Swarthmore College

Works

Philosophy Faculty Works Philosophy

11-1-1981

Review Of "The Nature Of Law" By A. Watson

Hans Oberdiek
Swarthmore College, hoberdi1@swarthmore.edu

Follow this and additional works at: https://works.swarthmore.edu/fac-philosophy

Part of the Philosophy Commons

Let us know how access to these works benefits you

Recommended Citation
Hans Oberdiek. (1981). "Review Of "The Nature Of Law" By A. Watson". University Of Pennsylvania Law
Review. Volume 130, Issue 1. 229-258. DOI: 10.2307/3311815
https://works.swarthmore.edu/fac-philosophy/123

This work is brought to you for free by Swarthmore College Libraries' Works. It has been accepted for inclusion in
Philosophy Faculty Works by an authorized administrator of Works. For more information, please contact
myworks@swarthmore.edu.
1981]

THE NATURE OF LAW. By Alan Watson. Edinburgh: Edin-

burgh University Press, 1977. Pp. 148. Price $8.50.
HANS 0BERDIEK t
What is law? Pontius Pilate had little time for a similar ques-
tion about truth, and it is unlikely that busy attorneys will find
much time for philosophical disquistions concerning the nature of
law. Nonetheless, those who ignore such questions do so to their
own detriment, for one's conception of law will certainly affect
one's understanding of what one is doing, whether it is worth doing,
and what one is becoming in the process. Only if one believes
that theory is irrelevant to practice can one sustain a dismissive atti-
tude towards inquiries into the nature of law. And such a belief
would itself embody claim& regarding the relationship of thought
and action that would require justification.
Despite its importance, legal philosophy is not treated with
much seriousness in most American schools of law. Jurisprudential
considerations enter the curriculum systematically primarily as ad-
juncts to other subjects. Courses devoted to philosophy of law
and jurisprudence are, when available, almost always optional, and
are often taught by professors whose primary professional interests
lie elsewhere. And for those who are convinced that legal philo-
sophy is an essential part of a legal education, there is the additional
problem of accessibility: so much that has been written recently is
arcane, obscure, incestuous, or polemical that even an eager stu-
dent hardly knows where to begin.
Alan Watson's The Nature of Law 1 provides a provocative
starting place: it is short, to the point, and presupposes only a basic
knowledge of alternative legal philosophies. The book's thesis is
emphatically presented, and will both attract and repel those ac-
quainted with modern philosophy of law. Professor Watson brings
impressive credentials to his study. He is an authority on Roman
law,2 and has also explored the ways in which custom becomes law,
how laws are exported from one legal system to another, 3 and how

t Professor of Philosophy, Swarthmore College. B.S. 1959, Ph.D. 1965, Univer-

sity of Wisconsin.
IA. WATSON, THE NATURE OF LAw (1977).
2 See A. WATSON, LAw MAKING IN THE LATER ROMAN REPUBLIC (1974); A.
WATSON, RoMAN PllxvATE LAw (1971); A. WATSON, THE LAw OF THE ANCIENT
ROMANS ( 1970 ).
3A. WATSON, LEGAL TRANSPLANTS (1974).

(229)

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
230 UNNERSITY OF PENNSYLVANIA LAW REVIEW lVoL 130:229

legal change occurs generally.• He is undoubtedly familiar with the

leading contemporary legal theories on both sides of the Atlantic.
The bulk of The Nature of Law is devoted to Watson's elabora-
tion and defense of a single, complex proposition:
The argument of this book is that the distinguishing
and sole necessary feature of law is the availability of an
institutionalised process, which has the essential function
of resolving actual or potential disputes, by means of a de-
cision, with the specific object of inhibiting further un-
regulated conflict. 6
Along the way, ·watson examines: the essential function of legal
processes; the essence, typical attributes, and particular virtues of
law; the relationship between law, order, and authority; the role
of legal rules, and respect and obedience for law. The Nature of
Law therefore represents a very ambitious undertaking, for its cen-
tral aim is no less than to supplant prevalent legal theories with a
novel approach that retains what is valuable in its rivals without
succumbing to their one-sidedness. For the sake of clarity, I will
begin with a sketch of Watson's general position before examining
his approach, assumptions, and argument in detail.
The only essential function of the legal process, according to
Watson, is to inhibit potential conflict by settling actual disputes. 6
More specifically, this essential function has three components: law
serves to "institutionalise disputes, validate the decisions given in
the process and inhibit unregulated conflict." 7 In Watson's view,
because order-rather than freedom, justice, or morality-is the es-
sence of law, legal rules do not inevitably reflect the political, social,
and economic needs and desires of the society as a whole or its rul-
ing elite. 8 The distinguishing and sole necessary feature of law,
then, is an institutionalized process by which a dispute can be re-
solved. More basically, it is the possibility or threat of recourse
to that process that furthers the essential function of inhibiting
conflict. 9
In Watson's scheme, legal rules are needed for two reasons:
to activate the process, and to validate-but not necessarily de-
termine-the resulting decisions. 10 Although Watson argues that
4 A. WATSON, SOCIETY AND LEGAL CHANGE ( 1977).
II A. WATSON, supra note 1, at i.
6 Id. 13. For a discussion of how the legal process is involved in resolving
potential disputes, see id. 18-19.
1 Id. 71.
s Id. i.
9 Jd. 45.
10 Id. L

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981) BOOK REVIEWS 231

no other features or functions of law and the legal process are

essential, he does recognize that there are several typical attributes
of law, and especially of the laws of mature legal systems. The most
important of these typical attributes are that "they form part of a
system that is more or less all-embracing, they are created by a
recognisable person or body that has the power to make law, they
are authoritative and authoritarian, they are backed by regulated
sanctions, and they receive obedience." 11 Watson argues that the
prime virtue of law is to offer a readily available dispute-resolution
process, 12 and that certain of these attributes further that end: "it
is a particular virtue for law to be formally enacted and published,
for its application to be within the financial and other means of the
parties affected, and for the legal rules to be clear." 13 Finally, to
fulfill its essential function, law must have recognized authority
in resolving disputes. Such authority derives both from the po-
tential use of force and from respect for the law. 14
Although he recognizes his intellectual debts to competing
theories, Watson distinguishes his theory from its rivals. Contrary
to natural law theorists, Watson declares that law as it is should
be sharply separated from law as it ought to be. Watson accepts
that law and morality interact, but not as classical or contemporary
proponents of natural law claim. Unlike legal positivists, he at-
taches the status of law to both international and primitive law
because, despite their lack of some or even all of the attributes and
virtues of law, both can and often do possess its sole necessary fea-
ture. And contrary to Marxists and proponents of other theories
viewing law as a means of social control, Watson maintains that
law often bears scant relation to the interests and power of the
ruling elite. In large part the law is firmly in the hands of lawyers,
who have their own interests and power!
Because Watson situates his own theory in the midst of its
three principal rivals, I shall consider the strengths and weaknesses
of each, as Watson sees them, before examining the alternative
that he offers.
I. THREE THEORIES OF LAW

A. Natural Law
Of the three theories of law that Watson considers, natural
law has by far the oldest and most distinguished pedigree, and
11 Id. 48.
12 Id. 61.
lSJd.
14 Jd. 71.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
232 UNNERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

Watson traces it to its ancient Greek roots. Cicero provided the

first systematic statement of this theory:
True law is right reason in agreement with nature; it
is of universal application, unchanging and everlasting;
it summons to duty by its commands, and averts from
wrongdoing by its prohibitions. And it does not lay its
commands or prohibitions upon good men in vain, though
neither have any effect on the wicked. It is a sin to try
to to [sic] alter this law, nor is it allowable to attempt to
repeal any part of it, and it is impossible to abolish it en-
tirely. We cannot be freed from its obligations by senate
or people, and we need not look outside ourselves for an
expounder or interpreter of it. And there will not be dif-
ferent laws at Rome and at Athens, or different laws now
and in the future, but one eternal and unchangeable law
will be valid for all nations and all times, and there will
be one master and ruler, that is, God, over us all, for he is
the author of this law, its promulgator, and its enforcing
judge. 16

Aquinas developed the theoretical foundations of the theory and

pointed out its chief implications; indeed, natural law in its
Thomistic form remains the principal theory of law and morals of
the Roman Catholic Church. Because of its centrality to Roman
Catholic thought, natural law continues to influence both juris-
prudential reflection and legal practice throughout the world.
Any ship that has picked up philosophical baggage in as many
ports as natural law has will inevitably have run a shifting course.
Yet the general direction of this theory has remained remarkably
constant. Three principal theses-each expressed in the passage
quoted from Cicero-can be identified. First, natural law is uni-
versal and unchanging: everyone, everywhere, is bound by it at all
times. Second, natural law is "higher" law; it is superior to man-
made, positive law because it derives from a superior source: na-
ture, reason, or God. None of this would be of much import,
however, unless the. content of natural law could be known. So
the third thesis is that natural law is discoverable by reason, prop-
erly educated.
From these theses, natural law theorists have drawn several
highly significant corollaries. To the old question whether law is
discovered or made, the proponent of natural law has an un-

11> Clcmm, DE RE PuBLICA DE LEcmus 211 ( C. Keyes trans. 1948).

