Electronic copy available at: http://ssrn.com/abstract=464420 Electronic copy available at: http://ssrn.

com/abstract=464420
BRIAN H. BIX
RAZ ON NECESSITY
(Accepted 19 April 2003)
INTRODUCTION
Joseph Raz is the foremost theorist in contemporary English-
language analytical jurisprudence.
1
His work has contributed
signicant and provocative new ideas in many areas of legal philos-
ophy, including that most ancient, and most obscure, inquiry what
is (the nature of) law? Razs discussions of the nature of law
have included claims regarding what is necessary or essential
for law,
2
as well as offering other conclusions that are grounded on
assertions regarding this nature. These sort of comments regarding
necessity raise basic questions regarding the type of claims that
are being offered, and that can be offered, within legal philosophy.
The other issue raised, indirectly, is what is at stake in the question
of whether we should or should not speak in terms of necessity in
jurisprudence.
In this paper, I will investigate discussions of necessity in juris-
prudence, with particular emphasis on theories of the nature of law,
and on the work of Joseph Raz. In recent years, Raz has been one of
the few theorists to explore the important meta-theoretical questions
in jurisprudence. This essay will give a critical overview of these
works, offering a guide to how the debate has developed and what
its signicance is. One conclusion that will be reached is that the
1
His books have all been important contributions to the current debates in
legal, moral, and political philosophy: (Raz 1979, 1980, 1986, 1990, 1994, 1999,
2001, 2003).
2
I follow Raz, the philosopher Frank Jackson, and most writers in using
necessary/necessity and essence/essential as roughly interchangeable when
discussing conceptual analysis in particular, the connections that exist or are
claimed to exist between a concept and constituent properties.
Law and Philosophy 22: 537559, 2003.
2003 Kluwer Academic Publishers. Printed in the Netherlands.
Electronic copy available at: http://ssrn.com/abstract=464420 Electronic copy available at: http://ssrn.com/abstract=464420
538 BRIAN H. BIX
possibility of necessity talk in jurisprudence goes hand-in-hand
with the possibility of conceptual analysis; if one concludes that one
is impossible or inappropriate in discussing law, then likely the other
is as well. As will be discussed, a defense of conceptual analysis
in jurisprudence must likely follow Razs lead, offering a notion
of necessity that is distinctly not Platonist, but is rather deeply
grounded in a communitys way of life or its self-understanding.
In this sense, one can have the paradoxically sounding necessary
truths that change over time. At the same time, there is still much
work to be done in articulating and justifying a conclusion that there
is (or we have) only one concept of law.
The investigation will proceed in stages, beginning with a
more general consideration of necessity claims in philosophy,
and then facing the specic question of whether such claims can
and should be made within theories about the nature of law.
Part I discusses necessity claims in philosophy. Part II introduces
conceptual analysis, offers an overview of its use in jurisprudence,
and considers one recent critique of that approach. Part III considers
a different line of criticism of conceptual analysis in jurisprudence
that there is no single concept of law and also evaluates how a
similar critique might apply to Razs view.
I. THE PROBLEM OF NECESSITY
What is Necessity?
In his essay, Authority, Law, and Morality, Raz writes: I will
assume that necessarily law, every legal system which is in force
anywhere, has de facto authority. That entails that the law either
claims that it possesses legitimate authority or is held to possess
it, or both.
3
This is a central point in Razs argument, though it
appears with little prior argument or stage-setting. Raz goes on to
assert: If the claim to authority is part of the nature of law, then
whatever else the lawis it must be capable of possessing authority.
4
3
Raz (1994, p. 199).
4
See Raz (1994, p. 199). The claim is repeated, often more emphatically, in
Razs later writings, see, e.g., Raz (1988, p. 3; 1996, p. 6), and a form of the claim
can be found in an earlier work (Raz 1979, pp. 2833).
Electronic copy available at: http://ssrn.com/abstract=464420 Electronic copy available at: http://ssrn.com/abstract=464420
RAZ ON NECESSITY 539
From a view about what it takes to be potentially authoritative, Raz
elaborates an argument for what has become known as exclusive
legal positivism: that by the nature of law, legal rules must be
identiable without resort to moral judgments.
5
My interest in this essay is not in evaluating the merits of Razs
argument-from-authority for exclusive legal positivism, a task many
others have already undertaken.
6
My concern is rather with the
character or status of Razs initial claim. He is making a claim
about, as he puts it, the nature of law, or the necessary elements
of everything which falls under the category law. Lest this use
of necessity be thought to be aberrational within Razs work, it
should be observed that the same sort of analysis appears, with
greater elaboration, in a number of his later works.
7
It should also be
noted that Raz is by no means the rst legal theorist to make claims
of this sort. John Austin distinguished between laws properly so
called and laws improperly so called based on whether an object
had all the qualities composing the essence of the class;
8
and
wrote of endeavour[ing] to resolve a law . . . into the necessary and
essential parts of which it is composed.
9
Still, reference to essence and necessity must strike a slightly
discordant note to modern ears. This may be exemplied by Ronald
Dworkins reference, in a somewhat different context, to the pecu-
5
See Raz (1994, pp. 199204); see also Waluchow (1994, pp. 123140)
(discussing Razs argument from authority for exclusive legal positivism).
6
See, e.g., Waluchow (1994); Coleman (1998b, pp. 267278); Leiter (1998,
pp. 540544).
7
See, e.g., Raz (1996, p. 2) (The universality of the theses of the general
theory of law is a result of the fact that they claim to be necessary truths, and
there is nothing less that they can claim); Raz (1998, p. 255) (What then counts
as an explanation of a concept? It consists of setting out some of its necessary
features, and some of the essential features of whatever it is a concept of. In our
case, it sets out some of the necessary or essential features of the law).
