operas on television. Suppose they gave us moral sentiments just because that
makes our endless agonizing more interesting to watch. We would then really
have those moral sentiments, but they would not be tracking any moral reality.
Suppose I discover this about the gods. What effect will that have on my
motivation? I will still have the moral sentiments, but I will come to see that
they are without justification, that they are an illusion. Surely I will then try to
discount them. I will feel the urge to help my neighbour in some sacrificial
way, and I will say to myself: ‘That’s just the gods again, curse them.’ Ruse
concedes that sometimes morality will tell us to do what is not to our adaptive
advantage as individuals. If I believed him about the illusion, and I felt the
moral prompting to sacrifice, I would probably say: ‘That’s just the genes
again, curse them.’ There are people whose moral commitment has in fact
been undermined by views like Ruse’s about morality being an illusion. While
at Lehigh University, I had a distinguished colleague, Donald Campbell,
president of the American Social Psychologists’ Association, who used to
have views rather like those of Ruse. But he also accepted the point that
these views tended to have an undermining effect, such as I have described.
He told us that it was for this reason he refused to write for the general public,

39
Robert H. Frank, Thomas Gilovich, and Dennis T. Regan, ‘Does Studying Economics
Inhibit Cooperation?’, Journal of Economic Perspectives, 7/2 (1993), 159–71.
 Divine Command and Evolutionary Psychology 277
and kept these views to the domain of scholarship, because he did not want
what he called ‘the preachments’ of the tradition about justification to lose any
force they currently had.40
But perhaps it is just an unfortunate truth that morality is an illusion and
that recognizing this tends to undermine its force. Perhaps it is like the sad
truth that it is really our parents and not Santa who put the presents under the
Christmas tree. What arguments does Ruse have for his scepticism? He has
basically two, and they are versions of the same arguments we saw in Mackie.
But here is the irony. Ruse ought not to accept either of them any longer
because of differences from his mentor that he has come to have in other parts
of his theory.
First, the argument from relativity. Ruse’s form of the argument makes a
significant shift from the factual to the counterfactual. In Mackie the argument
is from the variety we can actually observe between human societies. Ruse is
much less persuaded of this variety, and in fact, like much of contemporary
evolutionary psychology, he embodies a pendulum swing back to human
universals, encoded in our genes (with environmental triggers). He appeals
to what he calls ‘our shared psychological nature’, which includes a sense of
right and wrong, in order to explain how we can rely on each other morally
without any foundation or justification for why we should try to be morally
good.41 So his argument from relativity is counterfactual; we could have had a
quite different morality if our evolutionary history had been different. ‘As a
Darwinian, it is plausible to suggest that humans might have evolved with
the . . . kind of morality, where the highest ethical calling would not be love
your neighbor but hate your neighbor. But remember that your neighbors hate
you, and so you had better not harm them because they are going to come
straight back at you and do the same.’42 The argument is that, since evolution
could have taken a different path, there cannot be an objective set of values
that lies behind our moral practice. But for a divine command theorist this is
not a successful objection. God could use evolution to produce the kind of
creatures God wants to have, and this does not deny ‘random’ mutation of the
kind that Darwinian evolution proposes.43 Ruse concedes this, and agrees that
a Christian can, consistently with science, ‘be committed to a form of what is

40
See Donald Campbell, ‘Altruism: Biology, Culture and Religion’, Journal of Social and
Clinical Psychology, 3/1 (1985), 33–42.
41
Michael Ruse, Science and Spirituality (Cambridge: Cambridge University Press, 2010),
210.
42
DM 24. But the evolutionary psychologist can object that hating fellow-members of the
group is an implausible path for the evolution of a social animal. The naked mole-rat is a good
example of a mammal that is very different from us, but whose specific form of life requires
eusociality.
43
‘Random’ does not mean ‘uncaused’, but something like ‘caused in a way that does not
predictably confer adaptive advantage’.
278 God’s Command
known as the “divine command theory” of metaethics’.44 But then the fact that
humans could have evolved differently does not give us reason to think there is
no objective value; perhaps God willed us to evolve to recognize the values
there actually are, and gave us commands to supplement the limits of this
evolutionary history.
Ruse’s version of the argument from queerness is similarly undercut by his
later concessions. He does not use the term ‘queer’, but he does hold that it is
biological theory that requires us to take the sceptical position about justifica-
tion. At the causal level, what is going on, he says, is probably individual
selection maximizing our own reproductive ends, and there is no room here
for objective rightness and wrongness. But Mackie was a feisty Australian
atheist who thought theism was, as he quoted from Hume sarcastically, a
‘miracle’. Ruse, on the other hand, aims to expose the over-reaching character
of some contemporary militant Darwinism that wants to turn science into
metaphysics, and to make science the arbiter of all truth. Darwinism, he holds,
should not try to say everything. Whether there is or is not a God Ruse says he
does not know, and science does not tell him: ‘These central core claims [of the
Christian] by their very nature go beyond the reach of science. I do not say that
you must be a Christian, but I do say that in the light of modern science you
can be a Christian. We have seen no sound arguments to the contrary.’45 To be
consistent, Ruse ought to say the very same thing about objective morality.
Mackie’s argument from queerness required the premiss that anything that
has causal relations with the world must be accessible to science. Ruse at least
sometimes now wants to deny this, and, if he does deny it, then the foundation
of the argument from queerness disappears. There is a tension in Ruse’s
thought that can be resolved by rejecting the sceptical hold-over from the
less generous views of his mentor. Here is a general principle worth empha-
sizing. Antagonism to realist claims in ethics or theology that made sense
against the background of a thoroughgoing reductive empiricism makes no
sense once that kind of empiricism is rejected.

8.3.3. Sharon Street

In 2006 Sharon Street published an article, ‘A Darwinian Dilemma for Realist

Theories of Value’, which has been the subject of a considerable literature in
reply.46 Her argument relies on the primary claim that our normative
dispositions—that is, our dispositions to form certain normative beliefs rather
than others—are (largely) selected because they had some natural property

44 45
Ruse, Science and Spirituality, 210, 214. Ibid. 233.
46
Sharon Street, ‘A Darwinian Dilemma for Realist Theories of Value’, Philosophical Studies,
127 (2006), 109–66.
 Divine Command and Evolutionary Psychology 279
(for example, they contributed to reproductive success by promoting certain
kinds of cooperation among our ancestors).47 But from the perspective of
realism, accepting this claim defeats our epistemic entitlement to our norma-
tive beliefs (for which we should assume we have title, absent any evidence to
the contrary), because we will come to be aware of the unlikely reliability of the
processes that shaped those beliefs. This is the Darwinian dilemma: the realist
has either to deny the primary claim or to concede that her normative
judgements are, by her own lights, irrational. Street is not arguing here for
ethical scepticism, or for the impossibility of ethical knowledge. Rather, she is
trying to show that, if there is to be ethical knowledge, it has to be understood
on an anti-realist model.48 Her point is that all that natural selection needs is
our beliefs in the normative facts, not the normative facts themselves. The
same is true with theological beliefs. Perhaps the belief in a deity or deities has
adaptive value because it encourages social cohesion, but it can have this value
independently of the truth or falsity of the belief. If our normative and
theological beliefs are largely products of our evolutionary history, fitness-
enhancing beliefs about morality and gods will be adopted, regardless of
whether they are, in the realist sense, true or false. Some such beliefs the
moral realist may think false—for example, the belief that killing one’s step-
children is good—but they may nonetheless promote genetic propagation.
This is the challenge. But there is a good response to it. Even if we grant that
natural selection has given us normative belief-forming dispositions that are
not truth-tracking, and that have in fact given us a mixture of ‘nasty’ belief-
forming dispositions and corresponding behaviours alongside other ‘nicer’
ones, and even if we grant that therefore our normative beliefs are unreliable to
the extent that they are given to us by natural selection, nothing follows about
how many of our normative beliefs are formed in this way.49 What is in
question here is the primary premiss of Street’s reconstructed argument: our
normative dispositions—that is, our dispositions to form certain normative
beliefs rather than others—are (largely) selected, and so on. Consider the
analogy with mathematical beliefs. To what extent do we have the ability to
track truths about non-linear algebra? The point is that, even if we get our
cognitive equipment from evolution, we can use that equipment to reach

47
I am following here the reconstruction of her argument in Dustin Locke, ‘Darwinian
Normative Skepticism’, in Bergmann and Kain (eds), Challenges to Moral and Religious Belief,
220–36.
48
She is not arguing for a general anti-realism, because she thinks natural selection has given
us reliable sense-perception, for example, and reliable common-sense inductive beliefs. I am not
trying to pass judgement on this. Alvin Plantinga has arguments against this view, in ‘The
Evolutionary Argument against Naturalism’, in J. Beilby (ed.), Naturalism Defeated? Essays on
Plantinga’s Evolutionary Argument against Naturalism (Ithaca, NY: Cornell University Press,
2002), 1–12.
49
I am following the response by William J. Fitzpatrick, ‘No Darwinian Dilemma’, in
Bergmann and Kain (eds), Challenges to Moral and Religious Belief, 237–55, esp. 241–6.
280 God’s Command
beliefs that are independent of adaptive value. Suppose we grant that natural
selection operates upon us, especially in the Pleistocene, to give us a mixture of
‘nasty’ and ‘nice’ moral beliefs. It remains possible that cultural evolution has
been operating to refine our normative stance in a truth-tracking way. We
should have some scepticism about the notion of cultural evolution, if the
phrase is intended to imply a strict analogy between cultural units and genes
and between the randomness of initial acquisition in both cases. But, if we use
the phrase loosely, we can make the point that admitting a significant initial
effect of biological evolution on belief formation does not license the conclu-
sion that natural selection is the sole force in all our belief formation there-
after. When we come to think that killing stepchildren and slavery are wrong,
it may be because we have come to see the reasons there actually are for these
conclusions, and the presence of these reasons that do not seem especially
beneficial for survival and reproduction will be counter to Street’s initial
primary claim.
The initial effect of natural selection is still relevant, because, if we were
given cognitive equipment that was hopelessly and permanently vitiated, then
we could not hope to use this equipment to discriminate subsequently between
the beliefs in the initial mixture that we should endorse and the ones we
should reject. We would be, so to speak, fatally handicapped. But there is no
reason to think our situation is hopeless in this way.50 Someone might try to
make the counter-argument that we cannot rely upon subsequent rational
reflection to purify our initial belief set, because such reflection would just be
using some thoroughly contaminated moral beliefs to correct others. But this
depends on the assumption that all our normative beliefs are saturated with
evolutionary influence, and this assumption is exactly what is in question here.
To be sure, some of our preferences and our resultant normative beliefs are
contaminated. A person may well have a love of sweet and fat food (and
accompanying beliefs about how good it is) that results from the human
situation in the Pleistocene, and that is now harmful to him. But he can also
learn how to correct at least partially for these preferences in the diet by which
he actually lives.
The thought experiment about the Epicureans, introduced when discuss-
ing Ruse’s anti-realism, is again helpful. Suppose we discovered that the
gods had set us up with our moral sentiments because it made us more
interesting to watch. This would make us sceptical about whether we should
follow the promptings of these sentiments. Dustin Locke has a similar
thought experiment:

50
There is also the possibility that there has been faster biological evolution than Street allows,
and so conditions in the Pleistocene may not be determinative. Here I am not competent to
judge, but the claim that the melanism of the peppered moth has rapidly evolved has apparently
been vindicated.
 Divine Command and Evolutionary Psychology 281
Cammie has lots of beliefs about sports. However, she did not arrive at those
beliefs in any of the usual ways (watching television, reading books, browsing the
internet, etc.). Rather, a mad scientist programmed her with various sports
dispositions—that is, dispositions to believe various things about sports. More-
over, when the scientist was programming her with these dispositions, he ran-
domly chose the propositions she would be disposed to believe from a list of some
true and some false propositions about sports.51
If she discovered this, this would destroy any initial default justification she
may have had for supposing her sports beliefs reliable. Could science show
that our situation with respect to our normative beliefs is like Cammie’s
situation with respect to her sports beliefs? A prior question is whether our
current normative dispositions are all simply products of natural selection ‘and
not (partly or wholly) products of experience, reflection and reasoning guided
by moral reality as such’.52 But this is a metaphysical question, not one proper
to science in its own domain. Ruse’s recognition of this point is what separates
him from Mackie, in the passages I quoted from him earlier. We need to
distinguish the claims of science and the claims of ‘scientism’, which is the
attempt, as Ruse puts it, ‘to make science say everything’. Metaphysical
naturalism claims baldly that there is nothing beyond physical reality, but
this is a claim that requires philosophical justification and is not within the
proper sphere of science. Street’s argument does not give us any reason to
believe that metaphysical naturalism is true.

8.3.4. Paul Bloom

This subsection is about a different kind of anti-realism—namely, anti-realism

about God. It examines the question whether evolutionary psychology gives us
any reason to doubt the existence of God. If it does, then we should not appeal
to God’s command in understanding morality. So far this chapter has by-
passed this question, because the writers we have considered who are atheists
have simply assumed that there is no God.53 This book is not about whether
God exists, or whether belief in God is rational. Nonetheless it is worth
considering a discussion by Paul Bloom of the rationality of religious belief,
since the claim that it is irrational is a presupposition of much of the literature
we have been examining.54

51
Locke, ‘Darwinian Normative Skepticism’, 229.
52
Fitzpatrick, ‘No Darwinian Dilemma’, 248.
53
Mackie has argued against theism, in The Miracle of Theism, but his arguments do not
come in E., which is the work I discussed.
54
I will be using his discussion in Paul Bloom, ‘Religious Belief as an Evolutionary Accident’,
in Jeffrey Schloss and Michael J. Murray (eds), The Believing Primate (Oxford: Oxford University
282 God’s Command
Bloom distinguishes his view both from those who say that there is direct
adaptive value to holding religious belief (such as Dominic Johnson) and from
those who say that religion is the product of culture, and not biology (such as
Pascal Boyer).55 Rather, Bloom says, religion emerges as a by-product of
certain highly structured systems that have evolved for understanding the
social world. Another term sometimes used here is that religion is a ‘spandrel
effect’, where the spandrel is the space (sometimes decorated) between the
outer curve of an arch and the angle formed by the mouldings enclosing it, so
that the spandrel does not itself bear weight. Religion would be like the ability
to understand calculus, not itself emerging because of adaptive value, but
made possible by faculties that did emerge in this way. Bloom says that he is
trying to explain universal religious beliefs here, not those that vary from one
culture to another and not the rituals associated with beliefs. Whether there
are such universal religious beliefs is a question for a different discussion.
There are two tendencies with which humans have evolved that are relevant
here. The first is what Justin Barrett calls ‘a hypersensitive agency detection
device’ (HADD).56 Our tendency to find agency around us has no doubt arisen
for survival reasons: ‘Better to guess that the sound in the bushes is an agent
(such as a person or tiger) than assume it isn’t and become lunch.’ Relatedly,
Deborah Kelemen shows evidence that children prefer functional explanations
for why rocks are pointy—for example, to keep from being sat upon—and she
calls this ‘promiscuous teleology’.57 The second tendency, less firmly estab-
lished, is that we implicitly endorse a strong substance dualism of soul and
body, of the kind defended by Plato and Descartes, and that this endorsement
is a by-product of our possession of two distinct cognitive systems—one for
dealing with material objects, the other for social entities: ‘These systems have
incommensurable outputs, and dualism emerges as an evolutionary acci-
dent.’58 These tendencies might produce a belief that there is a supernatural
agent behind natural phenomena and that this agent like our own souls is
spiritual and not bodily.
We should consider what would be the theological implication of Bloom
being right about these two side effects. He claims that the project of

Press, 2009), 118–27. I am grateful to him for a private communication in which he makes it
clear that his paper is not designed to give an argument against the rationality of religious belief.
55
Dominic Johnson, ‘God’s Punishment and Public Goods: A Test of the Supernatural
Punishment Hypothesis in 186 World Cultures’, Human Nature, 16 (2009), 410–46. Pascal
Boyer, The Naturalness of Religious Ideas: A Cognitive Theory of Religion (Berkeley and Los
Angeles: University of California Press, 1994).
56
Justin Barrett, ‘Cognitive Science, Religion, and Theology’, in Schloss and Murray (eds),
The Believing Primate, 85.
57
Deborah Kelemen, ‘Are Children “Intuitive Theists”?: Reasoning about Purpose and
Design in Nature’, Psychological Science, 15 (2004), 295–301.
58
Bloom, ‘Religious Belief as an Evolutionary Accident’, 123.
 Divine Command and Evolutionary Psychology 283
explaining why people believe in God is not necessarily an atheistic one.
Psychologists are just as interested in the origin of true beliefs as of false
ones, and the question of why people come to believe, for example, that there
is intelligent life on Mars, is simply different from the question of whether
there is such life. However, Bloom thinks that, in the case of belief in God,
psychological explanations do challenge the rationality of those who hold such
belief. He mentions Freud’s argument that it would be irrational to believe in
something just because one wanted very much for it to be true. Bloom says
that religious believers faced with the objection that their belief is actually due
to a wishful desire for a father figure, ‘would be offended, and rightfully so’, but
he is not here passing judgement on whether Freud’s account is correct.59 He
is also not arguing that religious belief is irrational because it is unfalsifiable.
A theistic evolutionist might agree that religious belief is a by-product and so
‘accidental’ from an evolutionary point of view in the same way as beliefs
about higher mathematics, but might still hold that both kinds of belief can be
true and that the religious belief is part of God’s overall plan. No finding in the
cognitive science of religion could refute this, and Bloom’s objection does not
attack the rationality of such a combination in principle.
There is, he thinks, a testable hypothesis about the mental state of someone
who believes in God. The hypothesis is that there is a sense of the divine, the
Calvinist sensus divinitatis, which is innate in all human beings, and is
triggered by exposure to certain kinds of experience, such as the glories of
the natural world.60 Bloom objects that this hypothesis is false. Belief in a single
God is not universal in the context of such experiences, and it emerges within
some societies and not others. But what is the bearing of this upon the
rationality of belief in God? The empirical evidence of polytheistic societies
(like those of the ancient Greeks or modern Hindus) would make it irrational
to hold that the evidence of the pervasiveness of religion across culture
supports monotheistic belief. But this is not the Calvinist claim. The innate
sense of the divine is supposed by Calvin to make us ‘without excuse’ if we
reject God (following Romans 1: 20). God is supposed to reveal enough of the
divine nature so that we can be held accountable, but what is the minimal
content of this general revelation? Perhaps it includes the sense that the divine
is something good and powerful and something to be worshipped, and a
monotheist (who accepts general revelation) will probably suppose that what
is revealed in general revelation will be from the one God and so polytheistic
worship is a malfunction. But there is no need for someone who believes in
general revelation to suppose that all people have an innate sense (triggered by

59
Ibid. 125.
60
Bloom has Alvin Plantinga in mind as the defender of such a view, and he cites Warranted
Christian Belief (New York: Oxford University Press, 2000).
284 God’s Command
experience) that there is just one God. The evidence (much of which Calvin
knew perfectly well) suggests strongly to the contrary.61
So what is the bearing on the rationality of religious belief of the claim that
there is an explanation of such belief from the two side effects? We should ask
what kind of psychological explanation would resist being incorporated into a
larger, more comprehensive supernaturalistic explanation, and whether the
present explanation is one of these.62 It is hard to give a general account.
Perhaps this much is true. A psychological explanation of some phenomenon
would resist such incorporation if it postulated a kind of causation of that
phenomenon that would be inappropriate for God to employ. But there is no
reason to think that it is inappropriate for God to use randomness, in the sense
in which this is part of evolutionary theory.63 There is no reason to think that
God would not allow us to acquire our basic cognitive capacities by random
mutation plus natural selection.
So far this is a merely defensive manœuvre. But perhaps we can say more.
Following Justin Barrett’s work already cited, we might suggest that the
hypersensitive agency detection device is a form of access to religious belief
that fits our nature well. Throughout this book I have been arguing that the
moral law, though it cannot be deduced from our nature, fits that nature well.
Now we can suggest the same about our theistic belief acquisition. Barrett links
the agency detection device with naive physics, naive biology, and theory of
mind as ‘subsystems designed to carry out particular tasks that are important
for our species’ survival.’64 These subsystems give us non-reflective beliefs,
such as that objects cannot pass through solid objects, that animals bear young
similar to themselves, that self-propelled, goal-directed objects are intentional
agents, and that those agents act in order to satisfy desires. These beliefs are
standardly not taught to children through explicit verbal instruction.65 Con-
cepts that are ‘minimally counter-intuitive’ given the operation of these
subsystems will seem plausible, and will be easily remembered and transmit-
ted. This does not mean that these subsystems always yield true beliefs; we
cannot deduce the truth of a belief from its deliverance by one of these
subsystems. But these beliefs fit our nature, as constituted by these systems,
exceedingly well. For example, belief in a super-knowing god may be natural

61
See Plantinga, Warranted Christian Belief, 177, on belief de re about God.
62
I put the question this way following Peter van Inwagen, in ‘Belief in the Supernatural’, in
Schloss and Murray (eds), The Believing Primate, 135. He suggests that God using evolution
might be like a human designer of a vehicle using the waste heat from its engine as a way to keep
the passengers warm.
63
For a controversial discussion of the claim that God might use chance, see Peter van
Inwagen, ‘The Place of Chance in a World Sustained by God’, in Thomas V. Morris (ed.), Divine
and Human Action (Ithaca, NY: Cornell University Press, 1988), 42–65.
64
Barrett, ‘Cognitive Science, Religion, and Theology’, 79.
65
Cf. Wittgenstein, Philosophical Investigations (Oxford: Blackwell, 1963), #241: ‘That is not
agreement in opinions but in form of life.’
 Divine Command and Evolutionary Psychology 285
to us, given that we start off from ‘Theory of Mind’s default assumption that
[all] agents are super-knowing’.66 This would partially explain why children
are, in Kelemen’s term, ‘intuitive theists’. If Bloom is right about our innate
tendency to substance dualism (because of the incommensurability of our
cognitive systems designed to deal with social systems and physical objects),
the same sort of argument would apply. Moreover, Barrett suggests plausibly
that the connection between God and moral concerns is also intuitive.67
Fortunate and unfortunate events inevitably happen, and call for an explan-
ation, and, given the minimally counter-intuitive character of the concepts of
gods and super-knowledge, and given the intuitive character of general rules
forbidding ‘murder, adultery, theft, deception, treachery, and cowardice’, it is
natural to think in terms of gods commanding these rules, and punishing and
rewarding in accordance with them. In other words, the theist can legitimately
hold that God chooses means for our access to divine command that are not
inappropriate but entirely fitting to our nature, the kind of means that we
would expect creatures with cognitive subsystems like ours to use.68 The final
section of Chapter 5 suggested some ways internal to theology for correcting
the beliefs that these subsystems can give us. We should conclude that at least
from the evidence marshalled in the present section, there is no demonstration
that belief in God is irrational.