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 233

equivocal answer: the first principles and chief implications of law

are discovered by reason, not made by fiat. This corollary, how-
ever, does not mean that there is no room for legislative and judi-
cial innovation and discretion. Indeed, in every system of natural
law, and especially in Aquinas's, great legislative responsibility is
left to the good sense of men. Categories of law reserved to human
decision include "rules of the road," offences that are merely mala
prohibita (such as failure to apply for a vendor's license), and sig-
nificant determinations of morally sanctioned institutions (for ex-
ample, specific rules of inheritance).
Most natural law theorists also maintain that grossly unjust
law is not law: lex iniusta non est lex. And if a statute, decree, or
regulation does not represent true law, it is not legally binding.
A morally repugnant statute is not bad law; it is not "law" at all.
Depending on the kind and degree of injustice perpetrated under
the color of law, one may have a moral obligation to obey the
statute, but certainly not a legal obligation. According to Aquinas,
for example, one may have a moral obligation to obey an unjust
decree to avoid scandal or to prevent the collapse of civil order;
even these rationales have limits, however, for one is never justi-
fied in obeying a decree requiring one to act contrary to the ex-
plicit word of God. 16
By demanding that statutes must not conflict with natural
law, natural law theorists provide a framework for powerful attacks
on morally corrupt statutes. Similarly, natural law theory provides
a basis for defending those statutes regarded as morally acceptable.
But what considerations determine whether a particular statute ac-
cords or conflicts with natural law? According to the tradition,
natural law requires legislators, judges, and rulers to promote the
common good; that is, those conditions-as discovered by practical
reason-that enable all members of society to realize their human
potential. The nature and purpose of law is therefore to give de-
tailed structure to the common good and to determine justifiable
means for its pursuit.
The attractions of natural law theory are undeniable. It pro-
vides, within the concept of law itself, a moral basis for criticizing
and defending statutes. It links law with an order that is natural,
rational, and/or divine. And it supplies a satisfying ground for
both legal obligation and the rule of law. To many, however, the
defects of natural law theory have appeared incorrigible.
16 T. AQUINAS, Summa Theologica, in 2 THE BASIC WRITINGS OF SAINT THOMAS
AQUINAS 793-95 (A. Pegis ed. 1945).

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
234 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol.130:229

Watson's criticisms reflect two of the recurring major problems

with natural law:
How can a resolution that is passed in the same way as law
and is enforced in the same way as law, and is treated by
the state as law, reasonably be claimed not to be law?
Why can one not admit it is law but bad law? Secondly,
to determine whether a resolution is or is not law how can
one appeal to a standard outside the political structure
that passed the resolution? And how can one agree on
the outside standard to which appeal is to be made? 17

Watson, like the positivists, 18 insists on drawing a sharp and deep

distinction between law as it is and law as it ought to be. As re-
vealed in his first question, Watson criticizes natural law theory for
failing to identify and understand all law. His approach is that
of an anthropologist who casts about for a neutral, descriptive
theory of law, rather than a normative, evaluative, or prescriptive
analysis.
Watson's criticisms highlight the contrasts between his perspec-
tive and that of proponents of natural law. Natural law theorists
might argue, however, that their theory was never intended to be a
descriptive tool for legal anthropologists, but rather was designed
to provide a systematic statement of what law must be if it is to
meet the requirements of practical reason. Thus, Watson and his
natural law opponents disagree partly because their respective aims
and concerns are so different. Yet there is more to the disagree-
ment than that. The advocate of natural law views law from an
inner standpoint. From this vantage, the necessary interrelation-
ships among rationality, community needs, and the interests and
aspirations of individuals become apparent. Natural law theorists
argue that to look at law only from the outside as an anthropolo-
gist does, examining the shared characteristics of law as it is, un-
avoidably distorts these critical elements. Whether this allegation
can be sustained against Watson's analysis will be considered later.
Watson's cursory and scattered remarks on natural law reveal
that he views its disdain for the "is/ought" distinction as its critical
flaw. By ignoring this distinction, he asserts, adherents of natural
law fall into confusion. Why, he asks, are natural law theorists
"so insistent that the rules and principles with which they are
11 A. WATSON, supra note 1, at 1-2.
1s Watson recognizes his intellectual debt to the positivists, although he ex-
plicitly distinguishes his theory from that of the intellectual heirs of John Austin.
See, e.g., id. 46.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 235

concerned are rules of law and not simply rules of morality"? 19

Watson suggests that part of the answer is that they believe "there
is an overriding obligation to obey law-which must mean all valid
law-which is different in character from the obligation to behave
morally." 20 Limiting their analysis of law to only that which
imposes this significant obligation, natural law theorists fail to ac-
count for much of the legal rules and processes recognized by
societies as law throughout history. Another result of their failure
to recognize this distinction is uncertainty in the determination of
what is law. Natural law scholars "appeal to a standard external to
the political structure to determine what is law and what is not." 21
But, according to Watson, there is no agreement "as to what this
external standard is, and how appeal should be made to it." 22
Watson believes that legal theory gains clarity, scope, and persua-
siveness from concentrating on man-made law.
What, then, does Watson salvage from the natural law tradi-
tion? In attempting to rationalize the natural law "belief that a
man-made rule that is immoral cannot be law," Watson suggests
that law ... is an organ of the state or group, hence dis-
content with law brings one into emotional conflict with
the moral authority of the state; that law has an inherent
tendency towards the moral; and that law is society's at-
tempt to institutionalise justice. 23

States have moral authority because they claim it for themselves;

how much of this is deceptive will vary widely. Law has an in-
herent tendency towards the moral for the simple reason that it
promotes order, and "[o]rder is itself not a neutral quality but is a
prima facie good." 24 Another reason why morality seems to be an
essential part of law is that law is society's way of institutionalizing
justice: "The state seems to hold the ring between the quarrelling
parties and fixes fair rules. In a criminal trial, the force or violence
backing the law is very obvious; and so it should be. It is a basic
part of justice in common morality that retribution follows
wrongs." 211
111 Id. 122.
20 Id. 122-23.
21 Id. 16.
22Id.
23 Id. 124.
24 Id. 125.
211 Id. 126.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
236 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

But Watson ultimately rejects natural law's thesis that law and
morality are interwined. What Watson means by justice, morality,
and kindred terms is far removed from anything natural law scholars
would recognize:
We must ... be clear as to what is here meant by jus-
tice and morality. We are not talking in terms of an ab-
stract theory known to and appreciated by only a few phi-
losophers, but of what people feel and declare to be just
whether or not they base themselves on any explicit
theory. 26