Dickson (2001, p. 17 n. 24) notes the issue of Razs use of necessity, but does
not analyze the matter further.
8
Austin (1995, Lecture V, pp. 108109) (Austins Province, considered
by many to be the seminal work for legal positivism, was rst published in
1832).
9
Austin (1995, p. 117).
540 BRIAN H. BIX
liar world of legal essentialism.
10
What is such metaphysical
language doing in a discussion of a social institution?
11
Necessity Claims Generally
Is there a place for necessity within discussions of law? Some
philosophers have argued for necessity in the denition of certain
terms, when those terms denote some category whose boundaries
are arguably set out by the way the world is. These are natural
kind terms, like water and gold, and the debate within the liter-
ature, at least initially, was addressed to the question of whether
terms of this kind have their reference determined by peoples
beliefs about the items nature or by the way the world is.
12
Whatever the merit of a natural kinds analysis for terms that refer
to natural or physical entities, its applicability to human institutions
and social practices would seem to be much more problematic.
Gold may be a category whose boundaries are set by the world, and
its essence estimated by the best scientic theory we currently have;
there is, however, little reason to think that a similar approach would
work for baseball or for law. In what way could the world be
said to delimit what does and does not count as law?, and what
would it mean to have a scientic theory of the nature of law?
13
10
Dworkin (1984, p. 259); but see Kramer (1999, p. 179) (Either [Dworkins]
allegations about essentialism are false, or they apply as much to him as to any
positivists).
11
On the other hand, some legal theorists think that it is a way of thinking that
jurisprudence cannot do without:
Even though it is no longer fashionable to talk of essences, something
like the idea of the essence of a concept must be kept in mind in these
debates [between legal positivism and natural law theory], or we will no
longer be able to distinguish the peculiar science of jurisprudence from any
old discussion of a legal topic (Soper 1995, pp. 365366).
12
See generally Putnam (1975). For a critical analysis of attempts to apply
natural kinds theories to law, see Bix (1993, pp. 157173).
13
Michael Moore (1998, p. 312) suggests that H.L.A. Harts legal theory could
be seen as implying something analogous just as there are natural kinds in
the natural world, so there are social kinds in the social world, and law is one of
them but this still leaves us with the question of what it would mean for there
to be social kinds.
RAZ ON NECESSITY 541
Another analogy within the philosophical literature might be
Saul Kripkes idea of rigid designators: that in counterfactuals,
singular terms are intended to have the same reference in all possible
worlds.
14
Again, while the analysis is arguably persuasive as regards
proper names, it would be awkward, at best, if applied to a social
practice or social institution like law.
15
In the context of theories about the nature of law, and the use
of necessity within such discussions, the Kripke-Putnam theories
about reference and semantics do not seem helpful, except perhaps
by broad analogy.
16
There remains one path in philosophical
analysis to consider.
II. CONCEPTUAL ANALYSIS AND JURISPRUDENCE
Conceptual Analysis in General
One likely response to the discussion up to this point would be: Of
course, a jurisprudential discussion about the nature of law is not
an analysis of logical necessity, or even of a natural kind. It is a
conceptual analysis, and whatever necessary or essential claims
are involved are those of the inquiry into concepts.
17
14
See Kripke (1972).
15
One can accept Kripke and Putnams positions on a more general level, that
meaning has a social dimension, and is not individualistic (in the mind), even if
one does not accept that the world determines the meaning of our concepts. Raz
(1998, pp. 262264 & n. 26). The signicance of this compromise position for
the present analysis will become clearer later in the paper.
One should also note Nicos Stavropouloss recent efforts to ground objectivity
in legal interpretation in an analysis that draws upon, and is comparable to,
Kripke-Putnam semantics. Stavropoulos (1996).
16
See n. 15 above.
17
Of course, when the classical philosophers wrote of essential and accidental
properties, they were usually referring to the essential and accidental properties
of things, not of concepts. See, e.g., Aristotle, Metaphysics, Book VII, ch. 4
(Aristotle 1984, pp. 16251627).
Whether conceptual analysis must be in terms of necessary or essential
terms whether all concepts have that structure, or can be analyzed in that way,
or whether some can only be analyzed in terms, say, of paradigm cases B is a
question that goes beyond the terms of this paper. It is sufcient for our purposes
that some concepts can be analyzed this way. Whether law is best analyzed this
542 BRIAN H. BIX
Philosophical analysis of concepts is, of course, nothing new. For
example, there was a long-standing debate about whether knowl-
edge should be dened as justied true belief.
18
We do believe
that we can sensibly analyze our concepts, and, at least sometimes,
determine what their essential (and accidental) attributes are.
19
Also, conceptual analysis is certainly nothing new for jurisprudence
either: arguably the most important jurisprudential text published in
English in the last century was described by its title as being about
a concept, H. L. A. Harts The Concept of Law.
20
However, one might ask, why should we study the concept if
we can study the thing itself (the practice, the type of institu-
tion) instead?
21
This may seem like an empiricists (or an anti-
intellectuals) response to impractical, overly abstract philosophers.
At that level, the proper response is that conceptual analysis is a
prior inquiry we cannot study law until we know what we mean by
law.
22
Some might persist that the proper study of law a social
institution is through social theory. Law is a set of social prac-
tices, the argument would go, so its nature is best discovered, not by
armchair reections, but by an investigation of the actual practices
(a view that will be considered at greater length below). However,
should someone suggest that the investigation of the nature of law
way can be considered without resolving the question of whether all concepts
must be analyzed this way.
18
Gettier (1963).
19
Cf. Raz (1998, p. 273 n. 38), where Raz distinguishes those features of law
which are general, i.e., shared by all legal systems and the essential features of
law, features without which it would not be law.