8.4. TRANSCENDING OUR EVOLUTIONARY

SITUATION WITHOUT GOD

This section treats attempts to bridge the gap between natural selection and
moral obligation without bringing God into the picture. These attempts
have an interesting history, but it would make the chapter too long to go
into it now.69 This part of the chapter looks at two figures, Joshua Greene
(Section 8.4.1) and Philip Kitcher (Section 8.4.2).

66
Barrett, ‘Cognitive Science, Religion, and Theology’, 91. 67
Ibid. 88.
68
Bloom has suggested in private communication that it would be ‘sneaky’ of God to use a
technique of communication that would give the same result if God did not exist. But the
situation is more like a medical test that gives a large number of false positives, but gets the right
diagnosis in the key cases.
69
I had intended to write about John Stuart Mill, and his extraordinary optimism in
Utilitarianism, ed. Mary Warnock (London: Fontana (Collins), 1962), that ‘education and
opinion, which have so vast a power over human character, should so use that power as to
establish in the mind of every individual an indissoluble association between his own happiness
and the good of the whole’. I would then have discussed the relation of this to his views in Three
Essays on Religion, 249–50, which defends a version of what I called in Chapter 1 ‘the argument
from providence’, and says that ‘the indulgence of hope with regard to the government of the
universe and the destiny of man after death, while we recognize as a clear truth that we have no
286 God’s Command

8.4.1. Joshua Greene

We begin with Joshua Greene, and especially his book Moral Tribes.70 The
governing metaphor of this book is that of two tragedies, which Greene calls
‘the tragedy of the commons’ (after Garrett Hardin) and ‘the tragedy of
common-sense morality’.71 In Hardin’s parable, the tragedy of the commons
is produced when a single group of herders shares a common pasture, which
can provide food for only a limited number of animals. With every animal
that is added to the pasture its herder gets that animal’s additional price at
market, but the cost of supporting that animal is shared by all who use the
commons. It is rational, therefore, for each herder to go on adding animals,
and soon the grass on the commons will be completely used up. The
dilemma here is a version of a multi-person cooperation problem.72 Greene
proposes that morality is given us by evolution as a set of ‘gizmos’ in our
affective machinery that enables us to solve this kind of dilemma by cooper-
ating because we can trust each other to cooperate at least to some signifi-
cant degree. Two people would be able to trust each other in this way if they
had some confidence that they were both committed to some version of the
Golden Rule. But what would cause two people to care about doing or not
doing hateful things to each other in violation of such a rule? Greene
suggests, first, that the two people may be related and tied by ‘brotherly
affection’, or they may have a ‘tit-for-tat’ agreement that involves expected
reciprocity, or they may be friends, or they may care about reputation, or
they may fear the other’s built-in irrational desire for vengeance and retali-
ation. But, in addition, he thinks we have at least a small amount of care for
strangers and a readiness to help them ‘hard-wired’ into us, and he claims
that such ‘neighbourliness’ can be found in other primates and even in
capuchin monkeys.73 The problem is that tribal loyalty and self-interest
are stronger. For the first of these (tribal loyalty), Greene quotes the anthro-
pologist Donald Brown, whose survey of human cultural differences and
similarities identified in-group bias and ethnocentrism as universal.74 For
the second (self-interest), he quotes studies on what he calls ‘biased fairness’
in which our perception of reality and fairness is unconsciously distorted by
self-interest.75

ground for more than a hope, is legitimate and philosophically defensible’. Mill goes on to say
that without such hope we are kept down by ‘the disastrous feeling of “not worth while” ’.
70
Joshua Greene, Moral Tribes (New York: Penguin, 2013) (henceforth MT).
71
MT 1, 19. See Garrett Hardin, ‘The Tragedy of the Commons’, Science, 162 (1968), 1243–8.
72
The most famous example is the ‘prisoner’s dilemma’, in which cooperating is better for the
two prisoners collectively, but individually each prisoner is better off defecting. For the details of
the prisoner’s dilemma, see MT 28–30.
73 74 75
MT 39. MT 49. MT 84.
 Divine Command and Evolutionary Psychology 287
The ‘tragedy of common-sense morality’, on the other hand, results from a
higher-order dilemma. Greene imagines different tribes who have come to
adopt different economic and social practices, because their moral gizmos are
differently ordered. For example, some tribes are more egalitarian than others,
some have more active central governing structures than others, and some
have exclusive religiously based moral practices. What seems like moral
common sense within each tribe is different, and this leads to bloody conflict
between them when circumstances (a huge fire) require them to share the
same territory. The point of this parable is that the situation of these tribes is
our situation. We have different common-sense moralities, and this results in
conflict between us. What we need to find is a metamorality that can adjudi-
cate these disputes. It will help us, Greene thinks, to see what our moralities are
in origin—namely, mere evolutionary gizmos—so that we can more easily
transcend them, and move from nature to what he calls an ‘unnatural’
solution.76 Once we see the forces that gave rise to morality, we can ‘climb
the ladder of evolution and then kick it away’, as Wittgenstein says about his
method in the Tractatus.77 Greene argues that the unnatural metamorality we
should end up with, after kicking away the evolutionary ladder, is utilitarian-
ism. This is because utilitarianism trades only in the currency that is common
to all the tribes, and that currency is happiness and its maximization.
This picture raises three questions, deriving from the three arguments in
Chapter 1: the arguments from providence, grace, and justification. We can
take them in reverse order, because the answer to the question about justifi-
cation sets the frame for the other two. The first question is ‘Why should
I regard the conclusions of this metamorality as binding upon me?’ This is
Korsgaard’s so-called normative question, and Chapter 1 proposed that the
divine command theorist has an answer. The second question is ‘How can
I move to this metamorality, given that I am the mixture of motivations that
Greene has described?’ Chapter 1 proposed that the mixture that Kant calls the
predisposition to good and the propensity to evil makes such a move un-
attainable by our own resources, but attainable by God’s assistance. The third
question is ‘How can I reasonably believe that moving to this metamorality is
consistent with my own happiness, if it does not seem that other people are
moving in the same way?’ This question focuses on the cost of the moral
demand, construing it as Greene does in a utilitarian way. Chapter 1 suggested
that we get an answer to this in the doctrine of providence.
The first question asks for a justification. How can Greene justify the claim
that we should live under his form of the moral demand? He cannot appeal to

76
MT 147.
77
MT 25. See Wittgenstein, Tractatus Logico-Philosophicus (London: Routledge & Kegan
Paul, 1961), vi. 54.
288 God’s Command
religion (although he sometimes expresses some ambivalence here).78 The
reason is that he is an atheist.79 He says:
In short, the world’s religions enable their adherents to avert the Tragedy of the
Commons, to put Us ahead of Me. What religions don’t do—most of them at
least—is help us avert the Tragedy of Commonsense Morality. They exacerbate,
rather than ease, conflicts between the values of Us and the values of Them. For
our common currency, we must look elsewhere.80
But the exclusion of religion here is unfortunate, because it deprives him of
resources for justification that he needs. He is in fact ambivalent, as shown in
the phrase ‘most of them at least’. In this place he allows just one exception,
the Unitarian Universalist Church. But elsewhere he reports an episode related
in the Talmud, where Rabbi Hillel was asked to teach the entire Torah in the
time a person is able to stand on one foot. Hillel replied: ‘That which is hateful
to you, do not do to your neighbor. That is the whole Torah. The rest is
commentary. Go and study it.’81 This is not exactly the Golden Rule, but
Greene allows that versions of the rule are found in every major religion, and
that the Golden Rule (for example, at Matthew 7: 12) is not formulated in
terms of Us and Them. Most of the world’s population already belongs to one
of the major religions. It is hard to find accurate figures, but one estimate is
that, by 2050, 80 per cent of the world’s population will belong, at the present
rate of change.82 Surely we should be looking at the resources of those religions
to see if they can help us with common currency. My book is intended as a
contribution to this.
It is significant here that Greene has distorted the history of utilitarianism
by excising its religious roots. He says that it was founded by Bentham and
Mill.83 He ignores Hutcheson, who first writes of ‘the greatest Happiness for
the greatest Numbers’, and especially Paley, whose work preceded Bentham
and indeed the success of whose book at Cambridge provoked Bentham to

78
Greene also cannot use any justification that requires moral realism, because he, like
Mackie and Ruse, is an anti-realist. He says that recent evidence from neuroscience indicates
that moral judgement is often an intuitive, emotional matter, and that we know that this kind of
intuition is often ‘accompanied by a perceptual phenomenology’ (‘From Neural “Is” to Moral
“Ought”: What are the Moral Implications of Neuroscientific Moral Psychology?’, Nature
Reviews Neuroscience, 4/10 (2003), 849). This gives the beginnings of a debunking explanation
of moral realism, he suggests. But this is because he already assumes that perceptual-like
phenomenology can be veridical only in the case of physical objects, thus begging the question.
79 80 81
MT 179. MT 183. MT 30–1.
82
Micklethwait and Wooldridge, God is Back, 16. They say that the proportion of the world’s
population in 1905 was 67%, in 2005 73%, and at the current rate of change it will be 80% by
2050. We need a touch of scepticism about what counts as ‘belonging’ to a religion. But even so,
the figures are striking.
83
MT 155.
 Divine Command and Evolutionary Psychology 289
write his own version of the theory.84 The point is that utilitarianism starts
with Christians, and works out the view that, as Butler puts it, benevolence,
especially God’s benevolence, ‘seems in the strictest sense to include in it all
that is good and worthy’.85 Bentham, but not Mill, is cutting himself off from
the roots of his own theory. Indeed, the prizing of benevolence is ‘common
currency’ to all areas of the world in which the five major religions have
established a significant presence, though this claim needs support from
empirical research.86
What is Greene’s answer to the normative question? There are various
question-begging answers. One is that strengthening our sympathies for
distant strangers is ‘the honest response, the enlightened response’ to world
hunger.87 But the striking thing is that he does not squarely face this question.
The closest passage to an answer is right at the end of the book. In the final
paragraph he starts by saying that he does not wholeheartedly share Kant’s
marvelling at ‘the moral law within’, because he sees the moral laws (sic)
within us as a mixed blessing (since the fact that we have different ones leads
to ‘the tragedy of commonsense morality’). More marvellous to him is that
we can
question the laws written on our hearts and replace them with something better.
The natural world is full of cooperation, from tiny cells to packs of wolves. But all
of this teamwork, however impressive, evolved for the amoral purpose of suc-
cessful competition. And yet somehow we, with our overgrown primate brains,
can grasp the abstract principles behind nature’s machines and make them our
own. On these pastures, something new is growing under the sun: a global tribe
that looks out for its members, not to gain advantage over others, but simply
because it is good.88
There is an answer that is implied here: the love of what is good simply
because it is good, which Scotus calls the affection for justice, is already one
of the abstract principles behind nature’s working, and if we make it our own
we are being true to what this nature has made us. But there is a problem here.
There is another abstract principle behind nature’s working—namely, com-
petitive self-replication—and nature is a mixture. We cannot generate a

84
Francis Hutcheson, Inquiry into the Original of our Ideas of Beauty and Virtue, ed.
Wolfgang Leidhold (Indianapolis: Liberty Fund, 2004), 177–8. Sidgwick, in Methods of Ethics,
mentions Paley at the beginning and returns to him at the end, p. 509, with the suggestion
(though not the endorsement) of a god whose providence could make egoistic hedonism and
universalistic hedonism consistent.
85
Butler, Fifteen Sermons, 153.
86
Here is an illuminating anecdote about Buddhists releasing fish, which are then recaptured by
their irreligious neighbours <http://www.npr.org/blogs/parallels/2015/01/02/373286111/along-
shanghais-river-buddhist-tradition-meets-greedy-fishermen> (accessed February 2015).
87
MT 352.
88
I see the same kind of answer implicit at MT 65: ‘Out of evolutionary dirt grows the flower
of human goodness.’
290 God’s Command
justification of the obligation to follow a universalistic moral demand just from
the principles behind nature’s working, because we need to know which
principles.
The second question is ‘How can I move to this metamorality, given that
I am the mixture of motivations that Greene has described?’ Here Greene
again does not give us an answer, and he concedes that ‘our brains were not
designed to care deeply about the happiness of strangers’.89 He gives us, like
Haidt, sceptical arguments about the powers of our reason. Here is one
passage.
What’s more, the DLPFC [Dorsolateral Prefrontal Cortex], the seat of abstract
reasoning, is deeply interconnected with the dopamine system, which is respon-
sible for placing values on objects and actions. From a neural and evolutionary
perspective, our reasoning systems are not independent logic machines. They are
outgrowths of more primitive mammalian systems for selecting rewarding
behaviors—cognitive prostheses for enterprising mammals. In other words,
Hume seems to have gotten it right.90
What has Hume got right? We should remember Haidt’s continual trumpet-
ing of Hume’s victory over Kant; reason is the slave of the passions (the rider is
controlled by the elephant). Greene agrees that reason is the slave, but he
wants to allow more space than Haidt to reason; he wants reason to be able to
transcend the emotions, which he regards as automatic processes that tell us
what to do.91 But, if our reasoning process starts from emotional inputs as its
premisses, and this input is contaminated in the way Greene says it is, how is
the processing supposed to give us pure utilitarian theory as its output for how
we should live our lives? As they say, garbage in, garbage out. Perhaps we
should simply say that we do not yet know how this is supposed to work, and
we are dealing with a mysterious emergent property. But Hume is a telling case
here. He concedes that, if we had a society in which those whom we exploited
were not able to harm us because of their weakness, we would not be moved by
any abstract principle of justice to end the exploitation, even if they resented it.
This is Hume’s conclusion, that no inconvenience would result from the
exercise of such a power, and therefore the restraints of justice would be
totally useless.92 We might hope to be moved by the calm passions of
compassion and kindness, but the reach of our natural endowment of these
is, as Greene acknowledges, significantly limited. What is supposed to get us to
accept a higher standard?
What creates the problem here is the combination of optimism about the
new metamorality with pessimism about the input processed by our

89 90
MT 257. MT 368, footnote to p. 137.
91
MT 135. Greene has not mirrored Haidt’s move from the language of ‘emotion’ to the
language of ‘intuition’.
92
See Harrison, Hume’s Theory of Justice, 276.
 Divine Command and Evolutionary Psychology 291
reasoning. One solution is to be more optimistic about the sentiments. Frans
de Waal has criticized the denigration within sociobiology of human moral
capacity, and called this kind of denigration ‘Calvinist’, tracing the view back
to Calvin’s picture of the total depravity of human beings.93 He himself prefers
to think that our human ancestors had rules of right and wrong and tenden-
cies to help those in need from the beginning, pre-dating our modern religions
by hundreds of thousands of years (he does not say the rules pre-dated religion
in general). The roots of morality, he thinks, lie in empathy and reciprocity,
and are already present in primate sociality: ‘Human nature is not all selfish
and nasty, and we do not need religion to tame us into becoming moral beings.
We are evolutionarily equipped with moral sentiments, which put the virtuous
life within reach. In the effort to attain it, we are given an enormous helping
hand from our background as social primates.’94 For de Waal, the philosoph-
ical defender of moral sentiments is again Hume, and the enemy is Kant: ‘One
of the momentous developments of our time is the effort to wrest morality
from Kantian philosophy and put it back in touch with evolution.’ However,
de Waal is not consistent in what he says about religion. He concedes that
there is no human culture without religion, though humans had social norms
before they had our current major religions, and he says that, if we were able to
excise religion from society, it is doubtful ‘that science and the naturalistic
worldview could fill the void and become an inspiration for the good’.95 This
means that ironically, in terms of the second question, at least sometimes he
says we need religion (just as Kant does), even though he is not himself a
religious person. It also means that our sentiments in the absence of religion
are not sufficient to take us to a morally good life.
The third question we can deal with more briefly. How can I reasonably
believe that moving to this utilitarian metamorality is consistent with my own
happiness, if it does not seem that other people are moving in the same way?
In Chapter 1, this was the question that generated the argument from provi-
dence. We can find versions of this argument in both Mill and Sidgwick. For
Mill, the key text is Three Essays on Religion. For Sidgwick, the argument
comes at the very end of Methods of Ethics, and is not endorsed as a solution,
although the problem it addresses is recognized as a real problem: A utilitarian

93
De Waal, Good Natured, 17. De Waal is right that there has been what he calls a ‘veneer
theory’ within sociobiology, which holds that morality is a veneer, masking a core of self-interest,
but the construal of Calvin needs to be corrected. The doctrine of total depravity means that
every part of us is innately partly corrupt, but there remains ‘a seed of religion’ that survives in all
human beings, though it does not produce the fruit of a life pleasing to God without God’s
assistance (see Calvin, Institutes, I. 4. 1). Kant’s actual view of the innate mixture of predispos-
ition to good and propensity to evil is consistent with, though not the same as, Calvin’s
actual view.
94
De Waal, ‘The Animal Roots of Human Morality’.
95
Frans de Waal, ‘Morals without God?’, New York Times Opinionator, 17 October 2010
<http://opinionator.blogs.nytimes.com> (accessed May 2011).
292 God’s Command
needs to have something to say about how prudence (understood as the
pursuit of one’s own happiness) is consistent with the moral demand (under-
stood as the pursuit of the greatest happiness of the greatest number). This is
especially true if, like RMH’s defence of utilitarianism described in
Section 4.3.3, the defence of the view starts from prudence and then univer-
salizes it. RMH, like Mill, defended a version of the argument from provi-
dence. Without an argument like this it is not clear how Greene can hold that
his utilitarian metamorality and the pursuing of individual happiness are
consistent.