Watson thus reveals himself to be an ethical relativist: morality

is merely what people feel or declare to be just, not what is just.
Watson opts for relativism, I think, because of a deeply rooted
skepticism regarding the possibility of formulating and defending
a set of sound moral principles. Very early in The Nature of Law
he states: "it should be emphasised that no agreement [on an ex-
ternal standard of natural law] has been reached, and there seems no
way in which agreement can ever be reached." 27 Watson never
explains why agreement provides either a necessary or sufficient
criterion of soundness. Indeed, were consensus elevated to such a
role, it would prove disastrously unsuitable, for we would be
driven to skepticism even in the natural sciences, let alone in the
social sciences and legal theory. Although he fails to justify either
his relativism or his skepticism, I believe that both explain why
Watson has no alternative but to adopt the stance of the puzzled
outsider looking in. And this leads to some confusion in his own
ideas, as when he writes: "Nonetheless, it should again be stressed
that law and justice have no inevitable connection, and that in no
sense is the doing of justice an essential function of positive law,
whether of the rules of law or of the legal process." 28
What does this even mean? Is Watson saying that there is no
inevitable connection between what people feel and declare to be
law and justice, or between law and justice per se? And what does
"inevitable" mean here: a connection people recognize, or one that
obtains, whether or not recognized? A natural law adherent would
maintain that the doing of justice is and must be an essential pur-
pose of law, while recognizing that law often fails to secure this
aim. Analogously, one aim of formal logic is to formulate valid
26 Id. 125 ( emphasis added).
21 Id. 2.
28 Id. 126-27.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 237

rules of inference, although this aim is not always realized. Con-

sequently, just as there are numerous unsound rules of law, so
there remain invalid rules of inference.
Watson also mischaracterizes natural law theorists' understand-
ing of justice and morality. He poses a false choice between "an
abstract theory known to and appreciated by only a few philoso-
phers" and "what people feel or declare to be just." Contrary to
Watson's portrayal, those who maintain that moral judgments can
be justified do not demand that anyone accept any "abstract theory"
of their justification. Rather, they argue only that we recognize the
moral requiredness of certain precepts, such as that no one defraud,
calumniate, or bear false witness against another, or that no one
kill, maim, or enslave another at will. Each of these directives,
and others, needs further elaboration and defense, and that has
been provided for centuries by those untainted by any "abstract
theory." Indeed, it is the metaethics of relativism that has the
appearance of an abstract theory appreciated by only a handful of
advanced thinkers.
Many contemporary natural law theorists would regard Wat-
son's account of natural law theory as seriously misleading. For
example, John Finnis, in his defense of the doctrine, 29 denies most
of the claims Watson makes about natural law. Finnis contends
that critics who depict natural law theory, in its classical formula-
tions, as connecting "what is" to "what ought to be" are mis-
guided.30 He also argues that natural law, contrary to popular
images, does not posit that ethical norms are to be inferred from
facts, speculative principles, or metaphysical propositions about
human nature; rather, the first principles of natural law are self-
evident and indemonstrable. 31 Similarly, Finnis disagrees with
Watson's notion that proponents of natural law assert that bad law
is not law at all, 32 that there is an overriding obligation to obey
all valid law, 33 and that there is no rational, nonrevelatory way
to detemine the central precepts of natural law. 34 Finnis argues
persuasively that many of these errors could have been avoided
had proponents and critics of natural law alike concentrated on the
29 J. FINNIS, NATURAL LAW AND NATURAL RIGHTS ( 1980).
30 See, e.g., id. 55 n.6 ( criticizing R. UNGER, LAw IN MODERN SocmTY 79
(1976), and Weimeb, Law as Order, 91 ILuw. L. REv. 909,911 (1978)).
31 Id. 33.
a2 See id. 363-64.
33 See generally id. 297 -350.

M See generally id. 388-98.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
238 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

writings of Aquinas, rather than relying on those of Grotius 35 as

Watson does. But this is not the place to explore Finnis's defense;
suffice it to say that natural law appears much more plausible and
palatable than Watson's sketch would lead one to believe.

B. Law as Social Control

Watson also compares his theory of the nature of law with
those that picture law as a means of social control. Marxists, as
Watson points out, are most prominently associated with this doc-
trine, but it is not the private property of Marx, Engels, and their
heirs. Watson points, for example, to the related ideas of Friedrich
von Savigny, Rudolf von Ihering, Roscoe Pound, and A.S. Dia-
mond, and he could also have mentioned Emile Durkheim, Max
Weber, Talcott Parsons, H.A. Hayek, and Roberto Unger. ·watson
describes this shared viewpoint:
All these theories are united in the basic tenet that
law stands in a close rational or the natural (inevitable)
relationship with society, to the needs or desires of the
people or its ruling elite. Most of them, however, also
postulate or allow for some divergence of law from these
needs or desires. 36
According to these theorists, then, societal conditions, such as
economics, religion, history, and politics, will determine the legal
rules and processes.
Although Watson explicitly attacks these theorists, his own
position follows their common approach in significant respects.
Proponents of law-as-social-control theory study law in terms of
its functional relationship to a particular society. These theorists
argue that law serves to further the interests of the ruling group,
often at the expense of the rest of the population. Watson cer-
tainly recognizes the functionalist approach of this group. Unlike
natural law, he writes, "Law as Social Control is concerned not
with [the problem of the source of validity of law,] but with the
function of law or the source of the content of law." 37 In his
36See, e.g., id. 38-48.
36A. WATSON, supra note 1, at 85.
37Id. 4. This point is of critical importance. If it were firmly grasped. many
disputes would be seen for what they are: appeals to redirect legal thinking, cate-
gories, methods, and research programs. Arguments about "the nature of law"
obscure this. Had Watson himself traced the implications of his distinction, per•
haps he would have been more charitable to natural law. He might have seen,
for instance, that the choice between his own approach and that of natural law
depends in large part on one's purpose in constructing a theory of law-even a
descriptive theory.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981} BOOK REVIEWS 239

cr1t1c1sms of this theory, Watson underplays the similarity to his

own theory: in Watson's scheme, too, the essence of law is seen
largely in terms of its function. 88 Watson's emphasis, like that of
law-as-social-control theorists, is on what law does.
If Watson's theory thus has affinities to law-as-social-control
theory, why does he reject it? The main objection Watson offers
to law-as-social-control doctrine, especially in the Marxist version
but also in its other forms, is that the theory does not accurately
describe historical reality, because law does not always serve the
interests of the ruling elements in a society: "not only are many
legal rules socially neutral, but also . . . largely as the result of
inertia, a very great deal of law is and remains badly out of step
with the needs and desires of both society and its ruling elite." 39
This inertia exists, according to Watson, because "the essential
function of a rule is served whether or not the rule is the best pos-
sible provided only that it does not offend too greatly against the
ethic or the power base." 40 In Watson's view, then, Marxism and
other variants of law as social control fail as general theories be-
cause "they try to impose a pattern that does not exist. . . . The
accuracy of the theories must be tested against the social facts, not
by their intellectual neatness." 41
Watson's dispute with Marx's followers, at this level anyway,
turns on the extent to which there are legal rules that operate con-
trary to the interests of the ruling class. Watson recognizes that
Marxists are not committed to the proposition that every statute,
decree, or regulation necessarily serves, or is even intended to
serve, the interests of the ruling class. He writes: "Most [Marxist
theorists], however, also postulate or allow for some divergence of
law from (the needs or desires of the people or its ruling elite] ....
Among Marxists, Engels claimed that law seeks an internally co-
herent expression and thus fails to be a faithful reflection of eco-
nomic conditions." 42 Although Watson thus realizes that these
theories do attempt to account for some improper fit between law
and ruling class interests, he argues that the amount of divergence
that actually exists 43 undercuts the validity of the theory. 44
38 See infra text accompanying notes 55-62.
39 A. WATSON, aupra note 1, at 3-4.
40Id. 22.
41 Id. 86.
42 Id. 85-86.
43 The extent of this divergence is a factual issue that is beyond the scope of
this essay.
44 See A. WATSON, supra note I, at 87-95.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
240 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

Watson's criticism of the concept of law as social control fails

to account for modern interpretations of this theory. First, most
contemporary Marxists would regard lawyers as part of the ruling
elite. It would not be surprising, then, that many laws would favor
lawyers' interests at the expense of others in the ruling elite. Sec-
ond, most modern-day Marxists reject the early formulation of
their theory under which law was viewed as only part of the super-
structure of society, and not also part of the substructure, for law
necessarily enters into the organization of production. 45 Thus,
Watson's criticisms are more telling with respect to traditional
than modern formulations of law-as-social-control theory.