20
Hart (1994); see also Raz (1980) (The Concept of a Legal System).
21
Michael Moore seemed to express just that view when he wrote: General
jurisprudence should eschew . . . conceptual analysis in favour of studying the
phenomenon itself, law. Moore (1992, p. 206). However, given the source, a
moral and legal philosopher whose work is grounded on a metaphysically realist
approach, this was not the likely intention of the quotations challenge. The place
played by conceptual analysis in most schemas is played in Moores anlaysis
by Platonic essences, natural kinds, or functional kinds. Whether a metaphysical
realist like Moore can make do without conceptual analysis is a discussion for
another time. See also Moore (1998, p. 313) (reiterating the difference between
studying a phenomenon through concepts and studying it directly, but concluding,
at least as regards law, that the difference here is not great).
22
See, e.g., Jackson (2000, pp. 3031); cf. Coleman (2002, pp. 347351)
(offering a similar response to a naturalist critique of conceptual analysis).
RAZ ON NECESSITY 543
be purely empirical/sociological, that claim would be vulnerable to
the argument just offered: how can one have a sociological theory
of law if one does not have at least a rough prior notion of what is
or is not law?
23
There is thus a sense in which conceptual work must be prior
to empirical work.
24
For the focus is inevitably on the boundaries
of the category here, what makes something law or not law?
We are not asking empirical questions about particular institu-
tions: e.g., about the historical origins of common law reasoning in
the English legal system, or the interpretive practices of American
judges when construing statutes. Questions about specic institu-
tional practices would be social theory inquiries, which would call
for some combination of model building, observation, and statistical
analysis. However, the more general discussion of the nature of law,
if such discussion has any place at all, is not a comparably empirical
inquiry.
25
One might point out that if it would be mistaken to try to
ground a theory of the nature of law solely on empirical or soci-
ological grounds, without reference to conceptual analysis, it would
be equally mistaken to ground such a theory solely on conceptual
analysis, without reference to empirical and sociological truths.
26
Indeed, what sense or value could there be to a purported concept
of law if that concept had no relation whatsoever to the practices
we associate with legal systems? Razs own view is that the concept
23
One possible response is that while a prior notion of law is needed before
beginning other (empirical) work, simple intuitions and linguistic usage patterns
would be sufcient for that purpose. No thicker conceptual analysis is needed (or,
some commentators might add, possible). Leiter (2002).
24
However, there is also a sense in which the theorist doing conceptual analysis
must defer to the way the world is, at least in those cases where the theorist
is investigating the nature of an already-existing concept. The matter would be
different if we were positing some new concept or category, and then consid-
ering what empirical claims could be made about that concept. See Raz (1994,
p. 221).
25
None of this is to claim that sociological inquiry must be subordinate to
conceptual analysis. The fact that we have a rough sense of (e.g.) what is and
what is not law does not mean that social theories must be built on categories
that track those concepts.
26
This view has been forcefully advocated by Brian Tamanaha. See, e.g.,
Tamanaha (1997, 2000, 2001, pp. 132).
544 BRIAN H. BIX
of law is grounded on the perceptions and self-understandings of
people
27
self-understandings which, in turn, one presumes, reect
the social practices that help to constitute the social institution. The
connection between conceptual analysis and empirical truths will be
discussed further, below.
Skepticism About Conceptual Analysis: Naturalism and Family
Resemblance
Naturalism
As one nal stop in this brief overview of conceptual analysis, it
should be taken into account that some theorists challenge the whole
notion of concepts and conceptual analysis. A challenge of this sort
is offered (though only partly endorsed) by Brian Leiter:
What is a concept? A cynic might say that a concept is just what philosophers
used to call meaning back when their job was the analysis of meaning. But ever
since Quine embarrassed philosophers into admitting that they didnt know what
meanings were, they started analyzing concepts instead.
28
In a way, this challenge to conceptual analysis is related to a nomin-
alist critique, which will be considered in greater length in Part III.
In addition to the responses to the nominalist critique that will be
considered at that point, one might add (as Leiter himself does), the
concept of law has an advantage over the concept of the good, in
that there is an identiable set of practices and institutions to ground
our discussions.
29
The concept of law cannot easily be accused of
being an entirely mysterious entity, made up by metaphysicians in
their spare time.
30
Further, as Jules Coleman has argued, the search for analytic
truths that W. V. O. Quine criticized is quite different from what
27
See, e.g., Raz (1996, pp. 56).
28
Leiter (1998, p. 535). Leiter continues: The cynical view has, I believe, a
modicum of truth, but it is hardly the whole story. Id. Cf. Jackson (2000, p. vii)
(Properly understood, conceptual analysis is not a mysterious activity discred-
ited by Quine that seeks after the a priori in some hard-to-understand sense.
It is, rather, something familiar to everyone, philosophers and non-philosophers
alike); see also id., pp. 4446, 5255 (responding to Quine).
29
Leiter (1998, p. 536).
30
Compare J. L. Mackies famous accusation that moral objectivism depends
on the belief in queer entities. Mackie (1977).
RAZ ON NECESSITY 545
modern legal theorists were (and are) doing in their conceptual
theories.
31
Neither H. L. A. Hart nor Joseph Raz or Jules Coleman,
nor any other prominent legal theorist, could reasonably be under-
stood as trying to determine the analytical essence of some
trans-historical trans-empirical (Platonic) Idea.
32
Family Resemblance
Ludwig Wittgenstein famously introduced the notion of family
resemblance as a shorthand for the way that some concepts and
categories (Wittgenstein used the examples language, game, and
number) cannot be understood in terms of necessary and sufcient
conditions, but rather have a variety of different and overlapping
criteria.