8.4.2. Philip Kitcher

Philip Kitcher’s book The Ethical Project is ideally suited to the purposes of the
present chapter.96 Kitcher argues for a pragmatist naturalism that is governed
by the principle ‘No Spooks’. ‘Spooks’ include God, but also ‘a realm of values’
(of the kind that Mackie calls ‘queer’), ‘faculties of ethical perception’, and
‘pure practical reason’ (in Kant’s phrase). His book is not an argument against
belief in God, just as my book is not an argument for belief in God. I will come
to Kitcher’s positive proposal for getting to ethics without religion at the end of
this subsection. But he gives briefly two reasons for denying the existence of
God, and I will briefly respond to them.97 They are (first) that not all religions
can be true because they contradict each other, and there is no core set of
doctrines holding them all together, and (second) that the methods by which
people reach religious belief are unreliable. With respect to the first of these
claims, this book has not been about religion in general (since there is probably
no single account of what counts as a ‘religion’ that fits all the phenomena
usually described as religions), but about the Abrahamic faiths in particular.
There is a great deal of overlap here, though there are also substantial differ-
ences. This book has also defended a view of general revelation according to
which all human beings get enough revelation of the divine nature to be without
excuse if they reject God, even though they do not all have an innate sense
(triggered by experience) of a single God. But it has left vague what the minimal
content of this revelation might be. In general, it does not follow from the fact
that some set of beliefs contains beliefs that contradict each other that they are
all false, any more than disagreements across time about scientific claims show
that all scientific claims are false. What is needed in both cases is some account

96
Philip Kitcher, The Ethical Project (Cambridge, MA: Harvard University Press, 2011)
(henceforth EP). This subsection of the chapter has profited greatly from private communication
with him.
97
Kitcher’s arguments are given more fully in the last chapter of Living with Darwin (New
York: Oxford University Press, 2007) and the first chapter of Life beyond Faith: The Case for
Secular Humanism (New Haven: Yale University Press, 2014).
 Divine Command and Evolutionary Psychology 293
of a method by which errors can be removed. As to the claim about the
unreliability of the methods by which humans reach their religious beliefs,
Section 8.3.4 followed Justin Barrett’s suggestion that the methods are natural
to us, though not infallible. The last section of Chapter 5 described some of the
ways internal to theology for correcting some of these beliefs. The belief in a
divine being or beings standardly brings with it beliefs about how this being or
beings communicate with human beings. So we cannot settle the question of
whether the communication with that divine being is reliable independently of
a view about the existence of that being. In any case, the important question for
the present chapter is not the truth of atheism but what follows for ethics from
the assumption that God does not exist.
Kitcher starts from a distinction between different types of altruism. The
most important for the understanding of the ethical project is what he calls
‘psychological altruism’, which differs from ‘biological altruism’ and ‘behav-
ioural altruism’ because it involves the intention to promote what are taken to
be the wishes or the interests of others. Biological altruism involves acting in
ways that decrease an organism’s own reproductive success and increase the
reproductive success of another organism. Behavioural altruism involves
acting towards a beneficiary in ways that detract from the fulfilment of one’s
own desires and promote the fulfilment of the desires of the beneficiary; it does
not require any particular motivation for such behaviour. Kitcher then sug-
gests that ethics arises as a means of reducing psychological–altruism failure.
In the kinds of groups that we can imagine our first human ancestors to have
formed, it was crucial for survival to be able to trust each other not to defect
from the various forms of cooperation that constituted their way of life.
One key step in this development is what Kitcher calls ‘normative guid-
ance’, which is defined in terms of ‘the ability to apprehend and obey
commands’.98 He thinks that much of this normative guidance ‘may have
been mediated by respect for the supposed commands of transcendent
beings’.99 In other words, Kitcher makes the reception of ‘supposed’ divine
commands central to the development of ethics, even though he thinks there
is no transcendent being to give such commands.100 This is one reason his
book is so suited to the present chapter. He makes it clear that he thinks fear
is the central original motivation, the fear of divine punishment: ‘Unless
there were sanctions for disobedience, fear could hardly be central to the
initial capacity for normative guidance.’101 This fear then gets internalized as

98 99
EP 74. EP 84.
100
EP 115: ‘Far from being an irrational idiosyncracy, divine-command approaches to ethics
may reflect a deep fact about cultural competition.’
101
EP 87 (emphasis added). Kitcher does not deny that other emotions than fear may be
involved in ancient religion. But he thinks fear is central to normative guidance. My suggestion is
that something like awe is the primary emotion.
294 God’s Command
conscience, and ‘the commanding voice seems to come from within, initially
and crudely as the expression of fears’.102
There is, however, a difficulty with this account. On the supposition that
our original human ancestors were hunter-gatherers, it is important to notice
that the hunter-gatherer societies that we know about do not, on the whole,
have moralizing high gods. This claim is supported by consulting the ‘Stand-
ard Cross-Cultural Sample’ by George Murdock and Douglas White, which
surveyed over 1,250 societies, grouped them into clusters ‘with cultures so
similar . . . that no world sample should include more than one of them’, and
then listed a representative society of each of the remaining 186 sampling
provinces.103 All 186 societies have religions of one kind or another. Murdock
and White define ‘religion’ broadly to include worship of either ‘high gods’ or
else lower grades of divinity, ancestors who are still active and witches and
sorcerers.104 Of the 168 societies (out of the 186) in which there is enough
information to determine whether there is a high god or not, approximately
100 have high gods, of which three groups are distinguished. One is where a
high god is present, but is not concerned with human affairs. One is where a
high god is present and active in human affairs but not offering positive
support to human morality. The third is where a high god is present, active,
and specifically supportive of human morality (approximately 40 per cent of
societies with high gods). For present purposes, the key point is that we can
identify, on the cross-cultural sample, those societies that are, so to speak,
Pleistocene-appropriate.105 We will be looking for hunter-gatherers. If we
record the information from the cross-cultural sample and cross-check with
the Ethnographic Atlas, we can exclude societies with classifications including
advanced agriculture, horticulture, simple or shifting cultivation, domestic
animals, and exchange economies. Also we can look for independent local
communities, or stateless societies, excluding a single level of political inte-
gration transcending the local community, and two or more levels of supra-
community integration. After these exclusions, there are twenty three societies
left in the sample, and of these only one has ‘moralizing high gods’, the Yahgan
or Yamana.106

102
EP 94.
103
George P. Murdock and Douglas R. White, ‘Standard Cross-Cultural Sample’, Ethnology, 8
(1969), 329–69, esp. 331.
104
A ‘high god’ is defined as ‘a spiritual being who is believed to have created all reality and/or
to be its ultimate governor, even though his sole act was to create other spirits who, in turn,
created or control the natural world’.
105
I am grateful for work done on this question by Erik Santoro.
106
George P. Murdock, Ethnographic Atlas (Pittsburgh: University of Pittsburgh Press, 1967).
The University of Kent website <http://lucy.ukc.ac.uk/EthnoAtlas/ethno.html> (accessed Sep-
tember 2012), with more recent information, lists the settlement pattern of Twana (Nb2) as
semi-sedentary, and this reduces the number from 24 to 23.
 Divine Command and Evolutionary Psychology 295
Why does this matter? It matters because it suggests that worship of the
divine is much older than the narrative about an ‘unseen enforcer’ implies. The
idea that humans invented gods in order to enforce the law has a long tradition
behind it. Critias, the ancient Greek sophist, proposed that ‘some shrewd and
wise man invented fear of the gods for mortals, so that there might be some
deterrent to the wicked even if they did or said or thought something in
secret’.107 But the anthropological evidence does not support this. The soci-
eties that did not have moralizing high gods may have had ‘enforcers’, but
equally some emotion other than fear of punishment may have been the
primary emotion involved in their religion. Something like awe or respect or
reverence is a good candidate, and this was built into the story at the beginning
of this chapter. This would make ancient religion more continuous with our
own. We would then need to ask what accounts for this phenomenon. The
next section will suggest that an encounter with God is a good explanation,
though not the only one. What is striking in Kitcher’s account is the absence of
any recognition, especially for educated people, of the human desire for the
divine. When he lays out the shape of a possible good life, he says: ‘Educated
people may find the shape of their own lives (and of their selves) in a mix of
intellectual work, political activity, appreciation of nature and of art, and
(perhaps above all) communion with others.’108 But he has already dismissed
the ideal of a ‘spiritually fulfilled life’ as based on utterly false doctrines (about
gods). It is striking that a central desire of so many of the world’s people both
educated and not, and both now and in our history, is here excluded.
Another reason for worrying about making fear of punishment central to
religion is that this makes it contradictory to think, in Kant’s phrase, of
‘recognizing our duties as divine commands’.109 He gives an argument in the
Groundwork, directed partly against Crusius, that we cannot base our duties on
fear of divine punishment.110 But this is quite different from respecting God as
the head of the kingdom of ends, who can maintain the system in which good is
rewarded and evil is punished. Kant’s view of this matter is parallel to his view of
punishment by the state. The moral agent needs the state to punish, but not
because her moral motivation is fear of punishment. Rather, she values freedom,
and values punishment as a ‘hindrance to the hindrances to freedom’.111 By
analogy, the moral agent is not given moral motivation by fear of God’s
punishment or fear of hell. Rather, she aims at the highest good (the union
of virtue and happiness), and this requires the belief that the system by

107
Sextus Empiricus, Adversus Mathematicos IX, 54 (DK 88 B 2), trans. John Robinson, An
Introduction to Early Greek Philosophy (Boston: Houghton Mifflin, 1968), 270.
108
EP 315. There is a similar God-excluding list of ‘the ends of life’ in Griffin, Well-Being, 67:
accomplishment, agency, understanding, enjoyment, and deep personal relations.
109
Rel. vi. 154.
110
Gl. iv. 443. I have discussed this passage in detail in Hare, God’s Call, 87–119.
111
MdS vi. 396.
296 God’s Command
which virtue is consistent with happiness is in place and the apparent dispro-
portion of virtue and happiness that we experience in this life is not final. This is
the argument from providence discussed in Chapter 1. However, as with life in
the state, she has to live in a world in which not all her fellow humans are law-
abiding, and some of them can be deterred by the threat of punishment. This
enables us to distinguish two different motivations. One is fear, because ‘pun-
ishment can force the costs of free-riding above the costs of cooperation’.112 The
other (more satisfactory to the Kantian) is hope; a belief in punishment is part of
a belief in a ‘world morally governed’.113 There is a difference between being
motivated by fear of divine punishment and being motivated by love of justice,
which is a system that divine punishment maintains.
This consideration about Kantian morality will not persuade Kitcher, be-
cause he does not accept what he takes to be Kant’s views about moral
motivation. Thinking primarily of Kant, he says: ‘The entire conception of
the “ethical point of view” is a psychological myth devised by philosophers.’
However, his picture of Kant assumes that ‘pure practical reason’ (and so the
ethical point of view) requires that we not ‘recognize our duties as God’s
commands’. This is wrong about Kant, as just argued. Kitcher in fact preserves
the essential structure of Kant’s moral system, in requiring that psychological
altruism have no hidden ‘Machiavellian’ or self-indexed motivation, and that
morally endorsing desires involves ‘there being possible environments in
which they could be satisfied for all our fellows’.114
When Kitcher comes to consider concrete cases where ethical decision is
influenced by religious faith, he is concerned to deny that these cases involve
anything like ethical ‘insight’ (because this would be a ‘spook’). The case of the
Quaker John Woolman’s coming to see the evil of slavery is a key example.115
Kitcher quotes Woolman’s Journal, in which he says, after writing and signing
a bill of sale, ‘I was so afflicted in my mind that I said before my master and the
Friend that I believed slave-keeping to be a practice inconsistent with the
Christian religion,’ and later (after refusing to sign another bill of sale) ‘that
[the slaves] are human creatures, whose souls are as precious as ours, and who
may receive the same help and comfort from the Holy Scriptures as we do’.
The defect in Kitcher’s account of this case is that he does not locate it against
the background of the First Great Awakening, when George Whitefield came
to New England in 1740. One of the extraordinary features of this revival was
that slaves and their masters experienced the same outpouring of the Spirit. So

112
Johnson, ‘God’s Punishment and Public Goods’, 411.
113
The phrase is from Sidgwick, Methods of Ethics, 505.
114
EP 22, 223. For a Kantian, however, ‘fellows’ has to be extended to all members of the
kingdom of ends.
115
EP 158–9. Kitcher’s target in this passage is not realism about God but realism about value.
But my response applies to both.
 Divine Command and Evolutionary Psychology 297
Frank Lambert can say: ‘I saw the Book talk.’116 This is why Woolman can
report receiving the same comfort from the Holy Scriptures. Kitcher has two
reasons for denying that there is ‘insight’ here; one is that Woolman is
reflecting on the New Testament and not directly on experience, and the
other is that he does not mention the name of the slave whose sale ‘afflicted’
his mind. But neither reason is persuasive. First, religious experience is
standardly mediated by sacred tradition, and we come to see some goodness
as the sacred story reveals it. Second, insight can come from experience gained
about human beings as such as well as about an individual. Kitcher says that
Woolman does not ‘indicate how readers might put themselves into a position
to apprehend the simple and irreducible wrongness of owning other
people’.117 But he does. He indicates that they should read the passages
about Jesus’s compassion on the outcast, and think of slaves as those for
whom Christ died; then they will reach the same insight that he did.
Kitcher, having accepted that divine command theory may reflect a deep
fact about cultural competition, nonetheless rejects it. He has four main
objections.118 The first is Plato’s argument from the Euthyphro. As we saw
in Section 1.3, however, Plato does not actually give an argument for the key
premiss that the holy is loved by the gods because it is holy. Socrates simply
gets Euthyphro to agree to this without argument, and then argues to the
conclusion that the holy and the god-loved cannot be the same. If we try to
supply an argument, the best option is to go back to the previous argument,
where gods (and humans) are said to disagree or agree because they see in the
same thing different value properties. But Kitcher will not like this argument,
because it relies on a spook. The main problem here is that he has not
considered the versions of divine command theory that navigate between
the horns of Plato’s dilemma. Mackie had already seen how to do this, and
there are excellent versions in Adams and Evans.119
A second objection is that we get an infinite regress if we ask ‘Why should
we obey a divine command?’ I have discussed this already in relation to
Darwall in Chapter 1, and gave the reply from Scotus that the principle that
God is to be loved (and so obeyed) is known from its terms (and so does not
require prior justification).
A third objection is from horrible commands such as the commands to kill
Isaac or slaughter the Canaanites. If we simply follow orders, even orders from
God, are we not like the Nazis who obeyed the order to send victims to the gas
chambers? This objection ignores the sequence of the biblical narrative. As

116
Frank Lambert, ‘ “I Saw the Book Talk”: Slave Readings of the First Great Awakening’,
Journal of African American History, 87 (2002), 14. There was not, shamefully, a continuation of
the fellowship implied in this experience.
117 118
EP 196. The key passage is EP 166–70.
119
See E. 230–1, Adams, Finite and Infinite Goods, ch. 11, and Evans, God and Moral
Obligation, 89–91.
298 God’s Command
I suggested in Chapter 5, we have to take the sacred texts as a whole. If a
Christian now takes a command to kill her children or massacre her enemies
(as a people) to be God’s command, she is being unfaithful to the narrative.
She is now not in Abraham’s or Joshua’s situation because she now has
Abraham’s story and the story of Jesus and the Canaanite woman (as dis-
cussed in Section 5.6).120
There is a fourth objection, located earlier in The Ethical Project but relevant
here. This is the claim that religion leads to hierarchy of an oppressive sort,
and so undermines what Kitcher takes to be our initial situation of equality:
‘Those who can convincingly claim to have special access to the will of the
transcendent policeman—shamans, priests, and saints—come to have an
ethical authority others lack.’121 But, as far as we can tell, the hunter-gatherers
had religion but did not have this kind of hierarchy, and so the relation cannot
be essential. In fact, religious history shows a large fluctuation in political
structure, varying from ‘the priesthood of all believers’ to the cult of Jim Jones.
There is no reason to think that the tendency to oppressive hierarchy is
essential to the reception of divine command, and the position may be more
like the relation between religion and violence. Religion, just like any social
phenomenon, can be used for violent and oppressive purposes, but also for
peacefulness and inclusion. We can add that the corruption of the best is often
the worst.
Before leaving Kitcher, there is one more point that needs to be made,
perhaps the most important point. He gives us a method for ethics, and, if it
works, it promises to give us an answer to the normative question: ‘Why
should we be moral?’ The method depends on two basic principles, both
drawn from analogy with an account of progress in science. The first is that
‘an ethically adequate discussion discloses those features of an ideal deliber-
ation under conditions of mutual engagement that would prompt each par-
ticipant to reach consensus’.122 The notion of mutual engagement is key here,
and the conditions that have to be met before mutual engagement include the
second principle, that in their deliberations the participants do not rely on any
false beliefs about the natural world. The answer to the normative question
derives from the claim that we humans have throughout our history had an

120
I have not dealt with a central difficulty: was Joshua right to think God was commanding
the slaughter? Perhaps the whole conquest narrative is fiction. But if we do not want to say this,
there is some help in Nicholas Wolterstorff, ‘Reading Joshua’, in Bergmann et al. (eds), Divine
Evil? The Moral Character of the God of Abraham, 236–65. He compares the accounts in Joshua
and Judges of the same period, and concludes that Joshua is hyperbolic. But I do not pretend the
problem is solved. There is an excellent and provocative treatment in Baggett and Walls, Good
God, 125–42.
121
EP 115.
122
EP 343–4. The participants also know the consequences of their decisions, and they know
each other’s preferences. Their situation is in these ways like that of the archangel in RMH’s
account, discussed in Section 4.3.3.
 Divine Command and Evolutionary Psychology 299
ethical project whose method can be idealized in this way, and that we need to
‘appreciate how central the ethical project is to human life’.123 The sceptic may
ask why he should be bound by the rules emerging from this project. He may
say: ‘There is no compelling motivation for me to continue in any ethical
tradition.’ Kitcher responds that the ‘desires the skeptic wants to satisfy by
breaking the rules have been made possible only by the project he rejects, that
he fails to understand how the origin and evolution of ethical practice have
framed his life, giving him the options he wants to pursue’. Kitcher knows that
this reply will not silence the sceptic, but he adds that no ethical theory will do
that successfully; his account does as good a job as any of the others.
There is a problem with this reply. Kitcher’s answer to the normative
question belongs in the same family as Greene’s (‘We can grasp the principles
behind nature’s machines and make them our own’), and it is liable to the
same objection. The ethical project is central to human life, as we observe it,
but so is self-preference. Our nature as evolved is a mixture. That is exactly
why we need ethics; we are beset by psychological–altruism failure on all sides.
We cannot, therefore, appeal just to the fact of this mixture in order to
discriminate between the options we should take and the options we should
resist.
The most important point may be that Kitcher thinks that the second
principle of his method, the one that rules out ‘any false beliefs about the
natural world’, means that any modification of ethical practice ‘invoking the
commands of an allegedly transcendent being’ would rightly be rejected, and is
excluded from the outset.124 Religious conviction, which is to say most
people’s conviction, does not even get into the conversation. Suppose we
imagine a reversal of roles. Suppose Kitcher is a theist belonging to one of
the Abrahamic faiths, and I am an atheist. Kitcher now thinks, as do very large
numbers of people in the world, that humans receive divine commands. This
means that he thinks atheism false, a false belief about the natural world so far
as the divine interacts with it, and one that cuts us off from great good. Does
this mean that he should now exclude me from the outset, because ‘a funda-
mental cognitive condition on mutual engagement is violated’? Surely what we
need are the conditions for settling disagreements on these central concerns,
or, if we cannot settle them, for proceeding without agreement. This is what
Kitcher calls ‘conditional mutual engagement’, where each party acknow-
ledges that, if the rival party were correct in the disputed matter, they would
have succeeded in mutual engagement.125 He thinks that conversation about

123 124
EP 273. EP 350, 358.
125
EP 357. There is a slightly more hospitable view in Life beyond Faith, esp. ch. 3, where
Kitcher develops a notion of ‘refined religion’ that does make the threshold for conditional
mutual engagement. But this is because refined religion has already subordinated doctrine to
ethical constraints.
300 God’s Command
what we ought to do even when we disagree about the facts can legitimately
proceed on this basis. But he thinks that ethical proposals based on religious
grounds do not even make the threshold for conditional mutual engagement.
His account of ethical method would be a great deal more plausible, and more
consistent with his overall pragmatism, if he allowed that religious disagree-
ments (such as that between him and me) could be consistent with conditional
mutual engagement in this way.126

8.5. TRANSCENDING OUR EV OLUTIONA RY

SI TUATI ON W ITH G OD

We can now go back to the story with which this chapter began. In the initial
telling of the story, no mention was made of God, and that was because of
the constraints of the translation manual. The story was treated as a Kant-
like translation from biblical theology into the language of contemporary
(non-theological) anthropology, though it is still a story and not science,
because of its drastic foreshortening into a single narrative. We can now go
back and put God back into the story, and doing so helps make sense of the
story. We can do this in three moments: the encounter, the command, and
the punishment.
First of all, God meets our ancestors, though they were probably not
monotheists. The story described this in terms of their emotions of awe and
joy, and now a bit more can be said about what kind of awe this might be. We
can use some distinctions here from Robert Roberts’s excellent book Emo-
tions.127 When we feel awe, we have a sense of something’s greatness, and this
requires some standard of comparison. Frans de Waal describes rain dances by
chimpanzees, and reports Jane Goodall’s description of a chimpanzee male
acting similarly near a roaring waterfall: ‘As he gets closer, and the roar of the
falling water gets louder, his pace quickens, his hair becomes fully erect, and
upon reaching the stream he may perform a magnificent display close to the