C. Legal Positivism
Natural law may be the oldest, and Marxism the most icono-
clastic, theory of law, but surely positivism is the most deeply
entrenched. Because it is so well known, only a brief description
is necessary. Watson correctly states that positivists
insist that a rule is law precisely because it is created and
accepted as such by a particular human society. On this
approach the morality or immorality of a rule or any sup-
position of divine origin is irrelevant to the question
whether the rule is or is not a legal rule. 46

Watson points out that John Austin, a seminal positivist thinker,

"defined law as the general commands of a sovereign, supported by
the threat of sanctions." 47
Watson levels three objections at positivism. First, he objects
that neither international nor "primitive" law fits easily within the
positivist concept of law. Second, he argues that custom that be-
comes law cannot properly be regarded as a command of a superior.
Third, and most importantly, Watson claims that positive law does
not reveal anything about the purpose or function of law, and is
not designed to study these important elements of the nature of
law. 48
Unfortunately, Watson's objections apply not so much to pos-
itivism as to the Austinian verision of it. This particular version
has been pretty well discredited, in large part because of the in-

45 For an elaboration of this and other defenses of a general Marxist perspec-

tive, see G. CoHEN, KARL MARX'S THEORY OF HISTORY ( 1978).
46 See A. WATSON, supra note l, at 2-3.
41Id. 51.
4BJd. 3.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 241

fluential work of the leading contemporary positivist, H.L.A.

Hart. 49 In a footnote, Watson acknowledges that Hart's attack
on the classic positivist view is "compelling," and that Watson's
second objection does not apply to Hart's own position. 110 Wat-
son's first objection does not quite hit the mark either. Hart does
not totally exclude international law or primitive law from his
understanding of the nature of law. As to international law, Hart
explores the "sources of doubt" concerning its status and con-
cludes, with Bentham, that it is "sufficiently analogous" to municipal
law to "make the lawyers' technique freely transferable from the
one to the other." 111 Hart's position as to whether "primitive
law" is indeed law is unclear. He seems, however, to think that
primitive law is defective-because it lacks "a rule of recognition" 52
-rather than nonexistent. Watson's third objection also misfires,
because it is really a complaint that positivists are not exploring
a different line of analysis. It is true that Hart and his followers,
such as Joseph Raz, 53 believe that the analysis of legal concepts
and systems divorced from historical, sociological, and moral in-
quiries is not only possible but desirable. But they would not
deny that some other approach might also be worthwhile. Indeed,
they themselves occasionally indulge in such inquiries!
Watson does acknowledge that competing theories-natural law,
positivism, and law as social control-have all contributed to the
ideas he presents in The Nature of Law. 54 The considerable in-
tellectual debt he owes to the latter tvm theories will become more
apparent as we now turn to a more detailed examination of his own
provocative theory.

II. FUNCTIONALISM

Watson proposes to grasp the nature of law by exploring its

function. He begins by distinguishing three distinct meanings of
"function." Watson first presents the classical functionalist view
of Radcliffe-Brown, according to which the function of any re-
current activity denotes "the part it plays in the social life as a
whole and therefore the contribution it makes to the maintenance

49 See, e.g., H. HART, THE CONCEPT OF LAw (1961).

r.o A. WATSON, supra note 1, at 129 n.12.
51 H. HART, supra note 49, at 231.
52 Id. 89-92.
53 See, e.g., J. RAZ, THE CoNCEPT OF A LEGAL SYSTEM ( 2d ed. 1980).
o4 A. \V ATS0N, supra note 1, at 1.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
242 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

of the structural continuity." 511 As Watson notes, this perspective

intentionally excludes any notion of purpose. Many sociologists
and anthropologists have repudiated this approach, and Watson
mentions one reason: it cannot easily account for social change. He
does not, however, explore the other theoretical problems of func-
tionalism, such as that: not all institutions have useful or beneficial
effects; not all of the effects of useful institutions are themselves
useful, and not all beneficial effects are functions.
Watson describes and discusses two other functionalist ap-
proaches. Function can be defined in terms of intended purpose.
Although Watson himself favors this construct, 56 he levels a power-
ful objection at it:
If, in discussing what an institution does, one introduces
the idea of what it is intended to do one has to grapple
with the problem that intention may be manifest or latent,
explicit or implicit, conscious or unconscious, impossible
or realistic. What people say they intend may not be what
they actually do intend, and they may actually intend what
is impossible of achievement. . . . [I]n the final analysis,
what the institution is intended to do can only be judged
by what it actually does. 57
Function can also be analyzed teleologically; that is, in terms of
laws that express the tendency of a system to maintain its goal-state.
However, Watson dismisses this notion and its near relatives with
little discussion or argument. More will be said about this ap-
proach shortly. 118
Given Watson's reservations about each of these possible mean-
ings of "function," it is surprising that, in the end, he embraces all
three. "In this book," he declares, "I will use the word 'function'
not only to mean what an institution does, but to express the idea
of what an institution is intended to do, and can do." 59 Watson
maintains that his account of law and the legal process can be
readily formulated in terms of each approach, but emphasizes that
he believes that human purpose is essential to understanding social
institutions.
Surely Watson's all-encompassing definition of function is a
mistake. First, Watson goes beyond the classical definition of func-
11 5 A. WATSON, supra note 1, at 4 ( quoting A. RADcx.IFFE-BRoWN, STRUCTURE
AND FUNCTION IN PluMrnvB SOCIETY 180 ( 1952)).
116Id. 5.
111 Id. 4-5.
118 See infra text following note 66.
119 Id. 6.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
19811 BOOK REVIEWS 243

tion by including anything that an institution does. Radcliffe-

Brown, for example, does not equate "function" with what a re-
current activity does simpliciter, but with its contribution to con-
tinuity. 80 Not everything an activity or institution does can be
counted as its functions. Further, Watson's promiscuous position
must end in confusion and obscurity, for what an institution is in-
tended to do seldom bears much resemblance to what it actually
does or contributes, if anything, let alone what it potentially could
do. Given Watson's account, consider what the "functions" of law
schools might be!
As we have seen, Watson is convinced that law has the one
essential function of resolving disputes with the specific object of
inhibiting further unregulated conflict.61 The legal process may,
and typically does, fulfill other functions as well, but this is its
sole essential function. 62 Watson reaches this conclusion by a
process of elimination: six candidates are considered, and all but
one are rejected. 68 Some possible functions of a legal process are
rejected because they do not apply to international or primitive
law; others are discarded because they do not apply to trial by com-
bat, and still others are thrown out because the characteristics of
entire classes of legal processes undercut the importance of certain
functions. The one function common to all of the legal processes
Watson considered is that of resolving disputes with the specific
object of inhibiting further unregulated conflict.
If one insists on talking in terms of essential functions, it
would be better, even for a descriptive theory, to look at the func•
tions of a mature, flourishing legal system, not of a legal system
that is defective or rudimentary in critical respects. It is not merely
that we are primarily interested in these central cases, but that
deviant or underdeveloped institutions can only be understood ade-
quately by relating them to what they may become. We would
see, I believe, that law relates to a certain kind of order, an order
that at least embodies Hart's "minimum content of natural law" 64
and perhaps also Fuller's "eight demands of the law's inner
morality." 66 It would be better yet to set aside the search for any
60 See A. RAI>cunE-BRoWN, supra note 55.
e1 See g6flef'ally supra text accompanying notes 6-14.
62 By an "essential" function Watson means "one whose failure cannot be
structured into the system or whose constant failure cannot be accepted as tolera-
ble." A. WATSON, supra note 1, at i.
68 See id. 8-13.
M See H. HART, supra note 49, at 189-95; infra note 83.
61> L. FoLLER, THE MoRALrrY OF LAw 46 (2d ed. 1969). See generally id.
33-94.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
244 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