33
Wittgenstein was not claiming that all concepts were
family resemblance concepts, only that some were, and therefore it
would be a mistake to assume that there would always be necessary
and sufcient conditions for every concept.
34
A number of writers have suggested that law might be such a
family resemblance concept, with instantiations having no feature
in common and thus no necessary features.
35
Hart himself
suggested that the notion of family resemblance might be partic-
ularly relevant to legal terms,
36
and he broadly hinted early in The
Concept of Law that law might well best be understood in this
way,
37
though later in the same book he offered what appeared to be
a set of necessary and sufcient conditions for that term.
38
That noted, because no one claims that all concepts are family-
resemblance concepts, even if one accepts that some are, analysis
and debate must be developed concept by concept. One way to
disprove that law is a family resemblance concept is to provide
an analysis in terms of necessary and sufcient conditions, as Raz
31
Coleman (2002, pp. 343351).
32
Coleman (2002, pp. 350351).
33
Wittgenstein (1958, 6568). For a critical discussion, with further
citations, see Glock (1996, pp. 120124).
34
See Glock (1996, pp. 123124).
35
See, e.g., Burton (1985, pp. 19791980); Lyons (1983, p. 259).
36
Hart (1994, pp. 279280).
37
Hart (1994, pp. 1516).
38
Hart (1994, p. 81).
546 BRIAN H. BIX
and others have attempted to do. If the analysis succeeds, that
sufces to show that law is not a family resemblance concept.
39
The Connection with Practice and the Number of Concepts
In discussing the challenge of naturalism, I noted that theories about
the concept of law might be less vulnerable than other concep-
tual analyses to naturalist attacks, because the concept of law
is grounded in actual practices. However, to say that conceptual
analysis is connected with lived experience in some ways leads reas-
onably to the question a surprisingly difcult one of what that
connection is.
40
Raz suggests the following:
The concept of law is a historical product, changing over the years, and the
concept as we have it is more recent than the institution it is used to single out.
. . .
But the concept of law is not a product of the theory of law. It is a concept
that evolved historically, under the inuences of legal practice, and other cultural
inuences, including the inuence of the legal theory of the day.
41
In other words, todays concept of law is different from the concept
of law of some generations or centuries in the past. This in turn
raises the question of the quantity of concepts of law (more than one
over time?, more than one at any given time?), and their parochial
or universal nature.
When we are analyzing the concept law, the modier we place
in the description can be crucial. Are we describing, as in the title
to H. L. A. Harts book, The Concept of Law, implying that there
is (and has always been) only one? Or are we merely offering a
concept of law, implying that this is merely one possible concept
among many.
42
Also, even if it is only one possible concept among
many (and thus, in a sense, contingent, not necessary), is the
focus on this concept non-arbitrary that is, is there some good
39
Although, of course, the opposite is not the case: the failure of a particular
necessary-and-sufcient-conditions analysis does not prove that law is a family
resemblance concept, though it may help to fuel doubt in that direction.
40
I discuss the issue in Bix (2000).
41
Raz (1998, pp. 280, 281).
42
Someone once suggested that the two books, The Concept of Law (Hart
1994) and A Theory of Justice (Rawls 1999), might have usefully exchanged
articles.
RAZ ON NECESSITY 547
reason why we should look to this concept rather than another?
For example, might one argue that we are focusing on a partic-
ular concept among different possible concepts because it is our
concept of law though contingent, in the sense that there are
other concepts of law, this is the one that matches our communitys
linguistic practices or general self-understanding?
Jules Coleman, in a recent article, has advocated thinking in
terms of our concept of law, tying that position to a somewhat
deationary notion of necessity:
The descriptive project of jurisprudence is to identify the essential or necessary
features of our concept of law. No serious analytical philosopher . . . believes that
the prevailing concept of law is in any sense necessary: that no other concept
is logically or otherwise possible. Nor do we believe that our concept of law
can never be subject to revision. Quite the contrary. Technology may someday
require us to revise our concept in any number of ways. Still, there is a difference
between the claim that a particular concept is necessary and the claim that there
are necessary features of an admittedly contingent concept.
43
In what may well be more of a difference of label than of substance,
Raz writes of a concept of law that seems to be both contingent and
necessary (or, in his somewhat different terminology, both paro-
chial and universal
44
). According to Raz: (1) we have a concept
of law; (2) based on our societys self-understanding; and (3) our
concept of law has changed over time, in response to changes in
institutions, practices, attitudes, and even philosophical theories.
45
Let us look more closely at these notions within Razs analysis.
Raz is not a Platonist, and therefore does not believe that the concept
of law is some eternal Platonist Idea, which would be the same for
43
Coleman (1998a, p. 393 n. 59). While I am not entirely sure what Coleman
means by technology requiring the revision of a/our concept, the notion of a
contingent concept, on its own, seems understandable.
In an earlier work, I responded to Colemans quotation in the text by asking
about what work was being done by necessary or essential in the phrase the
essential or necessary features of our concept of law. Bix (1999, p. 28 n. 59).
While noting how similar Colemans position seems to be to that of Razs, I
retract the earlier criticisms. I would only add that it is still important to be able
to distinguish the features of legal systems, even the features of all legal systems,
from the necessary or essential features that make those systems legal. See
Raz (1998, p. 273 n. 38).
44
See, e.g., Raz (1996, pp. 17).
45
See Raz (1996, 1998, 2004).
548 BRIAN H. BIX
all people or for all times.
46
Therefore, it is natural to suspect that the
concept we investigate is our concept, the product of a specic
culture our own.
47
And since what counts as law (under our
concept) is independent of a societys possessing that concept, there
were likely earlier cultures or alien cultures that did not or do not
share or have our concept, yet still had law.