126
Richard Rorty, the eminent pragmatist, having complained about religion in the public
square as a ‘conversation stopper’, later came to characterize his remarks as ‘hasty and insuffi-
ciently thoughtful’. See ‘Religion in the Public Square: A Reconsideration’, Journal of Religious
Ethics, 31/1 (2003), 141–9. He conceded: ‘Both law and custom should leave [a believer] free to
say, in the public square, that his endorsement of redistributionist social legislation is a result of
his belief that God, in such passages as Psalm 72, has commanded that the cause of the poor
should be defended. . . . Attempts to find rules that are neutral between the two sides [of the
Kulturkampf] are pretty hopeless. So is the attempt to say that one or another contribution to
political discourse is illegitimate.’
127
Robert C. Roberts, Emotions (Cambridge: Cambridge University Press, 2003),
esp. 268–70.
 Divine Command and Evolutionary Psychology 301
foot of the falls.’128 She speculated that this might be a precursor of some
animistic religion. Perhaps the ape perceives the greatness of the waterfall,
relative to other streams he has perceived. There are, however, many kinds of
greatness. Kant distinguishes, for example, between the mathematical sublime
that responds to greatness in amount and the dynamic sublime that responds
to greatness in power. Both kinds of greatness can make everything else seem
small by comparison.129 It is probably impossible to specify a kind of greatness
that is the object of all kinds of awe. But in the case of our hunter-gatherer
ancestors it is plausible to think that what occasioned awe was something
personal. We are, after all, looking at agency detection. Awe directed at
persons is a specific kind of awe, and might best be called ‘reverence’. This
can attach to things that are not persons, but only because of their connection
with persons, in the way in which one might have reverence for the Torah
because it is God’s word in human language. It would not be enough to say in
ordinary English that one respects the Torah, and ‘respect’ may also not be
adequate as a translation of Kant’s Achtung, which is the feeling occasioned by
the moral law that we ‘recognize as God’s command’.130
Bringing in an encounter with God at this first moment explains how we
might arrive at the silencing or subordinating of self-interest. We can suppress
self-interest by not consciously thinking about it, but this is different, because it
can still be a prevailing motivation underneath (as discussed in Section 3.2.1).
This is different from self-interest being replaced as a prevailing motivation by
something else, the recognition of something so good that in comparison the
importance of any self-indexed good withers away. Section 3.2.2 invoked
Scotus’s distinction between the affection for advantage and the affection for
justice, which is a love of what is good in itself without any reference to the lover.
This does not mean that in the presence of what is good in itself we lose the
affection for advantage, but its salience can be radically decreased. This pro-
duces a double-source account of motivation. If our ancestors met God, perhaps
this produced what Kant calls ‘a revolution of the will’. To go back to the first
argument of Chapter 1, the encounter with divinity might have been with
something experienced as great, not merely terrifying but deeply attractive (in
Otto’s terms used in Chapter 1, fascinans as well as tremendum), something that
provided a sense of trust that relinquishing the priority of self-indexing was
consistent in a stable (though not necessary) way with their own happiness.
The second moment at which God enters the story is the command. This
command, in the story, is not connected in any intelligible way with nature. It

128
Frans de Waal, The Bonobo and the Atheist (New York: W. W. Norton & Company, 2014),
199–200. I had the experience in Zambia of seeing a group of baboons forming a perfect circle,
each grooming the one in front. Why should we deny that this is a ‘precursor’ of a religious
assembly?
129 130
KU v. 251, 260. Rel. vi. 154.
302 God’s Command
is an important part of the Genesis account that no attempt is made to say why
this kind of fruit should make its eaters wise. We are invited to think that
God selects within the divine prerogative (arbitrium) the fruit as a test, and the
test is to see whether the humans will try to usurp the divine function of
establishing what is good and bad, or (given the distinction in Chapter 1) what
is right and wrong.131 For present purposes, the significant feature of the
command is that it is not deducible from our nature or from any nature,
and it can therefore stand in for the whole series of divine commands that are
within God’s arbitrium in the same way. The version of the story at the
beginning of the chapter gave the decision not to eat the fruit a ‘groupish’
rationale as a symbol of group unity, though there is no warrant for this in the
Genesis text. Introducing God into the story allows us to say, like Rosenzweig,
that the basic command is not about the fruit, but is the command to love God
that comes out of the experience of being loved by God. Refraining from the
fruit is merely a symbol of that response. But, if we generalize to all the divine
commands for which we do not see the whole reason, we get some sense of
how introducing God into the picture might help from an explanatory point of
view. Evolutionary psychology is helpful in articulating our situation to which
the divine commands are addressed, and the need for group identification is
part of this situation. Religious communities sometimes arbitrarily select their
rituals; but it is also possible that sometimes they believe God selects them, and
this would help us understand why in those cases their apparent arbitrariness
or opacity does not undermine their perceived justification (to go back to the
argument from justification), which is that they express the love of God which
is good in itself.
The third moment is God’s punishment. In the Genesis story, after the
eating of the fruit, God asks ‘Where are you?’, because Adam and Eve have
hidden themselves.132 For present purposes, the importance of the punish-
ment is that it does not come with a divine withdrawal. There is in the Genesis
account the expulsion from the Garden, and the condemnation to wearisome
work, pain in childbirth, and distorted sexual relations. But God continues to
be present to the first humans and their descendants, even though the next
story is Cain and Abel and this is succeeded by the story of Lamech, and the
Flood, and Babel. The explanatory advantage here is that we need a way to

131
I am following Barth in this analysis of the command. See CD III/1. 257–8. As discussed in
Section 5.3.3, Barth distinguishes two kinds of knowledge of good and evil, in German wissen (for
God’s knowledge) and kennen (for appropriate human knowledge). God knows by way of
choosing and rejecting, and humans know by way of recognizing and endorsing. See McKenny,
The Analogy of Grace, 96–106.
132
Rosenzweig (Star, 175–6) suggests that God calls on Adam by name, and that Adam
recognizes the love in this second-person address, and after a series of third-personal excuses
(‘the woman did it’, etc.) responds with the same energy that had fuelled the previous defiance
but now with a pride in being loved.
 Divine Command and Evolutionary Psychology 303
understand the sense we have that, despite the cascade of evil that we experi-
ence (which in the theological story is our punishment), it would be possible to
be human and not be corrupted, that we are not as we are supposed to be, and
that there is nonetheless hope of becoming the way we are supposed to be.
This continuing hope would otherwise be mysterious, and we would expect a
moral system with a lower demand. The theistic version of the story tells us
that divine punishment does not exclude divine love, and that God intervenes
in our predicament to rescue us. The possibility of that redemption is already
implicit in the original encounter, but is made explicit in the form of covenant.
God goes on making initiatives towards us, and we go on refusing them.
Redemption returns us to the argument from grace in Chapter 1.
It is not surprising that the story fits the theistic explanation, because the
original version had God as a central character. But to the extent that the
translated version fits what actually happened to our ancestors, it is significant
if a theistic explanation is coherent and helpful. In none of these three cases is
the religious explanation of the story (in terms of encounter, command, and
punishment) the only one available. But evolutionary psychology, as we have
been looking at it in this chapter, gives us an excellent background against
which to see why bringing in God might give us a good explanation. There is a
fit between what we need and what God’s presence, guidance, and assistance
give to us.
We can go back through the discussions of evolutionary psychology we
have met, and see how our situation as evolved makes some independent
guidance helpful. In terms from Greene, we need something both to include
us, so that we can get beyond the ‘tragedy of the commons’, and to push us
beyond the group, so that we do not end up with mere within-group altruism.
The failures in psychological altruism that Kitcher posits as the origin of ethics
infect both our intra-group and our inter-group lives, and we can see the
preachments of the great religious traditions helping us with both.133 To take
first our lives within the group, we have seen earlier in this chapter the claim in
Spencer that the struggle for existence is a natural endowment from our
evolutionary history, and in Arnhart that this is true of our devotion to the
competitive goods such as wealth, power, and honour. We have seen Haidt’s
claim that because of our evolutionary background we care more about
reputation than about truth or sincerity, and that our reasoning is often better
seen as an ‘inner lawyer’ managing this reputation than an ‘inner scientist’
trying to work out what is right to do. We have seen Greene’s claim that from

133
I have not tried to deal with the controversy between the authors discussed in this chapter
about whether religious commitment has a genetic base (perhaps produced by group selection)
as opposed to a cultural base, or what the mixture looks like, if it is a mixture. See Bloom,
‘Religious Belief as an Evolutionary Accident’, 119, and Haidt, RM 251–2 (arguing for a fast pace
at which genetic evolution could work).
304 God’s Command
an evolutionary perspective our reasoning systems are designed for selecting
rewarding behaviours. We do not have to accept all these claims in order to
conclude that even within the group our ability to care for others is fragile.
This is not the focus of the works we have looked at in this chapter, but that is
not because these authors deny this point, but because they want to stress the
failures between groups. To the list of our tendencies to failure within the
group should be added unrighteous anger (in addition to the righteous anger
that makes us good punishers of defectors), importunate lust (carrying us
away from the people our loyalty commits us to), and craven fear (preventing
us from standing up to powerful violators of each other’s deepest well-being).
Many of these tendencies belong to our internal mental lives before they
manifest themselves in what we do or do not do outwardly. To make a list
like this of our characteristic failures is not just ‘Calvinist Sociobiology’ in de
Waal’s misleading phrase, because it is consistent with saying that we also have
tendencies to the good, ‘better angels’ of our nature, so that we end up a
mixture. But we need something other than just an appeal to our nature to get
us to follow the parts of the mixture that we should follow and not the parts we
should not.
Now consider the preachments of the traditions. Psalm 19 first imagines the
sun coming forth as a bridegroom out of his chamber, and circling through the
heavens, and there is nothing hidden from its heat. C. S. Lewis calls it
‘luminous, severe, disinfectant, exultant’.134 Then the psalmist thinks of the
law of the Lord in the same terms, that it gives us light and cleanses us, and
the psalmist rejoices in it, more than in gold or in honey. He does not specify
the contents of the law, because he assumes it known, but he tells us to love it.
Finally, he asks to be cleansed from his secret faults, and the presumptuous
sins that threaten to have dominion over him.
Jesus’s Sermon on the Mount is full of commands that go inside the mind. It
upholds the value of every jot and tittle of the law, but says that the prohibition
on killing extends to unjustified anger, and the prohibition on adultery extends
to lust. He commends giving alms and praying and fasting in secret, and not
just in public (where reputation is at stake). He commands not serving the god
of wealth, and not taking thought for material goods (what we will eat or drink
or wear), but emulating the lilies of the field, for even Solomon in all his glory
was not arrayed like one of these.
The Qur’an says (2: 235): ‘Know that God knows what is in your hearts.
So fear Him and know that God is All-Forgiving, All-Forbearing.’ It con-
tains versions of all the Ten Commandments except for the commandment
about the Sabbath.135 While it prescribes rituals—for example, facing the

134
C. S. Lewis, Reflections on the Psalms (New York: Harcourt Brace Jovanovich, 1958), 64.
135
See Miroslav Volf, Allah: A Christian Response (New York: Harper Collins, 2011), 106–7.
 Divine Command and Evolutionary Psychology 305

Sacred Mosque—it says (2: 177) that virtue is not to turn your faces
eastwards or westwards, but ‘to dispense money, though dear [or “out of
love from Him”], to kinsmen, orphans, the needy, the traveller, and beggars,
and for ransom [of slaves]’. We will consider benevolence again when we get
to inter-group preachments. For now, it is worth pointing out that the
Qur’an says that these gifts to the poor and needy are best done secretly,
and thus independently of reputation: ‘If you make public your gifts, freely
given, that would be worthy. But if you conceal them and donate them to the
poor, that would be even better for you. God will thereby pardon your sins’
(2: 271).
In all these ways, and there are many others, the resources of religious
traditions have responded to the problems within groups posed by our
evolutionary heritage. The same is true of the second class of psychological–
altruism failures, the failures between groups. Most of the authors we have
been looking at emphasize this kind of failure, but most of them either do not
acknowledge the role of religion in overcoming it (for example, Haidt, who
characterizes religion as a ‘hive switch’) or, if they do acknowledge that the
major religions do try to do this, they think it is utopian and does not fit our
evolved nature (for example, Arnhart and Mackie). For Greene, the tragedy of
our between-group hostility can be overcome by utilitarianism, but he cuts this
school off from its theological roots and the common ground they provide.
Kitcher acknowledges the roots, but is committed by his atheism to a rejection
of any use of them, even any ‘conditional mutual engagement’ in his sense
with these traditions.
Again we can look at the preachments of the Abrahamic traditions as fitting
our predicament very well. I am not sufficiently learned to go beyond this
boundary, though I suspect that the great religions of the East could be looked
at in the same way. Greene quotes the episode related in the Talmud, where
Rabbi Hillel says that the whole Torah consists in the prescription: ‘That which
is hateful to you, do not do to your neighbour.’ But who is the neighbour? This is
the question that Jesus answers with the parable of the Good Samaritan. But it is
also a question that we examined in the context of Judaism in Chapter 7.
According to David Novak, the Noahide laws are internal to Judaism—for
example, governing the relations of the people of Israel to resident aliens—but
also external, being given (at least the first six of them) to Adam. Chapter 7 did
not accept Novak’s conclusion about the order of justification, that the Torah is
justified by its conformity to natural law. But it did accept that universal human
obligations are internal to the authoritative Jewish texts; ‘universal’ here applies
both to the scope of those who are obliged and to the scope of those who are to
be the beneficiaries.
The Sermon on the Mount likewise contains a version of the Golden Rule
(Matthew 7: 12), and it also contains the command to love our enemies
(Matthew 5: 43–8):
306 God’s Command
You have heard that it was said, ‘You shall love your neighbour and hate your
enemy.’ But I say to you, Love your enemies and pray for those who persecute
you, so that you may be children of your Father in heaven; for he makes his sun
rise on the evil and on the good, and sends rain on the righteous and on the
unrighteous. For if you love those who love you, what reward do you have? Do
not even the tax collectors do the same? And if you greet only your brothers and
sisters, what more are you doing than others? Do not even the Gentiles do the
same? Be perfect, therefore, as your heavenly Father is perfect.
Jesus gives the parable of the Good Samaritan in Luke 10: 25–37 to the lawyer
who asked him ‘who is my neighbour?’, to illustrate that the Samaritan, for
whom Jews were enemies, nonetheless was the one who ‘neighboured’ the
wounded man lying by the side of the road. He ends the story: ‘Go and do
likewise.’
In Chapter 6 we discussed the Mu‘tazilite position that injustice, which is a
wrong injury done to another person, is to be avoided regardless of benefit or
harm to oneself. Even a heartless man, says ‘Abd al-Jabbar, would warn a blind
stranger against falling into a well. He recognizes that his opponents will
suppose that the warning is just to prevent the warner’s own distress. But he
insists that it is possible that all (extrinsic) reasons are removed, and the
heartless man still guides the wanderer, and this must be for the rightness in
itself of this action (since it cannot be for no reason at all). The obligation to an
unknown blind traveller takes us outside intra-group relations. The Qur’an
does not contain the Golden Rule, but the Hadith (the authoritative reports of
the prophet’s sayings) include the saying: ‘None of you has faith until you love
for your neighbour what you love for yourself.’136 The Qur’an also does not
include a command to love one’s enemies, and here there is no supplementary
hadith. But there are commendations of extraordinary generosity even to
those who have wronged you.137
All of these commands take the adherents of the Abrahamic faiths towards
a universal morality. These faiths both include their adherents into a commu-
nity, and then push them beyond it. But this does not yet show that universal
morality is not utopian (as Arnhart and Mackie, for example, suppose it is).
One of my earlier books, The Moral Gap, described the system of morality in
terms of a three-part structure: a moral demand; our mixed natural capacities,
which are inadequate to the demand; and God’s assistance, which intervenes
in our situation to make it possible for us to live by the demand. The present
book has dealt mainly with the first part of this structure, the divine com-
mands, and the second, our natural capacities. Chapter 1 discussed the argu-
ment from grace, which posits divine assistance. But this is not the right place
to launch into an account of what this assistance is like. The Moral Gap dealt

136
Sahih Muslim, Kitab al-Iman, 67–1, Hadith no. 45, cited in Volf, Allah, 159.
137
Volf, Allah, 178.
 Divine Command and Evolutionary Psychology 307
theologically with the work of the second person of the Trinity, and discussed
the doctrines of justification and atonement. We need an account of how to
deal with our own moral failure. It does not seem that we are in a good
position after such failure to forgive ourselves for it. From a Christian point of
view, God provides a remedy in Christ’s atonement. Any Christian account of
law needs to be supplemented by an account of gospel. But explaining that is
not a task for this book.
We can end this section of the chapter with a postscript. Does the picture of
divine command, mixed natural capacity, and divine assistance actually work
to produce morally better lives in those who accept it? There is some empirical
evidence that the answer is ‘Yes’. There are a number of studies in the
literature of social psychology, which are reported by Haidt in The Righteous
Mind. Richard Sosis studied 200 communes founded in the nineteenth century
in the United States.138 He compared those organized around religious prin-
ciples and those organized around secular principles. After twenty years 6 per
cent of the secular communes were still functioning, and 39 per cent of the
religious communes. Shared religious life binds people together. Stephen
Lansing reported on the Balinese rice farmers who had to rely on fair water
distribution from streams flowing down from a volcano.139 They built a large
temple at the top of the whole system of water channels, staffed with twenty-
four priests, dedicated to the worship of the Goddess of the Waters. The
common religion made possible the required trust. So far we have what
Haidt calls a ‘hive switch’, creating group solidarity and loyalty. But we can
go beyond this. Robert Putnam and David Campbell compared how religious
and non-religious Americans behave in terms of giving money and time to
charities and social organizations.140 The religious Americans gave more
money not just to religious organizations but to the American Cancer Society,
and they volunteered not just in church and synagogue and mosque but in
civic associations across the board. Putnam and Campbell conclude: ‘By many
different measures religiously observant Americans are better neighbors and
better citizens than secular Americans—they are more generous with their
time and money, especially in helping the needy, and they are more active
in community life.’141 When we look at the great movements towards the

138
Richard Sosis, ‘Religion and Intragroup Cooperation: Preliminary Results of a Compara-
tive Analysis of Utopian Communities’, Cross-Cultural Research, 34 (2000), 264–74.
139
J. S. Lansing, Priests and Programmers: Technologies of Power in the Engineered Land-
scapes of Bali (Princeton: Princeton University Press, 1991).
140
R. D. Putnam and D. E. Campbell, American Grace: How Religion Divides and Unites Us
(New York: Simon and Schuster, 2010).
141
Ibid. 461, quoted in RM 310. Note that it is observance that is measured, not belief. It is
true that Danes and Swedes, for example, score well on some of these measures of ‘niceness’ and
are less religiously observant than Americans (though still identifying themselves as Christian).
But there may be all sorts of reasons for this other than degrees of religious observance. It is
308 God’s Command
recognition of human value over the last sixty years, we will often find a
religious motivation. I am thinking of Martin Luther King and the civil-rights
movement, and the Lutherans in East Germany and the fall of the totalitarian
state. Why is this? It is because of the nature of the God they worship. It is true
that belonging to a community is very important, but the God of Abraham not
only includes us in community but pushes us out beyond community, to meet
the needs of the poor and the marginalized who are the objects of God’s care
just as much as we are. God commands both the inclusion and the moving-
out. And these do not need to be competing goals (though particular religious
assemblies can be better at one than the other). In Section 2.2, we discussed the
principle of providential proximity, according to which God puts us next to
the people God wants us to help. If we live in America, and one of our
community goes to a village in Zambia, that village becomes our village. The
point is that God wants to bless the whole of creation, and wants to bless it
through us. God’s command is a vehicle of this blessing.

better to stick within a single culture (albeit one with different subcultures) to make the
comparison.
 9

Summary

This final chapter is a brief summary of the theory of this book as a whole. The
book has been constructed through a series of dialogues with different figures
from the past or the present of the discipline. In this kind of series it is easy to
lose, as they say, the forest for the trees. Those with a systematic rather than a
historical bent find the constant reference to other people’s views tiresome,
and they want to be able to examine the theory on its own terms. As I said in
the Preface, I do not think we can escape the interlocutors even if we want to.
We are always in fact working through ideas that have been handed on to us,
and it is helpful and appropriate to acknowledge the sources when we can. It is
helpful, because we can then make clear the background in the surrounding
views of the author, and where we disagree with those views. Nevertheless,
I will end the book with an outline of the main points of the theory, referring
back just by section number to the more detailed and historically grounded
discussions in the rest of the book. There is a risk here that the theory will
seem crude and pretentious in this bare outline, without the nuance provided
in the main text. But the risk is worth taking. I will not give this outline of the
theory in the same order in which the theory was presented in the main text.
The logical order is not always the same as the order of initial intelligibility to
us. My hope is that, while there will not be any new ideas in this summary, the
connection of the ideas already present in the main text will become clearer.
The book is designed to defend the thesis that what makes something
morally obligatory for us is that God commands it, and what makes something
morally wrong for us is that God prohibits it. There are four terms in this
thesis: God, we human beings, the commanding (or prohibiting), and the
moral law (which states what is obligatory and what is wrong or forbidden).
I will start by saying more about each of these four terms.
I am myself a Christian. But I am assuming in this book (without arguing
for it) that God is the same God in Judaism, Christianity, and Islam, even
though there are fundamental differences between these religions in what is
said about God. The doctrine of the Trinity in Christianity, to take one central
example, is denied in Judaism and Islam. But most Christians will not take this
to mean that Judaism has a different God. They should not take it to mean that
310 God’s Command