single essential function, and look instead at the basis of functional

explanations and how they relate to ascriptions of functions. For
function statements and functional explanations are as intriguing
as they are baffling.
A function statement ascribes a function to something, as when
we say that it is a function of the heart to pump blood. Function
statements are explanations, but not necessarily functional explana-
tions, for it is not unusual for a function statement to explain what
something does without explaining why, and only the two together
comprise a functional explanation. In order for a function state-
ment to explain why, it must explain that whatever has the speci-
fied function exists in order to perform that function. A functional
explanation of the heart, for example, must specify not only the
function of the heart, but also that mammals have hearts because
they pump blood.
Functional explanations are therefore a subclass of teleological
explanations: 66 explanation not in terms of mechanistic operations,
but in terms of the result of an object's relationship to some end. 67
Unlike the teleological explanation described by Watson, however,
functional explanations need not express the tendency of a system
to maintain its goal-state-they may explain what leads the system
to change its goal-state. In addition, functional explanations differ
from classical teleological explanations in that they need not make
any reference to the conscious plans of an intelligent designer.
The functionalist perspective just described faces formidable
problems. Given that not all useful or beneficial effects, much less
all effects, can be designated as functions, how do we determine
which are and which are not functions? We have also seen that
only some function statements constitute functional explanations.
To make any progress, we must inquire, however briefly and
sketchily, into the structure of functional explanation. 68
Function statements are functional explanations only when
they answer a "why" question as well as a "what" question. But
this, though true, is not very informative. For a somewhat fuller
answer, we must introduce the twin notions of consequent explana-
tions and consequence laws. G.A. Cohen defines a consequence
law as "a universal conditional statement whose antecedent is a
66 See supra text following note 57.
67 A good introduction to the philosophical issues involved in functional and
purposive explanations may be found in Beckner, Teleology, in 8 ENCYCLOPEDIA OF
l':mLOSOPHY 88-91 ( 1967 ) .
68 For a useful bibliography on the topic of functional statements and explana-
tions, see C. CoHEN, supra note 45, at 356-61.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 245

hypothetical causal statement." 69 A consequent explanation re-

lies on consequence laws, and can take the following form: if it is
true of an object O that if it were F at ti, it would, as a result, be
E at tv then O is F at t3 •70
It might appear that consequent explanations purport to ex-
plain causes by their effects, but in this case appearances are de-
ce1vmg. This can be best grasped by considering an appropriate
example: let O = a particular society, F = "law-governed," E =
"inhibition of unregulated conflict," and t 1 = t 2 = t3 • The result-
ing consequent explanation would read as follows: If it is true of a
particular society that if that society were law-governed at a cer-
tain time it would be able to inhibit unregulated conflict at that
time, then that society was law-governed at that time. The effect-
the inhibition of conflict-does not explain the existence of the
cause-law. Rather, the existence of law is explained by a dis-
positional fact about society 0: that unregulated conflict would be
inhibited if the society were to have law. It is explanatory to cite
the effect of the law, not because its effect explains the law-which
is nonsense-but because the effect of inhibiting conflict permits us
to infer two important propositions about society 0: (a) that the
conditions of society 0 were such that law would have this inhibit-
ing effect, and (b) that these conditions caused the existence of
law. In terms of this discourse, functional explanations are simply
those consequent explanations that introduce functions (rather
than something else) as the explananda.
Functional explanations bristle with difficulties. Are disposi-
tional causes ever basic, or are they always reducible, if only in
principle, to episodic causes? Can dispositional causes ever pro-
vide either necessary or sufficient explanations? Can dispositional
causes overcome the considerable difficulties faced by all counter-
factuals? Can functional explanations ever be confirmed? Are
any functional explanations true? These questions bedevil all at-
tempts to construct functional explanations; no attempt will be
made to answer them here. Instead, we shall assume that the no-
tions of function statement and functional explanation are tolerably
clear and plausible, and try to determine Watson's position with
respect to them.
Throughout his book, Watson ascribes functions to law, the
legal process, and rules rather freely, even indiscriminately. But

69 Id. 259.
10 Id. 260-61.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
246 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:.229

this by itself does not establish that he provides us with any putative
functional explanations, because it is not clear that he is trying
to explain the existence of these items by reference to their func-
tions. Perhaps, like many sociologists and anthropologists, he be-
lieves that an explanation of the function of something is ipso facto
an explanation of its existence. If so, he is mistaken. 71 Further
evidence must be adduced to answer the "why" question.
Marx and his followers have attempted to do just that. They
have offered functional explanations of the existence of law, ex-
planations that specify dispositional facts about the tendency of
societies at certain stages of development to create legal systems
because those systems reinforce and give expression to the inter-
ests of the dominant class. It may be impossible for an explanation
along these lines, even when suitably qualified, to account for all the
facts. But it will not be refuted by showing, as Watson attempts
to do, that some laws are dysfunctional or even that some societies
-especially those in highly unusual circumstances-develop legal
systems for quite different reasons. For a Marxist explanation
need only account for the tendency of societies at certain stages of
development to develop legal systems of a particular sort. Watson's
criticism of Marxist functional explanations are, however, worth
pursuing and developing, because they do undermine a crude
Marxist view.
Is law best explained functionally? And, if so, which func-
tional explanation appears most convincing? These large questions
will not be pursued here. A great deal of careful philosophical
investigation must take place before we can be confident that
functional explanations are sound. One approach, of course, would
be to explore the presuppositions and implications of those ex-
planations, such as Marx's, that have been elaborated and defended.
But the Marxist's functional explanation is not the only candidate
in the field. Watson could, I think, develop his into another.
And still another variety explains law not as a technique of social
control, but rather as a way of structuring society to enhance human
development. On this account, law as a technique to inhibit con-
flict would be viewed as a less mature stage in the development of a
flourishing legal system. 72

71 There is no clear textual answer to the interesting question whether Watson

believes that the existence of law has a functional explanation in addition to its
functions. I suspect that he fails to address the question because he is unaware
of both the distinction it presupposes and the question's importance.
72 For my suggestions along this line, see Oberdiek, The Role of Sanctions and
Coercion in Understanding Law and Legal Systems, 21 AM. J. Jums. 71 ( 1976).

This content downloaded from

130.58.104.29 on Mon, 22 f on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 247

III. METHODOLOGICAL POSTURES

A. Watson's Methodological Essentialism

Underlying Watson's focus on the "function" of law, and to a
large extent predetermining the narrow scope of his functionalism,
is his adoption of what he calls "methodological essentialism." 73
Although he refuses to define or justify this theoretical approach,74
it is possible tentatively to reconstruct his meaning of this term.
Watson appears to endorse philosopher Karl Popper's understanding
of essentialism75-although Popper himself, along with Wittgen-
stein, opposes this method of study. 76 Popper writes that "meth-
odological essentialis[m] was founded by Aristotle, who taught that
scientific research must penetrate to the essence of things in order
to explain them." 77 Watson's book exemplifies this approach.
He searches for those features that all laws share and without which
law would not be law. Similarly, essential functions are identified.
According to Watson, "[t)he search for the essential function is ...
a search for the minimum function. That function will be what
law must perform if it is to be law." 78 Watson thus seems to reject
an alternative definition of essentialism, deriving from Locke, 79

73 A. WATSON, supra note 1, at 131 n.28.

74 Watson writes:
[S]ince Popper declares that the problem at issue ... is 'one of the oldest
and most fundamental problems of philosophy' ... and that among social
scientists there is no very energetic opposition to essentialism • . . , I
hope it will be understandable and excusable if I do not feel obliged to
justify methodological essentialism.
Id. ( citations omitted).
75 Acknowledging that his approach centers on a search for the essence of law,
and endorsing the use of a historical methodology for the study of social groups,
Watson states: "I must concede that my approach is realist or essentialist, on Pop-
per's terms." Id.
76 Watson implies that Popper opposes methodological essentialism because it
leads to historicism, id., and that Wittgenstein opposes it because of its futility, id.
6. Unfortunately, Watson does not adequately develop these views. Popper does
oppose historicism, which he defines as "an approach to the social sciences which
assumes that historical prediction is their principal aim, and which assumes that
this aim is attainable by discovering the 'rhythms' or the 'patterns', the 'laws' or
the 'trends' that underlie the evolution of history." K. POPPER, THE POVERTY OF
HisTORICISM 3 ( 3d ed. 1961) ( emphasis in original). But Popper does not oppose
the historical study of social groups. And Wittgenstein opposes a search for es-
sences because he believes it is more fruitful to look for "family resemblances."
See L. WITTGENSTEIN, I'RELIMINARY STUDIES FOR THE 'PHILOSOPHICAL INVESTIGA-
TIONS' [THE BLUE AND BROWN BooKS] (2d ed. 1969).
77 K. PoPPER, supra note 76, at 28.

