David O’Connor.

Hume on Religion
H. O. Mounce
Hume Studies Volume XXVIII, Number 2 (November, 2002) 309-313.

Your use of the HUME STUDIES archive indicates your acceptance of HUME
STUDIES’ Terms and Conditions of Use, available at
http://www.humesociety.org/hs/about/terms.html.

HUME STUDIES’ Terms and Conditions of Use provides, in part, that unless you have
obtained prior permission, you may not download an entire issue of a journal or multiple
copies of articles, and you may use content in the HUME STUDIES archive only for your
personal, non-commercial use.

Each copy of any part of a HUME STUDIES transmission must contain the same
copyright notice that appears on the screen or printed page of such transmission.
For more information on HUME STUDIES contact
humestudies-info@humesociety.org

http://www.humesociety.org/hs/
Hume Studies
Volume 28, Number 2, November 2002, pp. 309–313

Book Reviews

DAVID O’CONNOR. Hume on Religion. London: Routledge, 2001, Pp. xvi +

227. ISBN 0415201942, cloth, $75.00; ISBN 0415201950, paper, $15.95.

This book is a commentary on the Dialogues concerning Natural Religion,

Hume’s most famous work in the field of religion. O’Connor is a fine exposi­
tor, commanding a clear and readable style. The Dialogues is covered in detail,
so that students will gain a good sense of the structure. The weakness lies in
O’Connor’s attitude as it inevitably appears in his exposition. I do not mean
that this is wild or eccentric. Quite the contrary, it is the one which during the
last fifty years has become conventional, reflecting the empiricist tradition one
associates with Ayer and Flew. For example, O’Connor hardly questions Hume’s
distinction between the a priori and the a posteriori, implies indeed that he puts
it to devastating use, though it is now widely criticized even by philosophers
otherwise sympathetic to Hume. One wonders whether students need another
version, however well expressed, of views already so familiar. If their intelli­
gences are to be sharpened, they surely need to encounter some serious
criticism not simply of Hume’s opponents but of Hume himself.
Having that in mind, let us consider what O’Connor calls “evidentialism,”
an assumption he treats as lying behind Hume’s thought (9). This he explains
as the view that we should proportion our beliefs to the evidence. The trouble
with this formulation, however, is that it is ambiguous, being platitudinous in

Volume 28, Number 2, November 2002

310 BOOK REVIEWS

one sense and dubious in the other. It is platitudinous that one is not entitled
to ignore evidence where that is required. No one has denied it. What has been
denied is that every belief requires evidence in the implied sense. The reason is
that it leads to an infinite regress. If every belief requires evidence, any belief
advanced as evidence itself requires evidence and so on ad infinitum. In short,
no belief can ever be justified. If any belief is justified, some beliefs must be
fundamental, in the sense of not resting on further evidence. It is obvious, how-
ever, that fundamental beliefs are various and sometimes conflict. Such a
conflict cannot be settled by reference to other beliefs more fundamental, for
there are no such beliefs. It does not follow that all reasoning is at an end. But
the reasoning must be dialectical rather than foundational. One can proceed
only by considering whether there are difficulties or inconsistencies in one
position which are avoided in the other, or, supposing difficulties in both,
which are the greater.
The relevance of the above points to the philosophy of religion should be
obvious. The religious and the non-religious do not hold differing beliefs within
a common system. They hold different views of the world. The difference be-
tween them is itself an instance of fundamental conflict. It follows that a serious
study of religion is possible only to those who can occupy in thought both a
religious and a non-religious view. It is impossible for those who cannot imag­
ine that the other side may have an answer to difficulties that arise on their
own. No one can be delivered from a false system by further reasoning based
on the same system but only by having the imagination to conceive how that
system may appear when viewed from another. The procedure, in short, must
be dialectical not foundational. That, however, is very far from the procedure
we find, during the last fifty years, in the philosophy of religion. The proce­
dure, rather, is to consider whether there are difficulties in religion and then
to dismiss it when they appear substantial. It is not thought necessary to con­
sider whether on the other side there are difficulties as great or even greater.
Indeed worse; the reasoning is precisely foundational not dialectical, the foun­
dational view being the non-religious one. In short, religion is justified only if
it can be deduced from the opposing view.
A classic example of this fallacious procedure may be found in Hume’s
“Essay on Miracles.” His aim is to show that no miracle can serve as the basis
for a religion. His argument is that a miracle is contrary to common experi­
ence. But it is common experience which determines what is reasonable to
believe. What, however, if there is strong testimony that an event has occurred
which would be deemed improbable when judged by common experience?
Hume acknowledges that this is conceivable but argues that we still should