Islam has a different God either. The fact that divine command is so central to
all three, and that so many of the same problems arise in all three about the
relation between divine command and human reason, should be taken as
confirmation. God is taken in this book to be the supreme good. This goodness
manifests itself in three ways (Section 7.3). First, God is the creator of all that
exists other than God, and maintains it and is present to it once created.
Second, God gives us revelation, and for the purposes of this book the primary
revelation is of the divine will for our willing, which God gives us in command.
Finally, God redeems us, by bringing us to that union with God that is our
proper end (Section 1.2). These three functions (creation, revelation, and
redemption) can be expressed in terms of a threefold sovereign role that
God has over the created order, by analogy with human sovereignty. God
has legislative, executive, and judicial functions (Section 2.2): God makes and
promulgates the law by command; God runs the universe and sustains its
order; and God judges us and punishes and saves us.
Human beings are created as rational animals through the processes of
evolution (Section 8.1). We have the purpose of a kind of loving union with
God that is available only to rational animals (Section 4.1.1). Whether we are
the only kind of rational animal we do not know (Section 8.5). Each of us has,
however, not merely the purpose common to the whole species, but a
particular purpose (unique to the individual) of a kind of love of God
particular to that individual (Sections 1.4, 5.1.2). Our destination is a realm
in which all these individual kinds of love are conjoined. We all have the
same basic value because we all have a call from God of this unique kind
(Section 1.5). From the fact that we are commanded or called in this way it
follows that we are individual centres of agency, in time, free, and language-
users (Section 2.3.1). These features put constraints on what we should take
to be a divine command. From these same constraints, we can deduce a
presumption against taking anything to be a divine command that requires
breaching these constraints (Section 4.1.2). For example, we should have a
presumption against supposing that God is telling us to take an innocent life
or to lie. We are, however, born with not only a predisposition to respond to
the command, but a propensity to put our own happiness above the com-
mand (Section 1.2). We are in that way a mixture (Section 8.4.1). But the
predisposition is essential to us, and the propensity is not.
How does the propensity fit with the feature of human beings already
mentioned, that humans are free beings, and how does this feature of
ours fit with God’s sovereignty in the divine executive role (Sections
6.2.1–2)? We need to distinguish two kinds of power we have and two kinds
of divine decree (Section 6.2.3). We have the power to choose whether to
accept the divine command (which is the first kind of divine decree) or reject
it. But the power to act (the second power) is itself made possible only by
God’s sustaining power (the second decree), and God in the second decree
 Summary 311

brings all things (even our rejecting the first decree) to good. The relation
between our freedom and God’s power is that we are like a lake and God’s
power is like the flow in that lake from a hidden spring (Section 5.2.1). God’s
loving us has the power (though not the irresistible power) to call out our love
in response (Section 7.3.3).
Moral obligation can be both universal and particular. It is universal when it
has all human beings in the scope of the subjects who are commanded to act
and the scope of the beneficiaries or victims of that action (Sections 1.1 and
1.4). The scope extends beyond human beings to non-human animals as
beneficiaries and victims, but the book has not argued for that. We are
required to be able to will the prescription of an action together with
the reason for it as a universal law. We also are required to treat those who
are affected by our actions (the beneficiaries and victims) as valuable in
themselves, and this means making their purposes our purposes as far as the
moral law allows (Section 5.1.3). One way to understand this takes us to the
sense in which moral obligation can be particular. It is useful here to distin-
guish between the different positions in a moral judgement, especially between
the positions of addressee, agent, recipient, and action (Sections 2.1.1 and
5.1.3). Sometimes a moral judgement can make another person’s purpose the
agent’s own purpose by willing the prescription with a universal term in the
action position, but not in the other positions in the judgement. However,
there will usually also be, in such a situation, a moral judgement that applies
that is universal in all its term positions. It is a mistake to think that every
judgement that is essentially linked to an individual recipient (and so the term
in the recipient position is not universalizable) is self-indexed to the agent in a
way that is morally problematic (Section 3.2.3). The term in the action
position of a moral judgement, however, does have to be universal or univer-
salizable, because in prescribing the action morally we claim that it meets our
criteria for goodness and these criteria give types and not individuals.
Commands are a species of prescription, and we can distinguish five types
of divine prescriptions: precepts, prohibitions, permissions, counsels, and
directly effective commands (Sections 2.1.2–6). Divine precepts, prohibitions,
and permissions all have, by their meaning, internal reference to divine
authority. However, only precepts and prohibitions create obligation, and
have, by their meaning, internal reference to condemnation or punishment
for failure. Counsels do not create obligation (Section 2.1.5). Directly effective
commands are not a form of communication, and do not make something
obligatory (Section 2.1.6). ‘Authority’ is to be distinguished into objective
authority and subjective authority (Section 2.2). God has objective authority
over all human beings, whether they recognize it or not, because God’s
commands give all human beings rightful reason to comply, given
God’s threefold sovereign role already described. The reasons are rightful
because God’s commands make obligatory the good things that God
312 God’s Command

prescribes, all of which take us to our proper end by the path God has selected
for us, and our obedience is an expression of our love for God, which is good
in itself and our end.
So far we have a brief account of the four terms involved in the thesis that
what makes something morally obligatory for us is that God commands it, and
what makes something morally wrong for us is that God prohibits it. There are
at least five objections to this thesis. One objection is that it produces an infinite
regress (Section 1.4). Suppose I ask: ‘Why should I take the moral demand as a
demand upon me?’ A divine command theorist answers: ‘Because God com-
mands it.’ But suppose I then ask: ‘Why should I do what God commands?’ It
can seem that the divine command theorist has to give some further answer:
‘because you owe loyalty to your creator’, or ‘because you are God’s property’, or
something similar. But then why should I be loyal to my creator, or obey my
owner? The regress looms. But there is an answer to this worry. The principle
that God is to be loved is known from its terms: we know that if something is
God, (since God is supremely good) it is to be loved (Section 1.3). But to love
God is to obey God (to repeat in our wills God’s will for our willing), and so we
know from its terms the principle that God is to be obeyed.
A second objection is that the thesis makes morality arbitrary. Could not
just anything (for example, torturing babies) be obligatory if God were to
command it? The solution to this worry is to see that there is a distinction
between the good and the obligatory, and that God relates differently to them
(Section 1.3). The thesis of the book is that God’s command makes something
obligatory. God selects which good things to command; but it is consistent
with this to say that God selects only good things, and so the good puts a
constraint on what God commands. When a person judges that a thing is
good, she expresses an attraction to it and says that it deserves to attract her.
There is a prescriptivist or expressivist side to this and a realist side
(Section 4.3.1–2). The prescriptivist side is that the evaluative judgement
expresses some state of desire or emotion or will (she is attracted). The realist
side is that there is some value property that she claims belongs to the thing, in
virtue of which her state of desire or emotion or will is appropriate (the thing
deserves to attract her). Realism of this kind is not undercut by accepting our
evolution (Sections 8.3.1–3). The goodness (the value property) might reside
in resemblance to God. It might also reside in the union with God that is the
human destination, or what leads to this union, or what manifests God by
displaying God’s presence (Section 4.1.2). If God is supremely good, union
with God must also be good as an end, and so must the path to this end be
good as a means. God commands only what is consistent with this destination,
and thus the command is not arbitrary in the contemporary sense, in which
what is arbitrary ignores some consideration that is relevant to a decision
(Section 1.3). But this takes us to the third objection. If God commands only
what is good, is God’s command redundant?
 Summary 313

The third objection, from the redundancy of the divine command, can be
met by again making a distinction. The moral law cannot be deduced from our
nature, but it fits our nature exceedingly well. There are two kinds of deduc-
tion we should deny. It might be thought that we could fix the reference of
‘good’ by looking at what most people, most of the time, think is good
(Section 4.2). But this does not fit the fact that we could be, and in fact are,
wrong much of the time in our evaluation. An examination of ancient Greek
ethics and its stress on the competitive goods illustrates this. The second kind
of deduction we should avoid is the deduction of virtue from our human form
of life (Section 4.4.3), even though there is a goodness of organisms that can be
deduced from their simply being alive (Section 4.4.1). The human form of life
does indeed put a constraint on what we should conceive our virtues to be, but
a large part of our conception of virtue is constituted by our ideals
(Section 4.3.3), and these cannot be deduced from our form of life, unless
we have already screened our description of this form of life through our
ideals. The central reason for the failure of this deduction is the mixture in
both our natural inclinations and our ideals between what deserves to attract
us in this way and what does not so deserve (Sections 6.3.3 and 8.5). The
danger of some kinds of natural law theory is that God disappears into
creation, in the sense that, because we think we can get morality from our
nature, we think we do not need a personal divine commander (Section 7.3.2).
But creation itself, including our created nature, is not yet sufficiently com-
plete for us to deduce from it how we should live (Section 7.3.4). If we think of
reason as what looks at our nature, we can say that it is to some degree useful
in determining our duties, as a junior partner, and it is indispensible in
disputes between traditions, but its results are not sufficiently determinate to
tell us how to live, and we need the revelation of divine command in addition
(Sections 6.1.3, 6.3.3, and 7.1). For this revelation we should, therefore, be very
grateful.
A fourth objection is that we live in a pluralist society, and appealing to
God’s commands is inappropriate for conduct in the public square in such a
society (Section 7.2). On this view there is nothing wrong with appealing to
divine command within one’s own religious community, and this community
may even be one’s primary community; but when a democratic society is
making policy together, and we do not all believe in the same God or even in
any god, it is inappropriate to ground our prescriptions for the common life in
the command of a being that we do not all worship. Such grounding stops
conversation (Section 8.4.2). The reply to this objection is twofold. First, it is
discriminatory against religious believers to require them to shed their most
basic commitments in public dialogue (Sections 6.3.1 and 8.4.2). Second, there
is not enough common ground between all the parties to public conversation
so that we could get good policy by sticking to the lowest common denomin-
ator (Section 7.2.1).
314 God’s Command

A fifth objection is that, even if God were to give us commands, we are too
unreliable as receivers of them to make them the final arbiters of our moral
decisions. Too many insane and bad people have appealed to divine command
in justifying their actions. The question here is about what kind of access we
have to the commands, and how we know when we are receiving a command,
and when we are just making it up (Section 7.1.3). One way to proceed here is
to work out a rational ethical decision procedure and then say simply that God
commands us to follow it (Section 5.3.2). But the Abrahamic faiths have
additional resources in the content of the narratives they give us of God’s
dealing with human beings, in the procedures they prescribe for checking with
other members of the community, and in the phenomenology they describe as
characteristic of the reception of divine command (Section 5.3.3). They can
say that direct divine commands present themselves with clarity and distinct-
ness, external origination, familiarity, authority, and providential care.
Finally, we should deny another thesis found in some forms of natural law
theory, the thesis of eudaemonism that we should choose everything for the
sake of happiness. We need instead a dual structure of motivation, according
to which happiness is properly one of our ends, but we are also to be moved by
what is good in itself independently of our happiness (Section 3.2.2). The
notion of happiness is not just pleasure (Section 3.2.3). It includes an ideal
element, so that we would not count a person in a pleasure-machine as ‘really’
happy (Section 4.3.3). But it is self-indexed, in the sense that the agent pursues
it as her own good, and this makes eudaemonism unacceptably self-regarding.
Various defences of eudaemonism should be rejected. One defence is that
happiness includes sympathetic pleasures, and thus takes an agent outside
herself (Section 3.3.1). But sympathetic pleasures are limited in a way that
morality should want to transcend (Sections 8.2.2 and 8.4.1). A second defence
is that reason brings impartiality with it, and so our good as rational beings
requires that we follow the moral law (Section 3.3.2). But the notion of reason
here simply begs the question. A third defence is to propose that the interests
of the whole of creation form a nested hierarchy, so that, if the agent correctly
sees this order, she will see that her good is necessarily consistent with the
good of the whole (Section 3.3.3). But it is not hard to think of cases of real
conflict, or at least possible conflict, between interests, in which case the
question arises of whether any self-indexed good should take the priority. It
is not necessary that God should will to save all rational animals. Finally, we
can revise the third defence so that the agent perfects herself by identifying
with God who is self-transcending (Section 3.3.4). But, if she thereby loses
attachment to self-indexed goods, this revision becomes unacceptably self-
neglecting. We need a dual structure of motivation. We should hold that
happiness and morality are indeed conjoined, but not because of some neces-
sity in the nature of happiness or in the nature of morality, but because of the
free benevolence of the supersensible author of nature (Section 1.1).
 Bibliography