78 A. WATSON, supra note 1, at 8 ( footnote omitted).

79 See J. Loco, AN EsSAY CONCERNING HUMAN UNDERSTANDING 295-317,

438-71 (P. Nidditch ed. 1975); J. MACKIE, PROBLEMS FROM LOCKE 72-106 (1976).

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
248 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

which posits that the essence of a set of objects is that which all
objects of the set share and which explains the co-occurrence of their
properties. 80
Methodological essentialism almost certainly leads to distorting
simplification because of its overemphasis on the least common
denominator. Recall the fruitless, although occasionally humorous,
quests to find a single common element possessed by all humans.
Further, Watson appears to confuse a necessary condition for the
existence of something with its essence. For example, he writes: "An
essential function is defined as one whose failure cannot be structured
into the system or whose constant failure cannot be accepted as toler-
able." 81 The presence of oxygen is a necessary condition of human
life, but is hardly its essence. Methodological essentialism, more-
over, blurs the distinction between essence and identification. A
single feature, such as a set of fingerprints, may serve to identify a
particular person without defining his essence; such a feature may
be a consequence of his nature, rather than its essence. Finally,
one may suspect that those who opt for methodological essentialism
already have chosen their quarry and are simply laying self-fulfilling
traps to snare it. Would Watson, for example, have arrived at the
same conclusion if he had considered international law a caricature
or primitive law a parody?

B. An Alternative Approach
An alternative method of examining the nature of law, sug-
gested by Wittgenstein and pioneered among legal scholars by Hart,
offers greater hope of success. This method has three principal
features: centering attention upon the practical point of law; ex-
amination of the "central case" and "focal meaning"; and selection
of a viewpoint. 82
Generally, the first of these features means that scholars con-
centrate on the practical functions of law in a society. 83 Watson
so For example, typical samples of gold are yellow and malleable, melt at a
certain temperature, and have a certain density. The essence of gold under this
approach would be the feature that all pieces of gold share and that explains both
the typical characteristics-the yellowness, the malleability, the melting point, and
the density-and the fact that in most or all samples of gold these qualities occur
together. One candidate for the essence of gold under this approach is the atomic
structure of gold. Yellowness would not be the essence of gold, even though all
gold is yellow.
81 A. WATSON, supra note I, at i (emphasis added).
82 For a detailed description of the features of this alternative approach, see
J. FINNIS, supra note 29, at 6-18.
sa Finnis describes Hart's notion of the practical point of law as follows:
Law is to be described in terms of rules for the guidance of officials and
citizens alike, not merely as a set of predictions of what officials will

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 249

does examine the practical point of law when he specifies that the
object of law is the inhibition of unregulated conflict. Yet one
may question whether he goes far enough, for law does not seem
to be just any way of inhibiting conflict. To understand just how
Watson fails to emphasize adequately the practical point of law, a
little further elaboration of Watson's argument is necessary.
To some extent, ·watson's theory does recognize that law is a
particular way of lessening disputes. In detailing his argument,
he says that law settles disputes by an institutionalized process and
validates decisions in terms of rules. Watson enumerates some par-
ticular characteristics:
[A] process will tend to have some formalities, there will
be some regularity in handling dispute situations, there
will be an absence of total arbitrariness from the procedure
or from the decision or from both, and those involved in
a process ... will treat the dispute-solving with an air of
seriousness. 84

Watson's analysis of legal rules illustrates his failure to pay

proper attention to the practical point of law. Because he does not
try to understand how legal rules guide "officials and ordinary citi-
zens," giving them "practical reason for compliance" with law, 83 he
fails to account fully for the substantive role legal rules play in the
legal process. According to Watson, legal rules are required to
validate decisions but not to determine them. A judge, for ex-
ample, must cite appropriate rules in defense of his decision and
exude an air of seriousness, but it is not legally objectionable for
him actualJy to base his decision on class interest, personal pref-
erence, or prejudice. To Watson, then, legal rules merely ra-

do. A legal system is a system in which 'secondary' rules have emerged

in OTder to remedy the defects of a pre-legal regime comprising only
'primary rules'. Law must have a minimum content of primary rules
and sanctions in OTder to ensure the survival of the society or its members
and to give them practical reason f 01' compliance with it.
Id. 7 ( footnote omitted) ( emphasis in original).
84 A. WATSON, supra note l, at 39.
A legal process, however, need not be part of a legal system, though typically
it will be. As long as there is one validating legal rule with its appropriate process,
law exists. But Watson never explains why this cannot properly be considered just
a simple system. Of a hypothetical Eskimo group that had but one legal rule,
for example, Watson asserts: "[t]hat rule can properly be regarded as law, but it
is difficult to imagine it forming or being part of a legal system." Id. 49. The
issue, however, is not one of imagination but of purpose. What is gained, or lost,
by counting the example alternatively as a case of a prelegal system, a rudimentary
legal system, or a legal process sans system? Not much.
85 J. FL"INIS, supra note 29, at 7 (emphasis in original); see supra note 83.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
250 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

tionalize the decisions that inhibit conflict; the legal principles

recited need not actually determine the outcome. 86
More careful attention to the practical point of law might have
moved Watson closer to a position like that of Joseph Raz. Raz,
like Watson, believes that law is a societal technique for settling
disputes authoritatively. But Raz emphasizes that law accomplishes
this through rules that are legally binding both on law-making and
law-applying individuals and institutions, on the one hand, and on
law-receiving individuals and institutions, on the other hand. 87 As
a result of his distorted perspective, Watson has very little to say
about the way rules guide the conduct of ordinary citizens, except
as they enable people to avoid or gain access to the legal process.
The deficiency of Watson's account of rules shows up in
another way. Although he is preoccupied with the function of
law as a method of social control, he does not satisfactorily dif-
ferentiate the ways in which this can be accomplished. Lon Fuller
provides a needed corrective. Fuller distinguishes social orderings
where rulers exercise "managerial direction" over their subjects
from societies in which legal rules are followed:
The directives issued in a managerial context are applied
by the subordinate in order to serve a purpose set by his
superior. The law-abiding citizen, on the other hand, does
not apply legal rules to serve specific ends by the law-

8 6 What leads Watson to this remarkably cynical view? I think that there are
two causes. First, there are certain processes that Watson wishes to call law,
but which he believes do not provide rational means of arriving at truthful deci-
sions. Trial by combat-especially where the disputants supply surrogater-is
such a process. Watson could, of course, deny that trial by combat is a legal
process even though it is a way of settling disputes. Because trial by combat may
serve to inhibit further unregulated conflict, Watson could have examined the prac-
tical point of this legal process by inquiring into the participants' justifications for
their actions. Proponents of trial by combat might have had a theory that con-
vinced them that the process secured just outcomes. Perhaps they thought that
God would enable the person in the right to win-nations, after all, have often
espoused a similar view.
A second reason why Watson discounts the substantive role of legal rules is
that judicial decisions often seem to bear little resemblance to the rules that validate
them; it is natural, consequently, to assume that they do not determine the deci-
sions. But, as Hart's practical approach suggests, the most critical question is how
the matter looks to those involved in the dispute, especially to those engaging in
reasoned argument. Do they regard rules as legally binding, or as mere sop to
advance the cause of their client or to make a decision acceptable? If it is the
latter, as Watson might suggest, then it is difficult to make much sense out of
claims of legal right or out of judicial deliberations and critical reactions, except as
either cunning charades or the products of vast self-delusion. This troubling implica-
tion is precisely the problem that has prevented orthodox legal realism from gain-
ing plausibility as a general theory of law.
87 See, e.g., Raz, The Institutional Nature of Law, 38 Moo. L. REv. 469
(1975).