48
While the concept of law has changed over time not some
unchanging Idea we are discovering Raz treats the/our concept
of law as something unique, a matter about which we can be right or
wrong in our descriptions, and which we cannot simply re-invent for
our own purposes (though he does note that since concepts of law
are in ux, our theories of law, even mistaken theories, could inu-
ence the concept of law future generations have
49
). Similarly, Raz
rejects the notion that we (as theorists) can choose a concept of law
based, say, on its fruitfulness in further research,
50
or even according
46
Contrast Ciceros comments on natural law:
True law is right reason in agreement with nature; it is of universal applica-
tion, unchanging and everlasting; . . . And there will not be different 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 . . . .
Marcus Tullius Cicero, The Republic, Book III, xxii (Cicero 1928, p. 211). I do
not mean to imply that Ciceros view of an ideal law, or an eternal standard for
morally judging all positive laws, is the same as modern conceptual analyses of
law. I use Ciceros language only to exemplify a view of something unchanging
over time and independent of experience.
47
Raz (1996, p. 5).
48
Raz (1996, pp. 4, 5, 6).
49
Raz (1996, p. 7).
50
See, e.g., Raz (1994, p. 221):
[I]t would be wrong to conclude . . . that one judges the success of an analysis
of the concept of law by its theoretical sociological fruitfulness. To do so is
to miss the point that, unlike concepts like mass or electron, the law is
a concept used by people to understand themselves. We are not free to pick
on any fruitful concepts. It is a major task of legal theory to advance our
understanding of society by helping us understand how people understand
themselves.
Among those who appear to take a contrary view regarding choosing concepts
according to usefulness, see, e.g., Leiter (1998); Lyons (1983, pp. 5759); Tama-
naha (2000, pp. 283288). (In an earlier work (Bix 1999, pp. 928), I also seemed
RAZ ON NECESSITY 549
to its simplicity or elegance;
51
rather, it is a concept already present,
already part of our self-understanding. Raz refers repeatedly to the
concept of law which exists independently of the legal philos-
ophy which attempts to explain it,
52
and the nature of law which
general theories of law must strive to elucidate.
53
When these aspects of Razs view of the concept of law are
combined, they result in a position which might seem problem-
atic in two different ways. First, under Razs analysis, the concept
may apply to societies who do not or did not have the concept.
54
Raz emphasizes that nothing radical is implied or assumed by this
position: only that some ways of articulating our understanding of
ourselves develop slowly, as do concepts for understanding alien
cultures (such understanding requiring the development of concepts
which allow us to relate those cultures understanding of their prac-
tices to our understanding of our own practices).
55
As Raz points
out, we seem untroubled by this sort of analysis elsewhere: for
example, we can talk about the standard of living of a society
which existed long before that concept had been articulated.
56
The second problem is one that some readers might nd harder to
shake off: the way Raz combines references to necessity with talk
of historical contingency. This can be confusing, given the connec-
tions, mentioned earlier, within normal philosophical discourse
between necessity and the way things must be or the way
things must be in all possible worlds. The necessity in conceptual
analysis at least in Razs conceptual analysis is of a softer
kind, as it were. It means only that these are connections internal
to the concept in question (e.g., to be a legal system is to claim
authoritative status), a concept which is itself contingent and may
be tied to a particular community and time-period.
to endorse a contrary view, but I was, and am, more agnostic on this subject than
that text might imply.)
51
See Raz (2004).
52
Raz (1998, pp. 280281 (emphasis added)).
53
Raz, (1998, p. 2 (emphasis added)); see also Raz (2004).
54
See, e.g., Raz (1998, p. 4) (the concept of law is itself a product of a specic
culture, a concept which was not available to members of earlier cultures which
in fact lived under a legal system).
55
Raz (1998, pp. 45).
56
Raz (2004).
550 BRIAN H. BIX
III. CRITICAL RESPONSES
Beyond the skepticism about conceptual analysis discussed earlier,
there are a variety of possible critical responses to the Razian claim
that we can and should speak of necessary truths regarding our
concept of law. This Part will offer an overview of some of the
criticisms which have been brought, or could have been brought,
as well as the responses that might be made to those criticisms.
Nominalism and Pluralism
57
First, there is the view, hinted at earlier, that there is no single
concept of law, or at least none that should be given priority over
all the others. This view is well-presented by Brian Tamanahas
comment:
The project to devise a scientic concept of law was based upon the misguided
belief that law comprises a fundamental category. To the contrary, law is thor-
oughly a cultural construct, lacking any universal nature. Law is whatever we
attach the label law to.
58
This can be seen to be a nominalist attack on conceptual theory:
there is no category (natural or otherwise) law, law is whatever
we want it to be, so it is a strange exercise at best to wonder about
the nature or essential nature of something we have constructed
(and could construct a different way if we so choose). Perhaps juris-
prudence can only be, in a phrase used by one commentator, a
conjunction of lexicography with local history, or . . . a juxtaposition
of all lexicographies conjoined with all local histories.
59
One response
60
to this sort of nominalism
61
(though one more
modest or minimalist than Raz would likely offer) is that one need
not posit any sort of metaphysical grouping to justify theorizing
57
I do not mean to invoke the sociological/anthropological notion of legal
pluralism, with which the following discussion has at most a tangential
connection. On the social science notion, see, e.g., Tamanaha (1993).
58
Tamanaha (1997, p. 128 (footnote omitted)).
59
Finnis (1980, p. 4). Finniss position, of course, is that Jurisprudence is more
than just such a conjunction. See id. at pp. 318.
60
It is a response elaborated in Bix (2000, pp. 230234).
61
One writer has dened nominalism as The view that things denominated
by the same term share nothing except that fact. Blackburn (1994, p. 264).