‘Abd al-Jabbar, al-Mughni fi abwab at-tawhid wa l-adl, ed. I. Madkour et al. (Cairo,
1962–) (Mughni).
Adams, Robert M., Finite and Infinite Goods: A Framework for Ethics (Oxford: Oxford
University Press, 1999).
Alston, William, ‘Religious Experience and Religious Belief ’, Nous, 16 (1982), 3–12.
Annas, Julia, The Morality of Happiness (Oxford: Oxford University Press, 1993)
(MH).
Anscombe, G. E. M., Intention (Oxford: Blackwell, 1957).
Anscombe, G. E. M., ‘Modern Moral Philosophy’, Philosophy, 33 (1958), 26–42.
Aquinas, Thomas, Summa Theologiae, trans. English Dominicans (London: Burns,
Oates, and Washbourne, 1912–36; repr. New York: Christian Classics, 1981).
Aristotle, The Complete Works of Aristotle, ed. Jonathan Barnes (Princeton: Princeton
University Press, 1984).
Aristotle, Aristotle XX: Athenian Constitution, Eudemian Ethics, Virtues and Vices,
trans. H. Rackham (Cambridge, MA: Harvard University Press, 1986).
Aristotle, Aristotle: Nicomachean Ethics, Books VIII and IX, ed. and trans. Michael
Pakaluk (Oxford: Clarendon Press, 1998).
Aristotle, Nicomachean Ethics, ed. and trans. Roger Crisp (Cambridge: Cambridge
University Press, 2000).
Arner, Neil, ‘Theological Voluntarism and the Natural Law: The Integrated Moral
Theories of John Duns Scotus, John Calvin, and Samuel Pufendorf ’,
Ph.D. dissertation, Yale University, 2012.
Arnhart, Larry, Darwinian Natural Right: The Biological Ethics of Human Nature
(Albany, NY: SUNY Press, 1998).
Arnhart, Larry, Darwinian Conservatism (Exeter: Imprint Academic Press, 2005).
al-Ash’ari, Abu al-Hasan, Elucidation of Islam’s Foundation (al-Ibanah), American
Oriental Series, vol. 19, trans. Walter C. Klein (New Haven: American Oriental
Society, 1940).
al-Ash’ari, Abu al-Hasan, The Theology of al-Ash‘ari, trans. Richard J. McCarthy, SJ
(Beyrouth: Imprimerie Catholique, 1953) (Kitab).
al-Attar, Mariam, Islamic Ethics: Divine Command in Arabo-Islamic Thought (Abingdon:
Routledge, 2010).
Augustine, On Free Choice of the Will, trans. Thomas Williams (Indianapolis: Hackett,
1993).
Augustine, The City of God against the Pagans, trans. R. W. Dyson (New York:
Cambridge University Press, 1998).
Austin, J. L., How to Do Things with Words, ed. J. O. Urmson (Oxford: Oxford
University Press, 1965).
Baggett, David, and Walls, Jerry L., Good God: The Theistic Foundations of Morality
(Oxford: Oxford University Press, 2011).
316 Bibliography
Barrett, Justin, ‘Cognitive Science, Religion, and Theology’, in Jeffrey Schloss and
Michael Murray (eds), The Believing Primate (Oxford: Oxford University Press,
2009), 76–99.
Barth, Karl, The Epistle to the Romans, trans. Edwyn C. Hoskyns (Oxford: Oxford
University Press, 1968).
Barth, Karl, The Christian Life: Church Dogmatics IV/4; Lecture Fragments, trans.
Geoffrey W. Bromiley (Edinburgh: T & T Clark, 1981).
Barth, Karl, Ethics, ed. Dietrich Braun, trans. Geoffrey W. Bromiley (Edinburgh: T & T
Clark, 1981).
Barth, Karl, The Holy Spirit and the Christian Life: The Theological Basis of Ethics,
trans. R. Birch Hoyle (Louisville, KY: Westminster/John Knox, 1993).
Barth, Karl, Protestant Theology in the Nineteenth Century: Its Background and
History, trans. Brian Cozens and John Bowden (Grand Rapids, MI: Eerdmans,
2002).
Barth, Karl, Church Dogmatics, ed. and trans. G. W. Bromiley and T. F. Torrance,
14 vols (London: T & T Clark, 2009) (CD).
Bergmann, Michael, Murray, Michael J., and Rea, Michael C. (eds), Divine Evil?
The Moral Character of the God of Abraham (Oxford: Oxford University Press,
2011).
Biggar, Nigel, The Hastening that Waits: Karl Barth’s Ethics (Oxford: Clarendon Press,
1993).
Blackburn, Simon, ‘Errors and the Phenomenology of Value’, in Essays in Quasi-
Realism (Oxford: Oxford University Press, 1993).
Bloom, Paul, ‘Religious Belief as an Evolutionary Accident’, in Jeffrey Schloss and
Michael J. Murray (eds), The Believing Primate (Oxford: Oxford University Press,
2009), 118–27.
Blum, Lawrence, Friendship, Altruism, and Morality (London: Routledge & Kegan
Paul, 1980).
Blum, Lawrence, ‘Moral Perception and Particularity’, Ethics, 101/4 (1991), 701–25.
Boyd, Richard N., ‘How to Be a Moral Realist’, in Geoffrey Sayre-McCord (ed.), Essays
on Moral Realism (Ithaca, NY: Cornell University Press, 1998).
Boyer, Pascal, The Naturalness of Religious Ideas: A Cognitive Theory of Religion
(Berkeley and Los Angeles: University of California Press, 1994).
Brink, David O., Moral Realism and the Foundations of Ethics (New York: Cambridge
University Press, 1989).
Brittain, C., ‘Cicero’s Sceptical Methods: The Example of the De Finibus’, presented
at the 12th Symposium Hellenisticum, ‘Cicero’s De Finibus: A New Appraisal’,
September 2010.
Broad, C. D., Five Types of Ethical Theory (London: Kegan Paul, 1930).
Broadie, Sarah, Ethics with Aristotle (Oxford: Oxford University Press, 1991).
Buber, Martin, I and Thou (1923), trans. Walter Kaufmann (New York: Scribner,
1970).
Butler, Joseph, Fifteen Sermons (Charlottesville, VA: Ibis Publishing, 1987).
Calvin, John, Institutes of the Christian Religion, ed. John T. McNeill, trans. Ford Lewis
Battles, (Philadelphia: Westminster Press, 1960).
Campbell, Donald, ‘Altruism: Biology, Culture and Religion’, Journal of Social and
Clinical Psychology, 3/1 (1985), 33–42.
 Bibliography 317
Caputo, John D., ‘How to Avoid Speaking of God: The Violence of Natural Theology,’
in Eugene Thomas Long (ed.), Prospects for Natural Theology (Washington: Cath-
olic University of America Press, 1992).
Carson, Thomas, Value and the Good Life (Notre Dame, IN: University of Notre Dame
Press, 2000).
Cartwright, John, Evolution and Human Behavior (Cambridge, MA: MIT Press, 2000).
Cerić, Mustafa, Roots of Synthetic Theology in Islam: A Study in the Theology of Abu
Mansur al-Maturidi (d. 333/944) (Kuala Lumpur: International Institute of Islamic
Thought and Civilization, 1995) (Roots).
Chignell, Andrew, ‘Belief in Kant’, Philosophical Review, 116 (2007), 323–60.
Chignell, Andrew, ‘Rational Hope, Moral Order, and the Revolution of the Will’, in
Nancy Cartwright and Eric Watkins (eds), Divine Order, Human Order, and the
Order of Nature (Oxford: Oxford University Press, 2013).
Cicero, On Moral Ends, ed. Julia Annas, trans. Raphael Woolf (Cambridge: Cambridge
University Press, 2001).
Clark, Gordon H., Karl Barth’s Theological Method (Nutley, NJ: Presbyterian and
Reformed Pub. Co., 1963).
Clough, David, Ethics in Crisis: Interpreting Barth’s Ethics (Aldershot: Ashgate, 2005).
Cohen, Hermann, Charakhteristik der Ethik Maimunis, in Jüdische Schriften, iii
(Berlin: B. Strauss, 1924).
Cohen, Hermann, Religion of Reason out of the Sources of Judaism, trans. S. Kaplan
(New York: Ungar, 1972).
Copp, David, ‘Milk, Honey, and the Good Life on Moral Twin Earth’, Synthese, 124
(2000), 113–37.
Copp, David, ‘Realist-Expressivism: A Neglected Option for Moral Realism’, Social
Philosophy and Policy, 18/2 (2001), 1–43.
Cronquist, John, ‘Hare and Prescriptivism’, dissertation, Stanford University, 1972.
Cross, Richard, Duns Scotus (Oxford: Oxford University Press, 1990).
Cudworth, Ralph, A Treatise Concerning Eternal and Immutable Morality, ed. Sarah
Hutton, (Cambridge: Cambridge University Press, 1996).
Darwall, Stephen, The Second-Person Standpoint: Morality, Respect, and Accountabil-
ity (Cambridge, MA: Harvard University Press, 2009).
Davidson, Herbert, Alfarabi, Avicenna, and Averroes, on Intellect (Oxford: Oxford
University Press, 1992).
Derrida, Jacques, Dissemination, trans. Barbara Johnson (Chicago: University of
Chicago Press, 1981).
Descartes, René, Meditations on First Philosophy, trans. Donald A. Cress (Indianapolis:
Hackett, 1979).
Dodsworth, Christopher, ‘Understanding Divine Authority’, Faith and Philosophy,
28/2 (2011), 190–208.
Douglas, Mary, Purity and Danger: An Analysis of the Concepts of Pollution and Taboo
(London: Routledge & Kegan Paul, 1966).
Driver, Julia, ‘Gertrude Elizabeth Margaret Anscombe’, in Edward N. Zalta (ed.),
Stanford Encyclopedia of Philosophy <http://plato.stanford.edu/archives/win2011/
entries/anscombe/> (accessed 15 September 2014).
Drury, John, Music at Midnight: The Life and Poetry of George Herbert (Chicago:
University of Chicago Press, 2014).
318 Bibliography
Edwards, Jonathan, Religious Affections, ed. John E. Smith (New Haven: Yale Univer-
sity Press, 1959).
Emon, Anver, Levering, Matthew, and Novak, David, Natural Law: A Jewish, Chris-
tian, and Muslim Trialogue (Oxford: Oxford University Press, 2014) (Trialogue).
Engberg-Pedersen, Troels, Aristotle’s Theory of Moral Insight (Oxford: Clarendon
Press, 1983).
Evans, C. Stephen, God and Moral Obligation (Oxford: Oxford University Press, 2013).
Ewing, A. C., The Definition of Good (New York: Macmillan, 1947).
Fichte, Johann Gottlieb, Foundations of Natural Right, ed. Frederick Neuhouser, trans.
Michael Bauer (Cambridge: Cambridge University Press, 2000).
Finnis, John, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980).
Finnis, John, Aquinas: Moral, Political and Legal Theory (New York: Oxford Univer-
sity Press, 1998).
Fitzpatrick, William, ‘No Darwinian Dilemma’, in Michael Bergmann and Patrick
Kain (eds), Challenges to Moral and Religious Belief (Oxford: Oxford University
Press, 2014), 237–55.
Foot, Philippa, Natural Goodness (Oxford: Clarendon Press, 2001) (NG).
Fox, Marvin, Interpreting Maimonides: Studies in Methodology, Metaphysics, and
Moral Philosophy (Chicago: University of Chicago Press, 1990) (IM).
Frank, Robert H., Gilovich, Thomas, and Regan, Dennis T., ‘Does Studying Economics
Inhibit Cooperation?’, Journal of Economic Perspectives, 7/ 2, (1993), 159–71.
Freppert, Lucan, The Basis of Morality according to William Ockham (Chicago:
Franciscan Herald Press, 1988).
Gauthier, René Antoine, and Jolif, Jean Yves, L’Éthique à Nicomaque (Louvain:
Publications Universitaires, 1970).
Geach, Peter, ‘Good and Evil’, Analysis, 17 (1957), 33–42.
al-Ghazali, Faith in Divine Unity and Trust in Divine Providence, trans. David Burrell
(Louisville, KY: Fons Vitae, 2001).
Gibbard, Allan, ‘Hare’s Analysis of “Ought” and its Implications’, in Douglas Seanor
and N. Fotion (eds), Hare and Critics (Oxford: Clarendon Press, 1988), 52–72.
Gibbard, Allan, Wise Choices, Apt Feelings: A Theory of Normative Judgment
(Cambridge, MA: Harvard University Press, 1990).
Gilliam, Ann (ed.), Voices for the Earth (San Francisco: Sierra Club Books, 1979).
Goodman, L. E., The God of Abraham (Oxford: Oxford University Press, 1996).
Goodman, L. E., Love Thy Neighbor as Thyself (Oxford: Oxford University Press,
2008).
Greene, Joshua, ‘From Neural “Is” to Moral “Ought”: What are the Moral Implications
of Neuroscientific Moral Psychology?’, Nature Reviews Neuroscience, 4/10 (2003),
847–50.
Greene, Joshua, Moral Tribes (New York: Penguin, 2013) (MT).
Grice, Paul, ‘Meaning’, repr. in Studies in the Way of Words (Cambridge, MA: Harvard
University Press, 1989).
Griffin, James, Well-Being: Its Meaning, Measurement, and Moral Importance
(Oxford: Clarendon Press, 1986).
Gustafson, James M., Protestant and Catholic Ethics: Prospects for Rapprochement
(Chicago: University of Chicago Press, 1978).
 Bibliography 319
Haidt, Jonathan, ‘The Emotional Dog and its Rational Tail’, Psychological Review, 108
(2001), 814–34.
Haidt, Jonathan, The Righteous Mind (New York: Vintage Books, 2012) (RM).
Hardin, Garrett, ‘The Tragedy of the Commons’, Science, 162 (1968), 1243–8.
Hare, John E., Plato’s Euthyphro (Bryn Mawr, PA: Bryn Mawr Commentaries, 1985).
Hare, John E., ‘Eleutheriotes in Aristotle’s Ethics’, Ancient Philosophy, 8 (1988), 19–32.
Hare, John E., ‘Kant’s Philosophy of Religion Reconsidered’, book review, Faith and
Philosophy, 11/ 1 (1994), 138–44.
Hare, John E., The Moral Gap: Kantian Ethics, Human Limits, and God’s Assistance
(Oxford: Clarendon Press, 1996).
Hare, John E., ‘Augustine, Kant, and the Moral Gap’, in Gareth B. Matthews (ed.), The
Augustinian Tradition (Berkeley and Los Angeles: University of California Press, 1999).
Hare, John E., God’s Call (Grand Rapids, MI: Eerdmans, 2001).
Hare, John E., Why Bother Being Good? (Eugene, OR: Wipf and Stock, 2002).
Hare, John E., ‘An Essay on Divine Authority’, book review, Mind, 113/450 (2004),
375–9.
Hare, John E., ‘Prescriptive Realism’, in Andrew Moore and Michael Scott (eds),
Realism and Religion (Aldershot: Ashgate, 2007).
Hare, John E., God and Morality (Oxford: Blackwell, 2009).
Hare, John E., ‘The Supervenience of Goodness on Being’, in Kevin Timpe (ed.),
Metaphysics and God (New York: Routledge, 2009).
Hare, John E., ‘Animal Sacrifices’, in Michael Bergmann, Michael J. Murray, and
Michael C. Rea (eds), Divine Evil? The Moral Character of the God of Abraham
(Oxford: Oxford University Press, 2011), 121–49.
Hare, John E., ‘Karl Barth, American Evangelicals, and Kant’, in Bruce McCormack
and Clifford Anderson (eds), Karl Barth and American Evangelicalism (Grand
Rapids, MI: Eerdmans, 2011), 73–90.
Hare, John E., ‘From Morality to Metaphysics’, book review, Notre Dame Philosophical
Reviews (August 2013).
Hare, John E., ‘The Place of Kant’s Theism in his Moral Philosophy’, in Mark
Timmons and Sorin Baiasu (eds), Kant on Practical Justification (Oxford: Oxford
University Press, 2013).
Hare, John E., ‘Conscience and the Moral Epistemology of Divine Command Theory’,
in Michael Bergmann and Patrick Kain (eds), Challenges to Moral and Religious
Belief: Disagreement and Evolution (Oxford: Oxford University Press, 2014).
Hare, John E., ‘Morality, Happiness, and Peter Singer’, in Nigel Biggar (ed.), Peter
Singer and Christian Ethics (Oxford: Clarendon Press, 2014).
Hare, John E., ‘Kant on Practical and Pathological Love’, in Frederick Simmons (ed.),
Love and Christian Ethics (Georgetown: Georgetown University Press, 2015).
Hare, John E., and Joynt, Carey B., Ethics and International Affairs (London: Macmillan,
1982).
Hare, R. M., ‘The University Discussion’, New Essays in Philosophical Theology, ed.
Antony Flew and Alasdair MacIntyre (London: SCM Press, 1955).
Hare, R. M., The Language of Morals (Oxford: Oxford University Press, 1961).
Hare, R. M., ‘Meaning and Speech Acts’, in Practical Inferences (London: Macmillan,
1971), 74–9.
320 Bibliography
Hare, R. M., Applications of Moral Philosophy (London: Macmillan, 1972).
Hare, R. M., ‘Geach: Good and Evil’, in Essays on the Moral Concepts (London:
Macmillan, 1972) 29–38.
Hare, R. M., Freedom and Reason (Oxford: Oxford University Press, 1977).
Hare, R. M., Moral Thinking (Oxford: Clarendon Press, 1981).
Hare, R. M., Essays in Ethical Theory (Oxford: Clarendon Press, 1989).
Hare, R. M., Essays on Religion and Education (Oxford: Clarendon Press, 1992).
Hare, R. M., Objective Prescriptions (Oxford: Clarendon Press, 1999).
Hare, R. M., Sorting Out Ethics (Oxford: Clarendon Press, 1997).
Harinck, George, ‘How Can an Elephant Understand a Whale and Vice Versa? The
Dutch Origins of Cornelius Van Til’s Appraisal of Karl Barth’, in Bruce McCormack
and Clifford Anderson (eds), Karl Barth and American Evangelicalism (Grand
Rapids, MI: Eerdmans, 2011), 13–41.
Harrison, Jonathan, Hume’s Theory of Justice (Oxford: Clarendon Press, 1981).
Hart, H. L. A., The Concept of Law (Oxford: Oxford University Press, 1994).
Hauerwas, Stanley, and Pinches, Charles, Christians among the Virtues: Theological
Conversations with Ancient and Modern Ethics (Notre Dame, IN: University of
Notre Dame Press, 1997).
Hayes, Christine, What’s Divine about Divine Law? Early Perspectives (Princeton:
Princeton University Press, 2015).
Hector, Kevin, Theology without Metaphysics: God, Language and the Spirit of Recog-
nition (Cambridge: Cambridge University Press, 2011).
Henry, Carl, God, Revelation and Authority, iii (Waco, TX: Word Books, 1979).
Herman, Barbara, The Practice of Moral Judgment (Cambridge, MA: Harvard Univer-
sity Press, 1993).
Hitler, Adolf, Mein Kampf (My Struggle), trans. J. Murphy (London: Hurst and Blakett,
1939).
Hopkins, Gerard Manley, Gerard Manley Hopkins: Poems and Prose, ed.
W. H. Gardner (Harmondsworth: Penguin, 1953).
Hourani, George, Islamic Rationalism: The Ethics of ‘Abd al-Jabbar (Oxford: Claren-
don Press, 1971).
Hourani, George, Reason and Tradition in Islamic Ethics (Cambridge: Cambridge
University Press, 1985).
Hursthouse, Rosalind, On Virtue Ethics (Oxford: Oxford University Press, 1999) (OVE).
Hutcheson, Francis, Inquiry into the Original of our Ideas of Beauty and Virtue, ed.
Wolfgang Leidhold (Indianapolis: Liberty Fund, 2004).
Irwin, Terence, Aristotle’s First Principles (Oxford: Clarendon Press, 1988).
Irwin, Terence, The Development of Ethics (Oxford: Oxford University Press, 2011).
Jackson, Timothy, Love Disconsoled (New York: Cambridge University Press, 1999).
Jaeger, Werner, ‘Aristotle’s Use of Medicine as a Model of Method in his Ethics’,
Journal of Hellenic Studies, 77/1 (1957), 54–61.
James, William, Varieties of Religious Experience (New York: Modern Library, 1902).
James, William, ‘The Will to Believe’, in Pragmatism and Other Essays (New York:
Pocket Books, 1963).
Johnson, Dominic, ‘God’s Punishment and Public Goods: A Test of the Supernatural
Punishment Hypothesis in 186 World Cultures’, Human Nature, 16 (2009), 410–46.
 Bibliography 321
al-Juwayni, Abi’l Ma’ali, Kitab al-Irshad, ed. M. Y. Musa and A. A. Abd al-Hamid
(Cairo, 1950).
Kain, Patrick, ‘Self-Legislation in Kant’s Moral Philosophy’, Archiv für Geschichte der
Philosophie, 86/3 (2004), 257–306.
Kain, Patrick, ‘Kant’s Defense of Human Moral Status’, Journal of the History of
Philosophy, 47/1 (2009), 59–102.
Kain, Patrick, ‘Practical Cognition, Intuition and the Fact of Reason’, in Benjamin
Lipscomb and James Krueger (eds), Kant’s Moral Metaphysics (Berlin: DeGruyter,
2010), 211–30.
Kant, Immanuel, Natürliche Theologie Volckmann [Lectures on Natural Theology
(Volckmann)] (Berlin Academy Edition; Berlin: George Reiner, later Walter de
Gruyter, 1900–).
Kant, Immanuel, Religion within the Limits of Reason Alone, trans. Theodore
M. Greene and Hoyt H. Hudson (New York: Harper, 1960).
Kant, Immanuel, Speculative Beginning of Human History, in Perpetual Peace and
Other Essays, trans. Ted Humphrey (Indianapolis: Hackett, 1983).
Kant, Immanuel, The Conflict of the Faculties, trans. Mary J. Gregor and Robert
Anchor, in Religion and Rational Theology, ed. and trans. Allen W. Wood and
George di Giovanni (Cambridge: Cambridge University Press, 1996) (SF).
Kant, Immanuel, The End of All Things, trans. Allen W. Wood, in Religion and
Rational Theology, ed. and trans. Allen W. Wood and George di Giovanni (Cam-
bridge: Cambridge University Press, 1996).
Kant, Immanuel, Lectures on the Philosophical Doctrine of Religion (Pölitz) [Religion-
slehre nach Pölitz], trans. Allen W. Wood, in Religion and Rational Theology, ed.
and trans. Allen W. Wood and George di Giovanni (Cambridge: Cambridge
University Press, 1996).
Kant, Immanuel, On the Miscarriage of All Philosophical Trials in Theodicy, trans. George
di Giovanni, in Religion and Rational Theology, ed. and trans. Allen W. Wood and
George di Giovanni (Cambridge: Cambridge University Press, 1996) (M.).
Kant, Immanuel, Religion within the Boundaries of Mere Reason, trans. George di
Giovanni, in Religion and Rational Theology, ed. and trans. Allen W. Wood and
George di Giovanni (Cambridge: Cambridge University Press, 1996) (Rel.).
Kant, Immanuel, Lectures on Ethics (Collins) [Moralphilosophie Collins], in Lectures on
Ethics, ed. Peter Heath and J. B. Schneewind, trans. Peter Heath (Cambridge:
Cambridge University Press, 1997) (MC).
Kant, Immanuel, Critique of Pure Reason, ed. and trans. Paul Guyer and Allen
W. Wood (Cambridge: Cambridge University Press, 1998) (KrV).
Kant, Immanuel, Critique of Practical Reason, in Practical Philosophy, ed. and trans.
Mary J. Gregor (Cambridge: Cambridge University Press, 1999) (KpV).
Kant, Immanuel, Groundwork of the Metaphysics of Morals, in Practical Philosophy,
ed. and trans. Mary J. Gregor (Cambridge: Cambridge University Press, 1999)
(Gl.).
Kant, Immanuel, The Metaphysics of Morals, in Practical Philosophy, ed. and trans.
Mary J. Gregor (Cambridge: Cambridge University Press, 1999) (MdS).
Kant, Immanuel, Critique of the Power of Judgment, ed. Paul Guyer, trans. Paul Guyer
and Eric Matthews (Cambridge: Cambridge University Press, 2000) (KU).
322 Bibliography
Kant, Immanuel, Anthropology from a Pragmatic Point of View, trans. Robert
B. Louden, in Anthropology, History, and Education, ed. Günter Zöller and Robert
B. Louden, trans. Mary J. Gregor et al. (Cambridge: Cambridge University Press,
2007) (Anth.).
Katz, Leonard D. (ed.), Evolutionary Origins of Morality: Cross-Disciplinary Perspec-
tives (Exeter: Imprint Academic, 2000).
Kelemen, Deborah, ‘Are Children “Intuitive Theists”?: Reasoning about Purpose and
Design in Nature’, Psychological Science, 15 (2004), 295–301.
Kelsey, David, Eccentric Existence (Louisville, KY: John Knox Press, 2009).
Kenny, Anthony, Aristotle’s Theory of the Will (London: Duckworth, 1979).
Kent, Bonnie, ‘Scotus on Virtues’, in Thomas Williams (ed.), The Cambridge Com-
panion to Duns Scotus (Cambridge: Cambridge University Press, 2003).
Kierkegaard, Sren, Either/Or, ed. and trans. Howard V. Hong and Edna Hong
(Princeton: Princeton University Press, 1987).
Kierkegaard, Sren, Works of Love, ed. and trans. Howard V. Hong and Edna H. Hong
(Princeton: Princeton University Press, 1998).
Kitcher, Philip, Living with Darwin (New York: Oxford University Press, 2007).
Kitcher, Philip, The Ethical Project (Cambridge, MA: Harvard University Press, 2011).
Kitcher, Philip, Life beyond Faith: The Case for Secular Humanism (New Haven: Yale
University Press, 2014).
Korsgaard, Christine, ‘Two Distinctions in Goodness’, repr. in Creating the Kingdom of
Ends (Cambridge: Cambridge University Press, 1996), 249–74.
Korsgaard, Christine, with G. A. Cohen, Raymond Geuss, Thomas Nagel, and Bernard
Williams, Sources of Normativity, ed. Ononra O’Neill (Cambridge: Cambridge
University Press, 1996).
Lambert, Frank, ‘ “I Saw the Book Talk”: Slave Readings of the First Great Awakening’,
Journal of African American History, 87 (2002), 14.
Lansing, J. S., Priests and Programmers: Technologies of Power in the Engineered
Landscapes of Bali (Princeton: Princeton University Press, 1991).
Lawrence, Gavin, ‘The Rationality of Morality’, in Rosalind Hursthouse, Gavin Lawr-
ence, and Warren Quinn (eds), Virtues and Reasons (Oxford: Clarendon Press, 1995).
Leibniz, Gottfried Wilhelm, Opinion on the Principles of Pufendorf, in Political Writ-
ings, ed. Patrick Riley (Cambridge: Cambridge University Press, 1989).
Leopold, Aldo, Round River (New York: Galaxy Books, 1953).
Lewis, C. S., Reflections on the Psalms (New York: Harcourt Brace Jovanovich, 1958).
Lipscomb, Benjamin, ‘Power and Authority in Pufendorf ’, History of Philosophy
Quarterly, 22/3 (2005), 201–19.
Locke, Dustin, ‘Darwinian Normative Skepticism’, in Michael Bergmann and Patrick
Kain (eds), Challenges to Moral and Religious Belief (Oxford: Oxford University
Press, 2014).
Lucas, John, ‘Reasons for Loving and Being Loved’, in Freedom and Grace: Essays by
J. R. Lucas (London: SPCK, 1976).
Luther, Martin, The Bondage of the Will, trans. J. I. Packer and Olaf Raymond
Johnston (New York: Fleming H. Revell, 1957).
McCormack, Bruce, and Anderson, Clifford (eds), Karl Barth and American Evangel-
icalism (Grand Rapids, MI: Eerdmans, 2011).
 Bibliography 323
MacDonald, Scott, ‘Egoistic Rationalism: Aquinas’s Basis for Christian Morality’, in
Michael Beaty (ed.), Christian Theism and the Problems of Philosophy (Notre Dame,
IN: University of Notre Dame Press, 1990).
McDowell, John, ‘Virtue and Reason’, Monist, 62 (1979), 331–50.
McDowell, John, ‘Values as Secondary Qualities’, in Ted Honderich (ed.), Morality
and Objectivity: A Tribute to J. L. Mackie (London: Routledge & Kegan Paul, 1985).
McDowell, John, Mind and World (Cambridge, MA: Harvard University Press, 1994).
McKenny, Gerald, The Analogy of Grace (Oxford: Oxford University Press, 2010).
Mackie, J. L., Ethics: Inventing Right and Wrong (Harmondsworth: Penguin, 1977) (E.).
Mackie, J. L., The Miracle of Theism (Oxford: Clarendon Press, 1982).
Maimonides, Be’ur Millot ha-Higgayon, with commentary by Moses Mendelssohn
(Berlin: B. Cohen, 1927).
Maimonides, The Guide for the Perplexed, trans. M. Friedländer (1904; 2nd edn, New
York: Dover, 1956).
Maimonides, The Guide of the Perplexed, trans. Shlomo Pines (Chicago: University of
Chicago Press, 1963) (Guide).
Maimonides, The Eight Chapters of Maimonides on Ethics, ed. and trans. Joseph
I. Gorfinkle (New York: Columbia University Press, 1966).
Maimonides, Mishneh Torah: Yesodei ha-Torah, 1.1, ed. Rabbi Eliahu Touger (Jeru-
salem: Moznaim, 1989).
Malik-Shaybani, al-Muwatta (Lucknow: 1880, 1888–9).
Marion, Jean-Luc, L’Idole et la distance: Cinq études (Paris: Éditions Grasset and
Fasquelle, 1977).
al-Maturidi, Abu Mansur, Kitab al-Tawhid, ed. Fathalla Kholeif (Beirut: Dar al-
Mashriq, 1970) (Tawhid).
al-Maturidi, Abu Mansur, Tawilat Ahl al-Sunnam, ed. Ibrahim Awadain and al-Sayyid
Awadain (Cairo, 1971).
Meilaender, Gilbert, The Theory and Practice of Virtue (Notre Dame, IN: University of
Notre Dame Press, 1984).
Mellema, Gregory, Beyond the Call of Duty: Supererogation, Obligation, and Offense
(Albany, NY: SUNY Press, 1991).
Mendenhall, George E., Law and Covenant in Israel and the Ancient Near East
(Pittsburgh: Biblical Colloquium, 1955).
Micklethwait, John, and Wooldridge, Adrian, God is Back (London: Penguin Books,
2009).
Mill, J. S., Three Essays on Religion (London: Henry Holt, 1874).
Mill, J. S., Utilitarianism, ed. Mary Warnock (London: Fontana (Collins), 1962).
Moore, G. E., Principia Ethica (1903; Cambridge: Cambridge University Press, 1993).
Murdock, George P., Ethnographic Atlas (Pittsburgh: University of Pittsburgh Press,
1967).
Murdock, George P., and White, Douglas R., ‘Standard Cross-Cultural Sample’,
Ethnology, 8 (1969), 329–69.
Murphy, Mark, Natural Law and Practical Rationality (Cambridge: Cambridge
University Press, 2001).
Murphy, Mark, An Essay on Divine Authority (Ithaca, NY: Cornell University Press,
2002).
324 Bibliography
Murphy, Mark, God and Moral Law: On the Theistic Explanation of Morality (Oxford:
Oxford University Press, 2012).
Nagel, Thomas, The Possibility of Altruism (Oxford: Clarendon Press, 1970).
Nagel, Thomas, Mind and Cosmos (New York: Oxford University Press, 2012).
Newman, John Henry, The Grammar of Assent (London: Assumption Press, 2013).
Novak, David, Natural Law in Judaism (Cambridge: Cambridge University Press,
1988) (Natural Law).
Novak, David, Natural Law and Revealed Torah, ed. Hava Tirosh-Samuelson and
Aaron W. Hughes (Leiden: Brill, 2014) (NLRT).
O’Donovan, Oliver, Finding and Seeking (Grand Rapids, MI: Eerdmans, 2014).
Ogden, Stephen, ‘Receiving and Making Aristotle’s Intellect: A New Assessment of
Averroes and Aquinas,’ Ph.D. dissertation, Yale University, 2015.
Otto, Rudolf, The Idea of the Holy (Oxford: Oxford University Press, 1923).
Palmer, F. R., Mood and Modality (Cambridge: Cambridge University Press, 1986).
Parfit, Derek, On What Matters, 2 vols (Oxford: Oxford University Press, 2011).
Paton, H. J., The Categorical Imperative (Tiptree, Essex: Anchor Press, 1946).
Pieper, Joseph, The Four Cardinal Virtues: Prudence, Justice, Fortitude, Temperance
(Notre Dame, IN: University of Notre Dame Press, 1966).
Pigloicci, Massimo, ‘Primates, Philosophers, and the Biological Basis of Morality’,
Biology and Philosophy, 22 (2007): 611–18.
Plantinga, Alvin, Warranted Christian Belief (New York: Oxford University Press,
2000).
Plantinga, Alvin, ‘The Evolutionary Argument against Naturalism’, in J. Beilby (ed.),
Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument against Natur-
alism (Ithaca, NY: Cornell University Press, 2002), 1–12.
Plantinga, Alvin, ‘Games Scientists Play’, in Jeffrey Schloss and Michael J. Murray
(eds), The Believing Primate (Oxford: Oxford University Press, 2009).
Plantinga, Alvin, ‘Naturalism, Theism, Obligation and Supervenience’, Faith and
Philosophy, 27/3 (2010), 247–72.
Plato, Collected Dialogues, ed. Edith Hamilton and Huntington Cairns (Princeton:
Princeton University Press, 1961).
Pollock, Benjamin, Franz Rosenzweig’s Conversions: World Denial and World Redemp-
tion (Bloomington, IN: Indiana University Press, 2014).
Porter, Jean, Nature as Reason (Grand Rapids, MI: Eerdmans, 2005).
Porter, Jean, ‘The Desire for Happiness and the Virtues of the Will’, Journal of Moral
Theology, 3/1 (2014), 18–38.
Pufendorf, Samuel, On the Law of Nature and Nations, trans. C. H. Oldfather and
W. A. Oldfather (Oxford: Clarendon Press, 1934).
Putnam, R. D., and Campbell, D. E., American Grace: How Religion Divides and Unites
Us (New York: Simon and Schuster, 2010).
Quinn, Philip, ‘Divine Command Theory’, in Hugh LaFollette (ed.), Guide to Ethical
Theory (Oxford: Blackwell, 1999), 54–5.
Raleigh, Walter, History of the World (London: Walter Burre, 1614).
Raz, Joseph, The Morality of Freedom (Oxford: Oxford University Press, 1986).
Rist, John, Real Ethics: Reconsidering the Foundations of Morality (New York: Cambridge
University Press, 2002).
 Bibliography 325
Ritchie, Angus, From Morality to Metaphysics (Oxford: Oxford University Press,
2012).
Roberts, Robert C., Emotions (Cambridge: Cambridge University Press, 2003).
Robinson, John, An Introduction to Early Greek Philosophy (Boston: Houghton
Mifflin, 1968).
Rorty, Richard, ‘Religion as Conversation-Stopper’, Common Knowledge, 3 (1994),
1–6.
Rorty, Richard, ‘Religion in the Public Square: A Reconsideration’, Journal of Religious
Ethics, 31/1 (2003), 141–9.
Rosenzweig, Franz, The Star of Redemption, trans. William H. Hallo (Notre Dame, IN:
Notre Dame Press, 1970).
Ross, W. D., The Right and the Good (Oxford: Oxford University Press, 1930).
Rudolph, Ulrich, Al-Maturidi and Sunni Theology in Samarkand (Leiden: Brill, 2015)
(Theology).
Ruse, Michael, Can a Darwinian Be a Christian? (Cambridge: Cambridge University
Press, 2001).
Ruse, Michael, ‘Is Darwinian Metaethics Possible (And If It Is, Is it Well Taken)?’ in
Giovanni Boniolo and Gabriele De Anna (eds), Evolutionary Ethics and Contem-
porary Biology (Cambridge: Cambridge University Press, 2006) (DM).
Ruse, Michael, Science and Spirituality (Cambridge: Cambridge University Press,
2010).
Ruse, Michael, and Wilson, E. O., ‘The Evolution of Morality’, New Scientist, 1478
(1985), 108–28.
Schloss, Jeffrey, and Murray, Michael J. (eds), The Believing Primate (Oxford: Oxford
University Press, 2009).
Scotus, John Duns, Duns Scotus on the Will and Morality, ed. William A. Frank, trans.
Allan B. Wolter (Washington: Catholic University of America Press, 1986).
Seanor, Douglas, and Fotion, N. (eds), Hare and Critics (Oxford: Clarendon Press,
1988).
Searle, J. R., ‘How to Derive “Ought” from “Is” ’, Philosophical Review, 73 (1964),
43–58.
Sidgwick, Henry, The Methods of Ethics (Indianapolis: Hacker, 1981).
Singer, Peter, The Expanding Circle: Ethics, Evolution, and Moral Progress (New York:
Farrar, Straus and Giroux, 1981).
Smith, Michael, The Moral Problem (Oxford: Blackwell, 1994).
Sosis, Richard, ‘Religion and Intragroup Cooperation: Preliminary Results of a Com-
parative Analysis of Utopian Communities’, Cross-Cultural Research, 34 (2000),
264–74.
Spencer, Herbert, Social Statics; or the Conditions Essential to Human Happiness
Specified and the First of Them Developed (London: J. Chapman, 1851).
Stern, Josef, The Matter and Form of Maimonides’s Guide (Cambridge, MA: Harvard
University Press, 2013).
Strauss, Leo, ‘The Literary Character of The Guide for the Perplexed’, in Persecution
and the Art of Writing (Glencoe, IL: Free Press, 1952).
Street, Sharon, ‘A Darwinian Dilemma for Realist Theories of Value’, Philosophical
Studies, 127 (2006), 109–66.
326 Bibliography
Stump, Eleonore, Aquinas (New York: Routledge, 2003).
Sturgeon, Nicholas, ‘Moral Explanations’, in Geoffrey Sayre-McCord (ed.), Essays on
Moral Realism (Ithaca, NY: Cornell University Press, 1988).
Suarez, Francisco, A Treatise on Laws and God the Lawgiver, trans. Gwladys Williams,
Ammi Brown, and John Waldron, in Selections from Three Works of Francisco
Suarez, ii (Oxford: Clarendon Press, 1944).
Swinburne, Richard, The Coherence of Theism (Oxford: Clarendon Press, 1993).
Tanner, Kathryn, God and Creation in Christian Theology (Oxford: Blackwell, 1988).
Taylor, Charles, Sources of the Self (Cambridge, MA: Harvard University Press, 1989).
The New Oxford Annotated Bible: New Revised Standard Version, Fully Revised Fourth
Edition, ed. Michael D. Coogan, Oxford: Oxford University Press, 2010.
Van Inwagen, Peter, ‘The Place of Chance in a World Sustained by God’, in Thomas
V. Morris (ed.), Divine and Human Action (Ithaca, NY: Cornell University Press,
1988), 42–65.
Van Inwagen, Peter, Material Beings (Ithaca, NY: Cornell University Press, 1990).
Van Inwagen, Peter, ‘Belief in the Supernatural’, in Jeffrey Schloss and Michael
A. Murray (eds), The Believing Primate (Oxford: Oxford University Press, 2009),
128–38.
Van Til, Cornelius, The New Modernism (London: James Clarke, 1946).
Vasalou, Sophia, Moral Agents and their Deserts: The Character of Mu’tazilite Ethics
(Princeton: Princeton University Press, 2008).
Velleman, David, ‘Narrative Explanation’, Philosophical Review, 112/1 (2003), 1–25.
Verney, Stephen, Water into Wine (London: Fount Paperbacks, 1985).
Volf, Miroslav, Allah: A Christian Response (New York: Harper Collins, 2011).
Volf, Miroslav, A Public Faith: How Followers of Christ Should Serve the Common
Good, Grand Rapids, MI: Brazos Press, 2011.
Waal, Frans de, Good-Natured: The Origins of Right and Wrong in Humans and Other
Animals (Cambridge, MA: Harvard University Press, 1996).
Waal, Frans de, ‘The Animal Roots of Human Morality’, New Scientist, 192 (2006),
60–1.
Waal, Frans de, The Age of Empathy: Nature’s Lessons for a Kinder Society (New York:
Harmony Books, 2009).
Waal, Frans de, ‘Morals without God?’, New York Times Opinionator, 17 October 2010
<http://opinionator.blogs.nytimes.com> (accessed May 2011).
Waal, Frans de, The Bonobo and the Atheist (New York: W. W. Norton & Company,
2014).
Walzer, Michael, ‘Political Action: The Problem of Dirty Hands’, in Marshall Cohen,
Thomas Nagel, and Thomas Scanlon (eds), War and Moral Responsibility (Prince-
ton: Princeton University Press, 1974).
Warren, Andy, ‘Modality, Reference and Speech Acts in the Psalms’, dissertation,
Cambridge University, 1998.
Watson, Gary, ‘Responsibility and the Limits of Evil: Variations on a Strawsonian
Theme’, in F. D. Schoeman (ed.), Responsibility, Character, and the Emotions
(Cambridge: Cambridge University Press, 1987).
Webster, John, Barth’s Ethics of Reconciliation (Cambridge: Cambridge University
Press, 1995).
 Bibliography 327
Webster, John, Barth’s Moral Theology: Human Action in Barth’s Thought (Grand
Rapids, MI: Eerdmans, 1998).
Weinfeld, Moshe, ‘The Covenant of Grant in the Old Testament and in the Ancient
Near East’, Journal of the American Oriental Society, 90/2 (1970), 184–203.
Weithman, Paul, ‘God’s Velveteen Rabbit’, Journal of Religious Ethics, 37/2 (2009),
256.
Williams, Bernard, ‘Persons, Character, and Morality’, in Moral Luck (Cambridge:
Cambridge University Press, 1981).
Williams, Bernard, ‘Evolution, Ethics and the Representation Problem’, in Making
Sense of Humanity (Cambridge: Cambridge University Press, 1995).
Williams, Thomas, ‘From Metaethics to Action Theory’, in Thomas Williams (ed.),
The Cambridge Companion to Duns Scotus (Cambridge: Cambridge University
Press, 2003).
Wilson, Edward O., Sociobiology: The New Synthesis (Cambridge, MA: Harvard
University Press, 1975).
Winter, Tim, ‘Response to Christopher Insole’, Studies in Christian Ethics, 25/2 (2012),
219–22.
Wittgensein, Ludwig, Tractatus Logico-Philosophicus (London: Routledge & Kegan
Paul, 1961).
Wittgensein, Ludwig, Philosophical Investigations (Oxford: Blackwell, 1963).
Wolf, Susan, Freedom within Reason (Oxford: Oxford University Press, 1990).
Wolterstorf, Nicholas, ‘Conundrums in Kant’s Rational Religion’, in Philip Rossi and
Michael Wreen (eds), Kant’s Philosophy of Religion Reconsidered (Bloomington, IN:
Indiana University Press, 1991).
Wolterstorf, Nicholas, Divine Discourse: Philosophical Reflections on the Claim that
God Speaks (Cambridge: Cambridge University Press, 1995).
Wolterstorf, Nicholas, Justice: Rights and Wrongs (Princeton: Princeton University
Press, 2008).
Wolterstorf, Nicholas, ‘Justice as Inherent Rights: A Response to My Commentators’,
Journal of Religious Ethics, 37/2 (2009), 261–79.
Wolterstorff, Nicholas, ‘Reading Joshua’, in Michael Bergmann, Michael J. Murray,
and Michael C. Rea (eds), Divine Evil? The Moral Character of the God of Abraham
(Oxford: Oxford University Press, 2011), 236–65.
Wood, Allen, ‘Kant’s Deism’, in Philip Rossi and Michael Wreen (eds), Kant’s Phil-
osophy of Religion Reconsidered (Bloomington, IN: Indiana University Press, 1991).
Wood, Allen, Kant’s Ethical Thought (Cambridge: Cambridge University Press, 1999).
Zagzebski, Linda, Divine Motivation Theory (Cambridge: Cambridge University Press,
2004).
 Index of Biblical Passages