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981} BOOK REVIEWS 251

giver, but rather follows them in the conduct of his own

affairs, the interests he is presumed to serve in following
legal rules being those of society generally. The directives
of a managerial system regulate primarily the relations
between the subordinate and his superior and only col-
laterally the relations of the subordinate with third per-
sons. The rules of a legal system, on the other hand,
normally serve the primary purpose of setting the citizen's
relations with other citizens and only in a collateral man-
ner his relations with the seat of authority from which the
rules proceed. 88

Legal rules are distinguished from managerial directives, then, by

their generality and, more importantly, by the nature and scope
of the obligation to obey. Following this distinction, "primitive"
law will often be excluded from the sphere of law, because it is
largely managerial. Within this narrowed range, the role of legal
rules may be much more substantial than Watson describes. It is
only because Watson fails to pay proper attention to the practical
point of law, including the obligations of citizens, that he can
argue that rules do not determine legal outcomes. In Fuller's
analysis, legal rules are inextricably connected with determining
or reaching decisions, not merely with making them acceptable.
If legal rules do determine decisions, then in pursuing the
practical point of law one must single out which features are sig-
nificant for a descriptive theory of the nature or function of law.
And this, in turn, involves identifying the "focal meaning" of
law. The device of focal meaning, John Finnis explains,
involves a conscious departure from the assumption
that descriptive or explanatory terms must be employed
by the theorist in such a way that they extend, straight-
forwardly and in the same sense, to all the states of affairs
which could reasonably, in non-theoretical discourse, be
'called "law" ', however undeveloped those states of affairs
may be, and however little those states of affairs may mani-
fest any concern of their authors . . . to differentiate be-
tween law and force, law and morality, law and custom,
law and politics, law and absolute discretion, or law and
anything else. 89

Finnis's remarks are directed towards Hans Kelsen, but they apply
with equal force to Watson. We can follow Finnis in calling what
88 L. FULLER, supra note 65, at 207-08 ( emphasis in original).
~9 J. FINNIS, Stlp,ra note 29, at 9-10 ( emphasis in original).

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
252 UNNERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

is referred to by the concept of law in its focal meaning the "cen-

tral cases." These will be the legal systems of nations that strive,
albeit imperfectly, to operate according to the rule of law. This
distinction will enable us to differentiate
the mature from the undeveloped in human affairs, the
sophisticated from the primitive, the fluorishing from the
corrupt, the fine specimen from the deviant case, the
'straightforwardly', 'simply speaking' . . . , and 'without
qualification' from the 'in a sense', 'in a manner of speak-
ing', and 'in a way' ... -but all without ignoring or ban-
ishing to another discipline the undeveloped, pnm1t1ve,
corrupt, deviant or other 'qualified sense' or 'extended
sense' instances of the subject-matter .... 90

But how should we actually select the significant features?

It is not enough to spurn Watson's least-common-denominator ap-
proach; something better must be put in its place. Obviously, the
viewpoint from which to select these features is that which makes
the most sense in determining the practical point of law. And
here there are two longtime rival approaches, each of which sees
law as a technique of control pertaining to order. Watson be-
longs in the first camp, along with such disparate thinkers as Austin,
Holmes, Hart, and Marx. They argue that law pertains to social
order simpliciter. Aristotle, Cicero, Aquinas, Fuller, Dworkin, and
Finnis, among others, reside in the rival camp. They maintain
that law pertains to a social order of a certain sort: a moral order.
From either of these two viewpoints, certain cases will be seen
as undeveloped, primitive, corrupt, or deviant. But especially with
respect to the "corrupt" and "deviant" classifications, the cases will
almost certainly be sorted differently. Which point of view is
preferable, and why, remains a, if not the, perennial question in
philosophy of law. It is important, I believe, not to ignore it,
even if it means being less than generous to certain problematic
cases.

IV. LEGAL PROCESS AND LEGAL RIGHTS

In Watson's scheme, the legal process is called into play by

claims of legal right, power, or privilege. 91 Watson maintains that
it is precisely the availability of a process to resolve disputes that

90 Id. 10-11.
01 A. WATSON, supra note l, at 28.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981] BOOK REVIEWS 253

makes a right uniquely legal. 92 We are left to infer that other

types of rights, and particularly moral rights, do not have an insti-
tutionalized process of dispute resolution available to them. This
would help explain why so many claims of moral right go unheeded.
As is also true of many other assertions in The Nature of Law,
Watson's claims regarding the nature of legal rights are suggestive
but undeveloped. Watson sees quite clearly that rules create legal
rights, that such rights are logically prior to any legal process, and
that the creation of legal rights may be unrelated to the settling
of disputes. Regrettably, ·watson feels compelled nonetheless to
assert, without offering adequate support, that all legal rights must
be enforced by an appropriate legal process.
Without doubt, there could not be an effective system of legal
rights without the availability of a dispute resolution process. But
what must be true of a system, much less a presumptively effective
system, need not be true of each of its parts. Because he defines
legal rights only in terms of their use within a dispute resolution
process, Watson confuses the identity and existence of legal rights
with their effectiveness. An account of legal rights that does not
regard the availability of a dispute resolution process as essential
would recognize this important distinction. Legal rights could
then be explicated in terms of entitlements, that is, claims justified
within a system of legal rules. Moral rights could then be dis-
tinguished as claims justified on the basis of moral principles. Puta-
tive moral rights-and even legal rights, for that matter-may be
problematic, controversial, or nonexistent. Nevertheless, this does
not justify confusing questions of identity and existence with ques-
tions of effectiveness.

V. LAW AND LEGAL PROCESS

Watson's adoption of methodological essentialism, and the con-

sequent narrowness of his approach to the function of law, limit
both the generality of his theory of law and the validity of his con-
clusions. This is preeminently true with respect to his analysis of
the internal workings of legal systems. Where ·watson focuses on
the interrelationships among legal systems, however, his analysis
and conclusions are both thought-provoking and suggestive, though
incomplete. These consequences of Watson's narrow methodo-
logical posture will be examined in this section by concentrating
on several interesting and important portions of his overall argu-
ment and conclusions.
02 Id. 42.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
254 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol 130:229

A. Legal Rules and Legal Process

Just as Watson argues that there are no legal rights in the
absence of legal processes to effectuate them, he also argues that
rules become law only when disputes are institutionalized and a
process exists for their resolution: there can be no legal rules
without legal process. Watson admits that much modern legisla-
tion-regulatory provisions, for example-seems remote from legal
process. Nevertheless, he argues that always "the possibility of a
process is there . . . and if a dispute arose the norms set out in
these laws would appropriately be used both to initiate the process
and to validate any decision given in the process." 0.3 Watson here
again confuses identity and existence with effectiveness, just as he
does in his analysis of legal rights.
Perhaps the most noticeable deficiency in \Vatson's account of
legal rules is his failure to discuss the internal structure of legal
systems. Watson thus shows that he is indifferent, if not hostile,
to much that has been written under the separate but related
banners of legal positivism and analytical jurisprudence. As we
have seen, Watson is much less concerned with analysis of internal
relations than with understanding the functional dynamics of legal
process. 94 This concern and attention provides a much needed
supplement to the preoccupations of such philosophers as W.N.
Hohfeld, A.L. Goodhart, H.L.A. Hart, Hans Kelsen, and Joseph
Raz. Yet this omission at the same time diminishes the attractive-
ness of Watson's offering as a general theory of law.

B. The Scope and Purpose of Legal Process

Watson vacillates somewhat about whether the legal process
aims only to inhibit future, unregulated conflict, or whether it
also seeks to resolve existing disputes. He ultimately argues only
for the former, but why not have both, and more? In reality, law
sometimes inhibits future conflict, sometimes resolves existent dis-
putes-and sometimes permits or even encourages conflict! Wat-
son's quest for a single essential element thus once again prevents
him from acknowledging the complex aims of law.
Watson's refusal to accept arbitration as a form of legal process
is also troubling. Because nonbinding arbitration yields only a
recommendation, Watson maintains that it is not part of the legal
process. "[T]here is no decision saying 'this 1s the case' or 'you
oa Id. 32.
94 See supra text accompanying notes 55 & 61-63.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981) BOOK REVIEWS 255

must do this'. The claim that initiates a legal process also demands
a decision, not a recommendation." 911 Perhaps this is so, but
\Vatson offers little support for the proposition beyond its asser-
tion. Is a decision always required? In certain states, such as
Wisconsin, the Attorney General may offer nonbinding but legally
influential opinions on certain controversial, yet unlitigated, legal
issues. Such opinions are authorized by legal rules and often given
weight in subsequent judicial opinions. Given that these opinions
are more influential than recommendations, yet are surely not de-
cisions, they would seem to be excluded from Watson's scheme of
things. But is there any better reason for this exclusion than that
they would disturb the scheme?