RAZ ON NECESSITY 551
about concepts. However arbitrary the inclusion or exclusion of
items in our category law, if there is something interesting that
can be said about all (and perhaps only) the items in that category,
the process of theorizing will have value.
62
(One could also come at
the question from the other direction, as Frederick Schauer did, and
offer the suggestion that maybe there is a single concept, law, but
nothing interesting can be said about it.
63
)
One can invert the prior point: not that there should be more-or-
less arbitrary categories, about which there may or may not be some-
thing interesting to say, but rather that we should build or select
the categories which will have the best practical consequences.
64
Frederick Schauer, controversially, associates that position with
both H. L. A. Hart and Lon Fuller: Both Fuller and Hart appear
equally committed to the belief that giving an account of the nature
of law is not so much a matter of discovery as one of normatively-
guided construction, with the best account of the nature of law being
the one most likely to serve deeper normative goals.
65
62
See, e.g., Bix (2000, p. 231).
One should also distinguish a general skepticismabout concepts and conceptual
analysis from the criticisms American and Scandinavian legal realists brought
against judges and legal commentators who had reied legal concepts into objects
whose nature they studied, and used to justify particular legal conclusions. See,
e.g., Cohen (1935); Ross (1959, pp. 178183).
63
Schauer writes:
[N]ot every class that exists in the world is philosophically interesting
as a class. The classes residents of London, foods that begin with the
letter Q, and professional basketball players are all real even though
they are not natural classes, not ontologically primary, and not of great
philosophical interest. Similarly, law may exist as an analogously non-
ontologically primary aggregation of individuals, institutions, and practices,
undeniably part of the world but simply not having the philosophically inter-
esting core that philosophers of law have often supposed (Schauer 1994,
p. 508).
64
This is not to be confused with categories that have the best theoretical
consequences (consequences for research), a view associated below with Brian
Leiter.
65
Schauer (1994, p. 290 (footnote omitted)); cf. Hart (1994, pp. 207212)
(suggesting that between two rival concepts of law, one should choose the wider
one because it would help us in our confrontations with evil laws).
552 BRIAN H. BIX
Tamanaha has, in a recent article,
66
claried his original posi-
tion, discussed in the prior sub-section, and the clarication offers
a distinctly different possible critique. He argues that his earlier
comments should not be read as an attack on conceptual theory
as such, but rather as an attack on a particular kind of conceptual
theory. Instead of starting from the position that there is a single
concept of law or a core concept of law (e.g., state law), to
which other kinds of law (e.g., international law, customary law) are
peripheral, derivative, or defective instantiations, Tamanaha argues
for there being a multitude of specic manifestations of law, each
of which has its own characteristic features and none of which
would have any priority over the rest.
67
Tamanaha argues that there
is no good reason for giving state law priority over, say, interna-
tional law in ones theorizing about law.
68
Alternatively, it could
be pointed out that theorizing with state law at the core has likely
already achieved all the important insights which it can, and that
a different approach, centering on a different form of law (or a
different view of law) would likely produce new insights.
69
Whatever the merits of Tamanahas claim, one might argue that
it does not go to the core of Razs views. If there is one concept of
(state) law, the fact that there is also another or broader concept
that includes or overlaps that of (state) law is not something the
Razian legal theorist need deny.
Doubts About the Generality of General Jurisprudence
A different criticismis offered, albeit more implicitly than expressly,
in Ronald Dworkins work. Dworkin offers an interpretive approach
to law and legal theory, within which he asserts that the inter-
esting work will be at the level of interpretations of particular
legal systems, rather than at the level of general theories of law.
70
66
Tamanaha (2000).
67
Tamanaha (2000, p. 284).
68
Tamanaha (2000, p. 284).
69
Tamanaha (2000, pp. 284285). Tamanaha adds that a theory of lawin which
state law is predominant does not t as well to current practices, in which citizens
are often subject to multiple forms of law (from their home countries to the
European Union to various international tribunals). Id.
70
See, e.g., Dworkin (1986, pp. 102103; 1987, p. 16).
RAZ ON NECESSITY 553
Dworkins position is not so much that theories generally about law
are impossible or incoherent, but rather that they are not productive:
that there is nothing terribly interesting that one can say about all
legal systems, but that there are many things of value one can say
about particular legal systems.
71
(It may also be signicant that
Dworkin sees more general statements about law being tied to quite
specic claims made within daily legal practice. He famously states
that no rm line divides jurisprudence from adjudication or any
other aspect of legal practice.
72
)
One might respond to Dworkin the same way he has responded to
challenges to his right-answer theory based on global indeterminacy
or global incommensurability (incomparability). His response has
been that arguments cannot, or cannot easily, be made on a global
level, but must be made piecemeal. Dworkins argument is that for
a particular case, one puts up an argument for there being a (certain)
right answer, and it is up to the critic to show that for this question
there is no right answer, or that the values factored into a possible
answer are incommensurable.
73
The same sort of response could be
offered to Dworkins view on the proper scope of legal theory: once
a theory purports to say something interesting about (the concept
of) law generally, it will then be proper for critics to show that this
theory is faulty in some way.
74
Dworkins own work is, at best, doubtful support for this critique.
While it is true that he writes of the interpretation of particular
legal systems, and doctrinal areas within particular legal systems, he
simultaneously makes claims that apply to all legal systems:
75
most
importantly, that all legal systems indeed, all social institutions
are (should be) understood through constructive interpretation.
76
Also, while he offers one theory in discussions of the legal system
71
See Dworkin (1986, pp. 102103; 1987, p. 16).
72
Dworkin (1987, p. 14).
73
See, e.g., Dworkin (1986, pp. 266275; 1991, pp. 8990).
74
Cf. Raz (1998, p. 282).