Hume Studies
 BOOK R EVIEWS 311

not accept the miracle. That is because the weight of the testimony is at least
equally balanced by the improbability of the event itself.
Notice that this argument works only if at no point are we allowed to ap­
peal to religious considerations themselves. Suppose, as is conceivable, that
we are torn between the weight of testimony and the improbability of the event.
On the non-religious side, there is no clear way to resolve the conflict. But on
the religious side, there is no conflict to resolve. If miracles are real, precisely
what we should expect on some occasions is strong testimony in favor of an
event which would be deemed improbable when judged simply by common
experience. On such an occasion, the assumption of a miracle would be rea­
sonable not simply for the religious but for the hitherto non-religious. Indeed
so fallacious is Hume’s procedure that it would exclude not simply miracles
but any fundamental advance in the sciences. Take the conflict in astronomy
between the Ptolemaic and the Copernican systems. The central claim of the
Copernican system, when judged by the Ptolemaic, is wildly improbable. Con­
sequently, for one who holds the Ptolemaic, whatever is advanced in favour of
the Copernican will always be at least equally balanced by the improbability
of its central claim. Given Hume’s procedure, one who holds the Ptolemaic
will never have reason to accept the Copernican.
Procedures comparably fallacious run through the Dialogues. For Hume,
reasoning is exhausted by its deductive and inductive forms. In the former,
we may infer only what is already contained in our premises; in the latter, we
generalize from experience. The inductive in fact is only a looser form of the
deductive. As in the deductive we can infer only what is contained in our
premises, so in the inductive, we can infer only what is contained in our ex­
perience. Thus, we can infer rain from dark clouds only because we have
already experienced instances of dark clouds and rain. With these criteria in
place, it is unnecessary even to consider any argument in natural theology.
We may reject such an argument not because we have found it fallacious,
when considered independently, but because it is excluded beforehand by
our criteria. The arguments in natural theology seek to show that this world
does not explain itself but may be explained only by what transcends it. It is
therefore essential to natural theology that there should be principles of ex-
planation which transcend their data. These are excluded by Hume’s criteria.
For him, an explanation must be inferred from what its data already contain.
The inadequacy of these criteria will be evident to those who apply them
impartially. Thus applied, for example, they will lay waste to the theoretical
sciences. No theory in the sciences can simply be derived from its data,
whether by strict deduction or generalization. Every theory thus transcends

Volume 28, Number 2, November 2002

312 BOOK REVIEWS

its data. Throughout the theoretical sciences it is only so far as we transcend

in some measure what we seek to explain that we can in fact explain it.
The chief fault in O’Connor’s book is that he does not examine the cen­
tral weaknesses in Hume’s approach. But there are others. In the main, they
seem to me to reflect the fallacies or prejudices of our secular culture. Here
are some examples.
On page 7, the Scottish Calvinists are inevitably characterized as exhib­
iting “a grim joylessness.” The Calvinists no doubt had their faults, though
in that they are hardly exceptional. But they were neither trivial nor foolish.
Calvinism itself is one of the great formative movements in European his-
tory. For example, the Scottish tradition in philosophy, which is of the highest
distinction, is evidently indebted to it. Indeed Hume’s view of reason, as we
find it in the Treatise, it itself indebted to Calvinism.
On page 21, we are told that Darwin’s theory of evolution by natural se­
lection shows that nature is self-ordering. Perhaps this is loosely expressed.
In any case, as it stands it is plainly false. Natural selection cannot explain
natural order, for without natural order there would be nothing to select. It
may explain why it is a creature with such and such features rather than some
other which has survived. But it cannot be explained why it had those fea­
tures in the first place. We may illustrate the point. The explanation why
basketball players tend to be over six feet tall is that in basketball one tends
to select only players of that height. But the explanation for their being that
height is not that it gets them selected for basketball. They must already be
that height before they get selected. We may note that although students who
take philosophy invariably treat Darwin’s theory as infallible truth they rarely
have a precise idea of what it involves. The idea that natural selection ex-
plains natural order is in fact one of their commonest delusions.
In Hume’s work there figures a distinction between God’s existence,
which can be known with certainty, and his nature, which is unknowable.
The distinction is a standard one in traditional theology. O’Connor says it
is incoherent, since we cannot know an object exists unless we can charac­
terize it. But it is not incoherent, since an object may be characterized other
than through its nature. For example, we may characterize it through its
relations or consequences. The point will be evident to those who have stud­
ied traditional theology but it should be equally evident to those who have
studied modern physics. For example, the electron is characterized through
various models which were we to apply them to its nature would involve us
in contradiction.
Philosophy has been valued for its ability to criticize conventional pi­
eties or orthodoxies. In our age, the orthodoxy, at least in academic circles, is

Hume Studies
 BOOK R EVIEWS 313

certainly not religious. We find it rather in scientism and secularism, those

being effortlessly acquired by so many of our colleagues. In that respect stu­
dents will do well to study Hume’s Dialogues, so long as they do it in a critical
spirit. For it is this work as much as any in the last fifty years which has been
used to reinforce that orthodoxy.

H. O. MOUNCE
Department of Philosophy
University of Wales Swansea
Singleton Park
Swansea SA2 8PP, Wales

Volume 28, Number 2, November 2002