Genesis Psalms
1: 3 35, 48 1: 5–6 54
1: 4 233 19 304
1: 22 48 19: 7 229
1: 26–7 27 22: 1 73
1: 31 28, 103 33: 9 48
2: 16–17 42 72: 12–14 300
3: 1–17 262 84: 8 38
3: 9 253, 302 85: 8 48
3: 12–13 254 85: 10 89, 170, 180
9: 4 232 95: 7 254
9: 6 27 119: 176 1
9: 8–17 39 Song of Songs
15: 6 238 8: 6 248
18 238
22: 1–19 35, 106, 150, 152, 174, 254 Isaiah
27: 35 165 29: 13–18 164
40: 18 97
Exodus
12: 36 106 Matthew
20 17, 100–2, 152 5: 17–43 165, 304
32: 32 73 5: 43–8 45, 115, 164, 305
34: 28 36, 148 6: 1–34 304
Leviticus 6: 9 158
18: 14 231 6: 33 90
7: 12 288
Deuteronomy 11: 28 40
4: 8 229 14: 28 182
5: 1 174 15: 21–8 96, 164–6
5: 6–7 148 16: 26 42
6: 5 254 19: 16–22 44, 47
9: 4–6 36 21: 1–12 143
13: 6–10 268 22: 36–40 148
28–31 38–9, 48 25: 40 156
Joshua 27: 45 73
6: 4 35, 148 Mark
10: 28–43 298 1: 17 61
Judges 5: 21–43 107, 165
6: 23 48 8: 34 61
21: 25 247 8: 34–7 145
I Samuel 10: 41 154
3: 4–5 179 12: 42 115
18: 24 268 Luke
II Samuel 8: 45–56 107
7: 12–17 39 10: 25–37 28, 47, 52, 267, 306
12: 1–14 62 17: 10 45
Job John
27: 5–6 11 1: 1–3 48, 249
42: 3 11 1: 42 146
330 Index of Biblical Passages
John (cont.) Ephesians
3: 19 115–16 2: 10 107
4: 13 161 4: 1 40
10: 3 254 Philippians
10: 17–18 252 2: 3 115, 154
13: 4–17 155 2: 6 161
14: 19 145
14: 23–24 18 II Thessalonians
18: 15–27 146 3: 15 38
Acts I Timothy
9: 1–9 166 4: 12 154
17: 24 58 Titus
2: 6 38
Romans
2: 15 38
1: 20 283
8: 26–7 163 Hebrews
9: 3 73 11: 19 102
I Corinthians James
11: 27–32 180 2: 17 167
13 95 II Peter
Galatians 3: 13 105
2: 20 145 Revelation
4: 5 1 2: 17 29, 145
5: 1 160 21: 1 105
 Index of Names and Topics