75
Cf. (Raz 1998, p. 282): the book [Laws Empire] belies the modesty of
passages like the above [Laws Empire, at pp. 102103]. Time and again, from
its beginning to its very last section, it declares itself to be offering an account of
law, unqualied, in all its imperial domains.
76
See Dworkin (1986, pp. 4953). Dworkin denes constructive interpreta-
tion as a matter of imposing purpose on an object or practice in order to make
554 BRIAN H. BIX
of the United States,
77
he never indicates that a distinctly different
theory would be appropriate for some other, distinctly different legal
system (e.g., that of England, France, Iran, or Tibet).
Choices Still to Be Made (Normative Jurisprudence)
When Raz (and Coleman and others) try to defend a conceptual
jurisprudence unconnected with classical Platonism, this approach
has the advantage of not being burdened with a metaphysics many
people nd unlikely (at least where applied to social practices and
institutions). On the other hand, Platonismhas the relative advantage
of explaining why it is that there is a single (correct) answer to
conceptual inquiries about law. When we move from the concept
of law to our concept of law, there is more work to be done in
justifying the assumption or conclusion that there is only one such
concept.
Much of Stephen Perrys work is grounded on a claim similar to
the critique that could be raised against Raz: that there is more than
one tenable theory about the nature of law (grounded on different
tenable theories about the purpose of law), and the choice among
them must be made on moral or political grounds.
78
Brian Leiter
has offered a similar argument, grounded in part on Perrys work:
that there are a number of different concepts of law, more than one
of which could equally legitimately be claimed to be our concept;
given that fact, on what grounds can Raz (or anyone else) argue
that the concept he has picked out is the right one?
79
Leiter goes on
to advocate choosing the concept of law that gures in the most
fruitful a posteriori research programs (i.e., the ones that give us the
best going account of how the world works).
80
of it the best possible example of the form or genre to which it is taken to belong.
Id. at p. 52.
77
On some occasions, he makes passing references to the law of England (and
Wales), but he has not offered a distinct theory of English law.
78
See, e.g., Perry (1995, 1996).
79
Leiter (1998, pp. 545547). For arguments similar to those of Perry and
Leiter, grounded on somewhat different analyses, see Lucy (1999, pp. 7079);
Murphy (2001).
80
Leiter (1998, p. 547). Contrast Razs rejection of a functionally-driven
approach to theory construction, supra note 50.
RAZ ON NECESSITY 555
There is substantial argument in the literature about whether there
are in fact choices that need to be made among tenable theories
of law (or among the tenable purposes of law that ground these
alternative theories), and about whether such choices are necessarily
normative, or can be justied on conceptual or morally neutral meta-
theoretical grounds.
81
Razs references to the concept of law, and
even to the way concepts emerge within a culture at a particular
juncture,
82
seem to assume that there is only one concept of law
(or, perhaps more precisely, only one concept of law for us in the
present era), but the view is, of course, not self-evident.
A fuller articulation and defense of Razs ideas of the role of
necessity in jurisprudence will require participating in the debate
about whether there can ever be more than one tenable concept of
law, and whether, if there is more than one, the meta-theoretical
resources of conceptual analysis are sufcient (without recourse to
moral evaluation) to choose among the tenable concepts.
CONCLUSION
Legal theorists have come, rather late in the day one might think,
to explore seriously the methodology they use and the nature of
the claims they offer. Joseph Raz, especially in his most recent
work, has entered this meta-discussion, and offered some of the
most insightful and provocative ideas in the growing literature.
83
Two key and related questions are whether conceptual analysis
is appropriate and whether one can speak of necessary truths in
jurisprudence. Razs answer to both questions is yes. However,
it is key to understanding this claim that Razs idea of necessity
is distinctly different from what one nds elsewhere in philosophy:
sharply different from logical necessity, and almost as distant from
the type of necessity discussed in the context of Platonic philosophy
and natural kinds theories. It is perhaps a more Wittgensteinian
81
For a response to Perry, arguing that there are sufcient resources in con-
ceptual analysis to choose, see Coleman (2001, pp. 197210).
82
Raz (1996, p. 4).
83
Other important contributors to this meta-discussion have included Jules
Coleman (1998a, b, 2001, 2002), Julie Dickson (2001), Stephen Perry (1995,
1996), Brian Leiter (1998, 2002), and William Lucy (1999).
556 BRIAN H. BIX
(or Hegelian) notion, a necessity relative to a society and a time
or a way of life. However, this view of necessity requires further
elaboration, and it carries no automatic immunity to the recent
general criticisms of conceptual analysis in jurisprudence, nor does
it offer any easy answers to those challenges. Nonetheless, without
a clear understanding of what is meant by claims of necessity in
jurisprudence, we cannot begin the process of defending conceptual
analysis. And we should at least be open to the possibility that our
society contains multiple and conicting concepts of law; perhaps,
as W. B. Gallie suggested for the concepts of art and democ-
racy, our concept of law is essentially contested (grounded in
different tenable interpretations of a complex paradigm or set of
paradigms).
84
ACKNOWLEDGEMENTS
I am grateful to Jules L. Coleman, David Luban, Linda R. Meyer,
Lukas H. Meyer, Thomas H. Morawetz, Dennis M. Patterson,
Stanley L. Paulson, Joseph Raz, Brian Z. Tamanaha, and an
anonymous reader, for their comments and suggestions.
REFERENCES
Aristotle, The Complete Works of Aristotle (J. Barnes, ed., Princeton: Princeton
University Press, 1984).
Austin, J., The Province of Jurisprudence Determined (W. E. Rumble, ed.,
Cambridge: Cambridge University Press, 1995).
Bix, B., Conceptual Jurisprudence and Socio-Legal Studies, Rutgers Law
Journal 32 (2000): 227239.