C. Law and Coercion

Law is backed by recognized authority, and so it must be if it
is to fulfill its functions, Watson argues. In his view, authority
consists of two elements: force and respect. Law and force are not,
he argues, necessarily enemies, for law is unlikely to be effective
unless it is supported by the threat of force. This is especially
true with respect to the legal systems of nation states. Watson
goes so far as to contend that law does not cease to be law even
when it requires the backing of enormous violence. 96
Watson's discussion of the authority of law is at once proble-
matic and insightful. Contrary to his argument, law that requires
enormous force to keep it alive is law only in an attenuated sense,
for such law is moribund. Perhaps it can be revived and restored,
but if we are to call it "law" despite its grossly deformed state, we
should add "in a sense," or "sort of." Only then could we say, for
example, that Uganda had a legal system during the final years of
Idi Amin's reign of terror.
Watson does make the point, which is an advance over the
views of such sophisticated positivists as Hart and Raz, that the
concepts themselves of law and legal system do not require regu-
lated sanctions. At least in theory, he says, "[i]nternational law
enables us to envisage . . . a system of law not at all backed by
violence, whose existence rests on acquiescence and respect coupled
with non-violent sanctions." 97 This is true, and is also important
in that it invites us to perceive that inducements, as well as sanc-
tions, can support law-even the law of nation states. 98
96 A. WATSON, supra note 1, at 59.
96 See generally id. 70-82.
97 Id. 80.
98 See generally Oberdiek, supra note 72.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
256 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

Respect is the other element of authority. Watson argues that

law does not require consent or acceptance; at the very most it
demands acquiescence. From the citizen's point of view, this need
mean no more than that he fears the superior force of the state.
Watson views that situation, however, as a degenerate case. He
explains the more typical forms that respect for law can take:
Respect for law because it is law means, on the one hand,
that we can envisage legal systems (though not of a nation
state) which are not backed by force; on the other, that
(for instance, in a nation state) the great extent to which
law's validity depends on force is masked. 99
Regrettably, Watson again concentrates only on appearances: "We
will be concerned not with factors which show that law is worthy
of respect but factors which either demonstrate that law is thought
to be worthy of respect or derive from such feelings of respect." 100
As argued earlier, 101 this is a misguided approach, whether's one's
intent is to describe or prescribe.
D. Legal Transplants
Watson's discussion of legal transplants highlights his interest
in and insight into the history and sociology of law, and is one of
the best parts of his book. Because law is primarily developed by
lawyers, Watson argues, it is often borrowed from other systems
where models are readily available. The society that borrows con-
serves its own resources, and benefits from the quality and prestige
of the laws of the exporting systems. These benefits are especially
substantial, according to ·watson, if the transplant has the authority
of Roman law behind it.
The transplanted law is frequently second best, if not second
rate, for the borrowing system, however, for it will seldom precisely
fit the needs of either the society as a whole or of its ruling elite.
Because so much Roman law in particular survives in contemporary
legal systems, Watson disputes the Marxist claim that every stage
of economic development carries with it its own unique legal system.
Watson's generally realistic theory of law is supported by the ob-
servation that societies often import laws that are less than optimal
for them:
Society has the ability to tolerate a great deal of law that
is not the best that can be devised for that society, and
99 A. WATSON, supra note 1, at 114.
100 Id. 114-15 ( emphasis added).
1-01 See supra notes 17-35 and accompanying text.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
1981) BOOK REVIEWS 257

knowledge of the deficiencies does not automatically cause

intolerance. It appears that to a considerable extent what
matters more is that there are rules than that the rules are
the best suited for that society. 102
After questioning the suitability of certain legal rules for their
particular societies, Watson rather complacently concludes that
daily courtroom battles to determine the best legal rule to apply in
a particular case may be a poor use of society's resources:
If the argument is correct it should mean that it is not
necessarily very important for society that a judge in a law-
suit get his law right . . . . It is important that the attempt
be made to get the law right (or cynicism will ensue),
and that the decision be acceptable. But what is of sig-
nificance is that the law best suited for a society's needs
is by no means inevitably that which exists in the society,
and, to the extent that it is not, the expenditure of great
effort in a law suit to achieve the legally correct result may
be a misuse of resources. 103
Watson's analysis and conclusions regarding legal transplants are
provocative and are worth further investigation and reflection.

VI. THE UNDEVELOPED PoTENTIAL oF \VATsoN's ANALYSIS

If we purge The Nature of Law of its methodological essen-

tialism, define and develop its functionalism, and expand its account
of legal rules, then we can see the outlines of a challenging synthesis
of legal positivism and sociological realism, with a gesture towards
natural law. The features that Watson thinks typify law, after all,
are just those that positivists stress: both view law as comprehensive,
created, authoritarian, supreme, backed by sanctions, and regularly
obeyed. Futhermore, Watson's view of the function of law-to
settle disputes-is broadly shared by positivists and realists. At the
same time, Watson's sociological concerns turn him away from the
positivists' fascination with the formal, internal structure of legal
systems, and towards the subtle and diverse ways legal systems shape,
and are shaped by, other social institutions, including foreign legal
systems.
"\Vatson even acknowledges, however slightly, the perennial
appeal of natural law. For law, Watson grants, possesses a set
of particular virtues: formal enactment and publication of regula-
tions; accessibility of the legal process; clarity of legal rules; rea-
102 A. WATSON, supra note l, at 112.
103 Id.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms
258 UNNERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 130:229

soned decisions; established procedures, and consistency in legal

judgments. 104 Unlike natural law theorists, however, he refuses to
follow Lon Fuller in calling these moral virtues: 106 "the high value
[that Watson attributes] to the 'particular virtues' derives from the
concept of the essential functions of law and not from any argu-
ments as to morality or justice." 106
Why is Watson so reluctant to describe these particular virtues
of law as moral? In the absence of any discussion of morality, it is
difficult to say. My guess is that Watson believes that in order to
have a descriptive theory of law one must avoid describing it from
a moral point of view. If Watson's view is mistaken, or at least
not essential, the way is open to argue that legal systems lacking one
or more of these virtues, or possessing them only partially, are de-
formed, deficient, or deviant as legal systems. The way is also
open to include two important virtues that Watson excludes be-
cause of their avowedly moral nature: equal treatment of disputants
and proscription of ex post facto laws. 107

VII. CONCLUSION

The Nature of Law, despite its brevity, illuminates many of

the principal issues in contemporary philosophy of law. Its claims
are bold, far-reaching, and clearly expressed. It rightly stresses
the signal importance of process to an understanding of law, al-
though this preoccupation sometimes borders on obsession. In this
way, perhaps, it mirrors the curriculum of many schools of law,
which require courses in criminal procedure, for example, but not
criminal law.
One cannot fail to learn about law and the legal process from
Watson's book. Because it is informed by its author's command of
Roman law, his understanding of modern sociological theories, and
his familiarity with contemporary analytical jurisprudence, The
Nature of Law brings together related interests that are too often
kept apart.
The Nature of Law, despite and perhaps because of its flaws,
deserves a wide audience. I especially hope that it finds its way
into the hands of those who think that legal philosophy must be
turgid, abstract, and irrelevant. Watson's book gives the lie to
that old canard.
104 Id. 61-69.
105 See supra note 65 and accompanying text.
106 A. WATSON, supra note 1, at 69.
101 Id.

This content downloaded from

130.58.104.29 on Mon, 22 Nov 2021 16:34:07 UTC
All use subject to https://about.jstor.org/terms