‘Abd al-Jabbar 22n68, 188–95, 204–6, Barrett, J. 282–5, 293

213–15, 230, 232, 306 Barth, K. 13, 29, 40, 95, 142–3, 185,
Abraham 165, 238–9, 245n84 248, 302
binding 35, 57, 102, 106, 150n27, 174–5, access to commands 173–83
182, 242, 254 and Kant 166–9
Abrahamic Faiths 1, 27, 53, 57, 103n15, four constraints 35–7, 105, 107–9, 139,
178–9, 185, 223, 268, 305–6 143, 245
accountability 29, 46, 54, 60–1, 157, 171 freedom 157–64
action 6, 35–6, 68–9, 84–9, 105, 119, 147–51, particularity 143–57, 251
155, 162–3, 188–92, 199–203, 204–5, beauty 11, 21, 75, 104, 105, 201, 212, 263
209–10, 254, 311 Bentham, J. 9, 269, 288–9
Adam 42–3, 44, 232–4, 253–4, 302 Biggar, N. 85n65, 152n30, 177n127, 179n136
Adams, R. M. 19, 20, 24, 57nn64–5, 60, Blackburn, S. 122
102–3, 106, 111, 112–16, 171, 174, 181, blame 46, 188–9, 191–2, 199–200, 204–5
203, 297 Bloom, P. 281–5, 303
Alston, W. 182–3 Blum, L. 149–50, 193
altruism 267, 293, 296, 299, 303 Bonaventure 101n7
analogy 49, 51–2, 217, 220 Boyd, R. 122n75
Annas, J. 67n8, 78n47, 80n52, 81, 83, 95 Boyle, J. 109n29
Anscombe, G. E. M. 33–4, 131, 135–6 Brandt, R. 25n82
Anselm 72, 74, 88 Brink, D. 119n65
anti-realism 121–4, 273–85 Brittain, C. 79n48
Aquinas 20–1, 34, 49, 73, 86–94, 95, 97, 99, Broad, C. D. 127n89
103, 110, 130, 178, 181 Broadie, S. 71n23, 227n20
arbitrariness 24, 102–3, 106, 302, 312 Buber, M. 144, 178
Aristotle 14, 21, 24, 27, 33n6, 56, 74, 80, 97, Butler, J. 50, 71–2, 80, 88, 127n89,
130, 131, 145, 150, 202, 204–6, 220, 181n144, 289
240, 244, 269, 270–1
friendship 18, 81, 90 call (God’s) 25, 29–31, 40, 46–7, 108, 131–2,
happiness 64, 67–72, 90, 96, 114–15, 266 146, 153n33, 199, 253–4, 260
the mean 226–30 Calvin, J. 15, 46, 58, 73, 93, 129, 283–4, 291
the noble 69–70, 90, 115, 205, 228 Campbell, D. 276
religion 57, 73, 97, 103, 252 Camus, A. 9
Arner, N. 58n66, 250n105 Caputo, J. 33n3
Arnhart, L. 137–8, 266–7, 303, 306 Carson, T. 111n36
Arius Didymus 83–4 Cartwright, J. 134n107
al-Ash‘ari 13, 195–7, 206–9, 215–17 Ceric, M. 184
al-Attar, M. 187n7, 191n25, 194n36, 204n68, Chignell, A. 9n19, 172n112
205n75 Chu Hsi 14
Augustine 12, 14, 179, 251n108 Cicero 78–9, 83
Austin, J. L. 19, 34, 119n66, 194 Clark, G. H. 167n84
authority Clough, D. 155n39, 156n43
divine 9, 38–41, 49–55, 58, 111, 156, Cohen, H. 226n10, 236, 244, 248, 253, 259
180–1, 206 commands 20, 33–7, 55–6, 193
human 38, 221, 228–30, 247, 298 see, divine commands
autonomy 46n34, 60, 157, 245, 259 condemnation 38, 44
Averroes 97n100, 186 conscience 40n, 59–60, 169–73, 177–8,
al-Baqillani 222 181n144, 271, 294
332 Index of Names and Topics
consequentialism eudaemonism 2, 63–98, 111, 192, 206, 242
see, utilitarianism Evans, C. S. 16, 19, 23, 40, 62, 108n26,
Copp, D. 123n79 229n22, 297
counsels evil 9–11, 12–13, 50, 92, 116, 140–1, 160, 168,
see, divine commands 188n10, 194, 206–8, 212–13, 233, 252,
covenant 36, 38–9, 54, 96, 148, 159, 164–5, 265, 298n120, 303
238–9, 252, 303 propensity to 12, 26, 129, 170, 245, 259,
creation 10–11, 28, 48–9, 95, 104–5, 153n33, 272, 274, 291n93
180, 200, 212, 237, 240–1, 244, 248, evolution 91–2, 120, 132, 137, 261–308
249–51, 256–60 Ewing, A. C. 22n68
Crisp, R. 68n13 expressivism 113, 118–24, 312
Cronquist, J. 35n12
Cross, R. 75n37 faith 9–11, 128, 139, 163, 167n90, 216,
Cudworth, R. 17n51, 58 218n122, 224, 237–8, 252–3, 296, 306
the Fall 74, 101n7, 129, 251, 254n118, 259
Darwall, S. 17n51, 44, 46, 53–4, 56n62, 57–60, family 16, 40n24, 70, 81–2, 83, 140, 153n33,
110, 297 156, 260
de Waal, F. 85, 129, 134, 291, 300 Fichte, J. G. 56n62, 181
deductivism 100–2, 112–17, 121, 124, Finnis, J. 109, 250
128–41, 203, 265–6 Fitzpatrick, W. 279n49, 28
Derrida, J. 168 Foot, P. 100, 106n22, 117–41, 227, 251
Descartes, R. 9, 176, 219, 282 forgiveness 1, 13, 15, 307
desire 20–3, 33–5, 57, 76, 89, 97–8, 113, Fox, M. 97, 212n99, 224–35, 243, 245, 257–8
118, 121–3, 124–6, 137–8, 155, 233, Francis of Assisi 101, 126
266, 293 Frank, R. 276
dignity freedom
see, human dignity divine 93, 108, 112, 157, 195–6, 249
divine commands 1, 18, 54, 61–2, 108, 151, human 6, 12, 26, 52, 56, 64n1, 74, 157–66,
172, 199, 202, 268, 274, 293, 302 203–12, 252, 295
phenomenology 41, 162, 171, 173–83 Freppert, L. 56n61
varieties of 37–49 Freud, S. 283
counsel 40, 44–8, 101n7, 174, 202 friendship 16, 18, 70–2, 78–82, 83, 90, 93,
directly effective 48–9, 57, 190, 202, 208 109–10, 149–51, 250, 286
permission 42–4, 202 function 68, 106, 130, 132, 135, 227, 282
precept 37–41, 105, 127, 143, 176, 202 fundamentalism 187
prohibition 41–2, 110, 136, 176, 201–3,
210–11, 231, 304 Gauthier, R. A. 68n13
divine command theory 2–4, 18, 22, 23n73, Geach, P. 105, 130, 133, 135
57, 140, 192 Gibbard, A. 119n67, 126
objections God
arbitrariness 24–5, 102–9, 194, 297 as commander, see divine commands
autonomy 157–64, 179–80, 187, 297 as just, see justice, divine
pluralism 217–20, 236–8, 298 as lover 28–9, 94, 139, 160–1, 181–2, 249,
redundancy 20, 63, 193 251–3, 300–3
regress 17–18, 57–60, 297 as sovereign
Dodsworth, C. 53 judge 53–4, 169–71, 180
Douglas, M. 237n52 legislator 30, 172, 174, 301
Driver, J. 131n98 ruler 52–3, 170–1, 211–12
as Trinity 30, 48, 95, 101, 161, 185
Edwards, J. 73n30 see, Holy Spirit, Jesus Christ
election 13, 73, 93, 95, 157, 182 Goodman, L. E. 245n84
emotion 47, 121–3, 150, 227, 233, 268, 272, goodness 20–5, 28–9, 74–5, 103–6, 113,
290, 293–5, 300 122, 125, 130–5, 137–41, 150, 155,
Engberg-Pedersen, T. 70n17 187–203, 237
Epictetus 83–4 divine 17–18, 94–5, 194–5, 234
Epicureanism 7, 78–82, 276, 280 highest good 7–9, 59, 90, 171–3, 295
 Index of Names and Topics 333
predisposition 12, 26, 92n88, 129, 245, 259, human nature 4, 48, 99–109, 112–17,
272, 287, 291n93 128–41, 200–2, 235, 245–7, 265–6,
Good Samaritan 15, 47n37, 52, 267, 306 273–4, 291
grace 1, 11–16, 89, 93n92, 129, 159, 162–3, Hume, D. 10n27, 122, 126n86, 137, 170n103,
168–9, 214, 252–3, 259, 306 269, 272, 273, 278, 290–1
Greene, J. 286–92, 299, 303, 305 humility 115, 154, 252, 254
Grice, P. 35n10 Hursthouse, R. 49n41, 64n4, 81n55, 100, 122,
Griffin, J. 66, 76, 79n50, 295n108 126, 128–41, 252
Grisez, G. 109n29 Hutcheson, F. 288
Gustafson, J. 177
ideals 46, 79, 124–8, 137–41
haecceity 98, 101n6, 103, 145–7, 149–51 image of God 27, 97, 233, 255
Haidt, J. 267–72, 290, 303, 307 intellect 33–4, 74, 95, 97, 101, 213, 225,
Hanifah, A. 186–7, 209 231n29, 233–4, 240, 244
happiness 5–11, 34, 124, 127–8, 137, 146, intuition 81, 122n76, 127n89, 129, 163,
193n34, 227, 258, 285n69, 287–8 218–19, 245, 268–71
see, eudaemonism Irwin, T. 58, 114
Hardy, T. 120–1 Israel 36, 96, 164–5, 230n28, 234, 236,
Hare, J. E. 6n10, 13n38, 14n43, 15n47, 16n50, 238, 305
21n65, 22n66, 24n76, 40n22, 59n75,
65n5, 82n57, 84n65, 114n49, 123n78, Jackson, T. 72n27
167n84, 168n92, 169n99, 227n15, Jaeger, W. 226n12
235n42, 254n121, 295n110 James, W. 11n33, 180
Hare, R. M. 5n6, 19, 32, 35–7, 64–5, 79, 80, Jesus Christ 15, 18, 29, 40, 44–7, 61, 73, 96,
99, 103, 105, 116, 117–36, 140, 147, 106, 107, 114–16, 143, 146, 148, 154,
232, 269–70, 292 156, 161, 164–6, 230, 297, 304–6
Harinck, G. 167n84 Job 10–11
Harrison, J. 137n119 Johnson, D. 282, 296n112
Hart, H. L. A. 111n39 Joynt, C. 82n57
Hauerwas, S. 115n55 al-Jubba’i 186, 195
Hayes, C. 230n28, 252 justice
heaven 14, 47, 72, 95, 104, 114n48, 146, 246 divine 10–11, 13, 111, 171, 195, 207–8, 212,
Hector, K. 33n4, 61n78 221, 231, 250
hedonism 78, 129, 289n84 human 34, 121, 127n89, 136–7, 190–2,
Hegel, G. W. F. 85, 168 200–1, 214, 238–9, 290, 296, 306
Heidegger, M. 145, 250 justification (of morality) 16–25, 29, 58, 65,
Heidelberg Catechism 41 231–2, 242, 259–60, 170–1, 275–8,
Henry, C. 167n84 287–90
Herdt, J. 98n102 al-Juwayni 201–2, 222
Herman, B. 76n43
Hierocles 84n63 al-Ka‘bi 210
Hillel 288, 305 Kain, P. 25n82, 26n84, 60n76, 172n112
Hitler, A. 265 Kant, I. 2n3, 4, 51–2, 233, 301
hive-switch 268, 305, 307 freedom 159–60, 206
Höffe, O. 25n82 kingdom of ends 8, 25, 52, 54, 59–60, 94,
holiness 24, 30, 93n92, 109, 111 156, 170–1, 246, 259, 295–6
Holy Spirit 48, 61, 144, 163 moral theology 45–6, 59–60, 61, 74, 91–4,
Homer 32, 228–9 111, 162–3, 166–73, 174–5, 181, 218,
honour 69–70, 114–15, 303 255–6, 295–6
hope 9, 14, 91, 95, 140–1, 242n71, 285n69, moral theory 7–16, 18, 25–7, 64–6, 75–7,
296, 303 85, 87–9, 96, 106–9, 147–51, 154–6,
Hopkins, G. M. 45–6, 104, 146–7 173–4, 189n16, 191n28, 244–6, 253,
Hopkins, S. 73, 93 264–7, 269, 290–2
Hourani, G. 185, 187, 190–2, 201n62, 203n66, radical evil 136, 259, 272
213n104, 217n120, 232 translation 30–1, 129, 178–9, 262
human dignity 25–31, 125, 187, 245–6 Katz, L. 274n34
334 Index of Names and Topics
Kelemen, D. 282, 285 Merkel, R. 26n82
Kelsey, D. 29, 157 Micklethwait, J. viin1, 288n82
Kenny, A. 74n33 Mill, J. S. 9, 285n69, 288–9, 291–2
Kent, B. 159n57 Milton, J. 44, 140, 181
Kholeif, F. 186n4 mixture (of good and bad) 1, 128, 138, 147,
Kierkegaard, S. 24n74, 110n30, 145, 146, 245n84, 262, 275, 279–80, 287,
155–6, 160–3, 254–5 289–90, 299, 304
Kitcher, P. 292–300, 303, 305 Mohammed 195–7, 215–16, 219
knowledge 10, 75, 90, 174, 177, 189–90, Moore, G. E. 23, 29n96, 99–100, 116, 131, 188n8
201n61, 207–8, 213–14, 217–19, Moses 17, 38, 73, 196, 216, 229, 232
238–41, 248, 279, 302n131 motivation 15n46, 58, 63, 65–6, 67–75, 77,
Korsgaard, C. 16–17, 26, 59, 85–6, 187, 83–4, 95–7, 118–21, 159, 192–3, 276,
189–90 287, 290, 293, 296, 301
Kropotkin, P. 135 murder 36, 102, 106–8, 212, 227, 231n28,
237, 285
Lambert, F. 297 Murdock, G. 294
Lansing, S. 307 Murphy, M. 51n46, 53–4, 57, 102, 110n33,
Lawrence, G. 117n61 111n39, 112, 250n105
Leibnitz, G. W. 58, 93, 150 Mu‘tazilites 185–7, 195–7, 200, 202–3,
Leopold, A. 104n17 215–16, 223, 306
Levinas, E. 246 see, ‘Abd al-Jabbar
Lewis, C. S. 304
Lipscomb, B. 58n69 Nagel, T. 86n73, 131n99
Lipscomb, D. 279n47, 280 natural law 2, 4, 17, 40n24, 100, 109–10, 180,
Lombard 164n75 192, 226–7, 231n28, 236, 238, 242–3,
love 246–7, 250–1, 254, 257
divine, see God as lover naturalism 99–100, 281, 292
for God 17–18, 23, 29–30, 39, 51, 73, 74, nature 6, 8, 26, 83, 91–3, 111, 130, 159, 169,
86, 94–5, 110, 146, 148, 160–1, 242n, 226–8, 239–40, 257–9, 265, 289–90
252–4, 259–60, 302 see, human nature
for others 65, 70, 76, 78, 80, 86–7, 110, 115, Newman, J. H. 177n125
147, 149–51, 160–1, 164, 255, 277, 305–6 Nietzsche, F. 113, 116, 161
Lucas, J. 150 non-cognitivism 118, 232–3, 235
Luther, M. 12, 14 Novak, D. 59, 230, 232n33, 235–47, 248,
lying 43, 91, 108–9, 190–2, 196, 208, 214 256–8, 305
Novalis 168
MacDonald, S. 99n1
MacIntyre, A. 85n68, 227n17 obedience 15, 17–19, 31, 37–43, 44, 51–2,
Mackie, J. L. 25n77, 59, 123, 273–5, 275–8, 56, 58, 143, 178–80, 204, 232, 255,
281n53, 297, 306 275, 312
Maimonides 18n55, 27n86, 55n60, 97–8, 175, obligation 5–7, 15, 17–18, 19–20, 23–4,
212n99, 224–35, 237, 239–43, 243–5 25–30, 37–8, 46, 54–5, 75, 86, 109–12,
Marion, J. L. 250n99 131, 136, 140, 149, 170, 179, 181,
Malik, C. 187 189–90, 194, 204, 213, 230, 231, 242,
al-Maturidi 14, 55, 197–203, 209–12, 305–6
217–22, 235 Ockham 18n52, 32n1, 43, 56n61
McCormack, B. 13 O’Donovan, O. 177n127
McDowell, J. 50, 122, 123n80, 227n17 Otto, R. 11, 301
McKenny, G. 29n94, 152n30, 155n41,
156n43, 159, 161n65, 177n129, Paley, W. 9, 288
179n133, 302n131 Palmer, F. R. 37n15, 42n25
McMahan, J. 26n82 Parfit, D. 79n50, 85n67, 122n76, 123n81,
Meilaender, G. 115n56 126n86
Mellema, G. 188n11 particularity 23, 143–57, 177, 251
Mendelssohn, M. 233 Paton, H.J. 147, 149
Mendenhall, G. 39n19 perfect duties 45–6, 58, 109, 189n16
 Index of Names and Topics 335
perfection, moral 8n14, 20, 34, 45, 69, 87–9, reward 9, 54n56, 212–14, 217, 242, 285,
95, 98, 101, 108, 115, 145–6, 159, 167, 295–6
229, 233, 244 rights 28, 54, 138
personhood 6–7, 18, 25–7, 250 Rist, J. 251n107
Pieper, J. 115 Ritchie, A. 123n81, 274n34
pietism 12, 14, 16, 162 Ritschl, A. 167
Pinches. C. 115n55 Roberts, R. 300
Plantinga, A. 22n70, 279n48, 283n60, 284n61 Rorty, R. 236–7, 300n126
Plato 24, 68n13, 69, 90, 113, 117, 130, 168, Rosenzweig, F. 241, 247–60, 302
229, 234, 240, 268–9 Ross, W. D. 189n15, 190–1, 232
pleasure 21, 23, 66, 69, 76, 78–82, 132, 136, Rudolph, U. 184
191, 199, 217 Ruse, M. 264–5, 275–8, 281
political relations 38, 68–9, 90–1, 94, 137, 237,
272, 295, 298, 300n126 Saadiah 243–4
Pollock, B. 248n92 Sahih Muslim 306
Porter, J. 80–94, 161n, 193, 250 sanctification 46, 76, 129, 182, 257
positions of judgement 35–6, 147–51, 155 Satan 74, 140, 181, 191
power Schleiermacher, F. 144
divine 52, 55, 60, 127, 158, 166n82, 170, Scotus 49n40, 55, 70n18, 92, 94–5, 191n25,
181, 196, 249–10, 260, 301 217n121, 221–2
human 50, 114–15, 138, 196–7, 204–5, the affections 34, 71, 72–5, 77, 88–9, 93,
207–8, 209–10, 221, 266, 290 96–7, 159, 289, 301
praise 22, 76, 189–91, 204–5, 209, 234 co-lovers 23, 29, 39, 54, 58, 79, 111n34,
prayer 158, 173–4, 178, 191, 213–14 139, 146
prescriptions 19, 23, 25, 32, 33–49, 105–6, natural law 17–18, 23, 100–9, 297
118, 154, 173–6, 305 particularity 98, 145–7
prescriptive realism 22, 50, 123–4, 273 will 34, 159, 164n75, 182, 206
promising 6–7, 19–20, 39, 54–5, 135–6, 140 scripture 14, 40n24, 164, 187n7, 201–2, 213,
providence 5–11, 66–7, 108, 141, 173, 207, 217, 231, 238, 268, 296–7
258–9, 268, 287, 291–2 Searle, J. 19, 135
providential proximity 52, 82, 156n, 308 self-interest 9, 94, 129, 276, 286, 291, 301
prudence 125, 129, 140, 292 Sextus Empiricus 295
Pseudo-Dionysius 108n27 al-Shaybani 203n66, 217
Pufendorf, S. 46, 56n62, 57–9 Shelley, P. B. 95
punishment 39, 44, 54, 170n104, 195, 214, Sidgwick, H. 9, 67, 78, 81–2, 84–5, 129, 289,
232, 242, 282, 293–6, 302–3 291, 296n113
Putnam, R. 307 Siep, L. 25n82
Pythagoras 69 Singer, P. 84–5, 126n86
slavery 130, 148, 280, 296–7
queerness (argument from) 274–5, 278 Smith, M. 33n5
Quinn, P. 18, 19n56 Socrates 17, 24, 41, 90, 145, 166, 229n22, 297
Sosis, R. 307
Raleigh, W. 38n18 Spencer, H. 264–6
reason 4n5, 34, 60, 67, 69, 74–5, 83–6, 89, Spener, P. J. 12
110n32, 150, 185, 190–1, 196, Spinoza, B. 5n8, 92
198–200, 209–10, 213–22, 226, 230, Stoicism 7, 67, 83–6, 96, 168n93
240–1, 256, 268–72, 290–1 Strauss, L. 224–5
reasons for action 50–3, 54, 137 Street, S. 278–81
practical reason 9, 10, 11, 13, 93n92, 108, Stump, E. 21–2, 130
128, 167–8, 171–3, 178, 244, 296 Sturgeon, N. 123
Regan, T. 25n82 Suarez, F. 9n20, 34, 51n45, 111
regulative principles 167, 170–1, 179 supervenience 21n65, 22, 190
relativity (argument from) 116, 271, 274, 277–8
revelation 11–12, 30, 32, 61, 115–17, 167–69, Tanner, K. 157–8
175, 190, 213–22, 230, 235, 237, 240–3, Taylor, C. 108n25
251–8, 283 teleology 131–2, 244–5, 258, 282
336 Index of Names and Topics
Ten Commandments 17, 36, 41, 43, 47, 100, Walzer, M. 90n83
105–6, 110, 148, 152, 304 Warren, A. 37n15
theft 15, 75, 101–2, 106, 227, 247, 285 Warren, M. A. 26n82
theodicy 10–11, 207, 212 Watson, G. 56n62
Thompson, M. 130, 134 wealth 47, 82n57, 83, 114, 137, 266,
Tillich, P. 181 303, 304
torture 136, 194, 205, 312 Webster, J. 157n49
total depravity 291 Weinfeld, M. 39n19
tragedy 43, 286–8, 303 Weithman, P. 28
will
universalism 13–14 divine 4n4, 8, 32, 37, 43, 52, 55, 57, 59,
universality 6, 35–6, 37, 59, 77, 85, 101, 108, 72–3, 89, 101–2, 163–4, 178, 180–1,
125–7, 147–51, 151–7, 197, 229, 236, 187, 202, 204, 206, 210–12, 221, 229,
239, 242n72, 254, 256, 261, 264–6, 230–7, 259, 298
269–70, 273–4, 282–3, 290, 305–6 human 6, 12–13, 18, 22, 34, 52, 60, 72–5,
utilitarianism 9, 66, 79, 82, 125–6, 140, 76, 87, 95, 100–1, 121, 140, 143–4, 152,
265–6, 287–92 160, 168, 171, 206, 245, 271, 274, 301
Williams, B. 81, 116, 138n123
van Inwagen, P. 131n100, 284nn62–3 Wilson, E. O. 268n15, 275n37
Van Til, C. 166–8 Winter, T. 187n6
Vasalou, S. 185n3, 188n10, 189n15, 190n17, Wittgenstein, L. 56, 121, 137, 284n65, 287
193n35 Wolf, S. 86n73
Velleman, D. 108n25 Wolterstorff, N. 20, 27n87, 28, 54, 62, 69n16,
virtue 7–8, 11, 46n34, 49n41, 63–5, 69–71, 71n21, 104, 168n93, 298n120
89–91, 114–15, 124, 126, 135–6, 139, Wood, A. 5n9, 25n82, 27n87, 85n68,
154n35, 159, 193, 210, 226–9, 253, 272, 168n92
295–6, 305
Volf, M. 237–8, 304n135, 306nn136–7 Zagzebski, L. 123n79