Bix, B., Jurisprudence: Theory and Context, 2nd ed. (London: Sweet & Maxwell,
1999).
Bix, B., Patrolling the Boundaries: Inclusive Legal Positivism and the Nature
of Jurisprudential Debate, Canadian Journal of Law and Jurisprudence 12
(1999): 1733.
Blackburn, S., The Oxford Dictionary of Philosophy (Oxford: Oxford University
Press, 1994).
Burton, S. J., Law, Obligation, and a Good Faith Claim of Justice, California
Law Review 73 (1985): 19561983.
84
See Gallie (19551956).
RAZ ON NECESSITY 557
Cicero, M. T., Cicero: De Re Publica, De Legibus (C. W. Keyes, trans.,
Cambridge, Mass.: Harvard University Press, 1928).
Cohen, F., Transcendental Nonsense and the Functional Approach, Columbia
Law Review 35 (1935): 809849.
Coleman, J. L., Incorporationism, Conventionality, and the Practical Difference
Thesis, Legal Theory 4 (1998a): 381425.
Coleman, J. L., Methodology, in Jules L. Coleman and Scott Shapiro (eds.),
Kenneth Einar Himma (assoc. ed.), Handbook of Jurisprudence and Legal
Philosophy (Oxford: Oxford University Press, 2002), pp. 311352.
Coleman, J. L., The Practice of Principle (Oxford: Oxford University Press,
2001).
Coleman, J. L., Second Thoughts and Other First Impressions, in B. Bix (ed.),
Analyzing Law (Oxford: Clarendon Press, 1998b), pp. 257322.
Dickson, J., Evaluation and Legal Theory (Oxford: Hart Publishing, 2001).
Dworkin, R., Laws Empire (Cambridge, Mass.: Harvard University Press, 1986).
Dworkin, R., Legal Theory and the Problemof Sense, in R. Gavison (ed.), Issues
in Contemporary Legal Philosophy (Oxford: Clarendon Press, 1987), pp. 920.
Dworkin, R., On Gaps in the Law, in P. Amselek and N. MacCormick (eds.),
Controversies About Laws Ontology (Edinburgh: Edinburgh University Press,
1991), pp. 8490.
Dworkin, R., A Reply, in M. Cohen (ed.), Ronald Dworkin and Contemporary
Jurisprudence (London: Duckworth, 1984), pp. 247300.
Finnis, J., Natural Law and Natural Rights (Oxford: Clarendon Press, 1980).
Gallie, W. B., Essentially Contested Concepts, Proceedings of the Aristotelian
Society 56 (195556): 167220.
Gettier, E. L., Is Justied True Belief Knowledge?, Analysis 23 (1963): 121
123.
Glock, H.-J., A Wittgenstein Dictionary (Oxford: Blackwell, 1996).
Hart, H. L. A., The Concept of Law, 2nd ed. (Oxford: Clarendon Press, 1994).
Jackson, F., From Metaphysics to Ethics: A Defence of Conceptual Analysis
(Oxford: Clarendon Press, 2000).
Kramer, M. H., In Defense of Legal Positivism (Oxford: Clarendon Press, 1999).
Kripke, S. A., Naming and Necessity (Cambridge, Mass.: Harvard University
Press, 1972).
Leiter, B., Naturalism in Legal Philosophy, in E. N. Zalta (ed.), Stanford Encyc-
lopedia of Philosophy, http://plato.stanford.edu (Spring 2003 edition) (last
revised July 15, 2002).
Leiter, B., Realism, Hard Positivism, and Conceptual Analysis, Legal Theory 4
(1998): 533547.
Lucy, W., Understanding and Explaining Adjudication (Oxford: Oxford Univer-
sity Press, 1999).
Lyons, D., Book Review (reviewing Neil MacCormick, H. L. A. Hart), Cornell
Law Review 68 (1983): 257268.
Lyons, D., Ethics and the Rule of Law (Cambridge: Cambridge University Press,
1983).
558 BRIAN H. BIX
Mackie, J. L., Ethics: Inventing Right and Wrong (Harmondsworth: Penguin,
1977).
Moore, M. S., Harts Concluding Scientic Postscript, Legal Theory 4 (1998):
301327.
Moore, M. S., Law as a Functional Kind, in R. P. George (ed.), Natural Law
Theory: Contemporary Essays (Oxford: Clarendon Press, 1992), pp. 188242.
Murphy, L., The Political Question of the Concept of Law, in J. L. Coleman
(ed.), Harts Postscript (Oxford: Oxford University Press, 2001), pp. 371409.
Perry, S. R., Interpretation and Methodology, in A. Marmor (ed.), Law and
Interpretation (Oxford: Clarendon Press, 1995), pp. 97135.
Perry, S. R., The Varieties of Legal Positivism, Canadian Journal of Law and
Jurisprudence 9 (1996): 361381.
Putman, H., The Meaning of Meaning , in Mind, Language and Reality:
Philosophical Papers Vol. 2 (Cambridge: Cambridge University Press, 1975),
pp. 215271.
Rawls, J., A Theory of Justice, rev. ed. (Cambridge, Mass.: Harvard University
Press, 1999).
Raz, J., The Authority of Law (Oxford: Clarendon Press, 1979).
Raz, J., The Concept of a Legal System, 2nd ed. (Oxford: Clarendon Press, 1980).
Raz, J., Engaging Reason (Oxford: Clarendon Press, 1999).
Raz, J., Ethics in the Public Domain (Oxford: Clarendon Press, 1994).
Raz, J., Legal Theory, in M. P. Golding and W. A. Edmundson (eds.), Black-
well Guide to The Philosophy of Law and Legal Theory (Oxford: Blackwell,
forthcoming).
Raz, J., The Morality of Freedom (Oxford: Clarendon Press, 1986).
Raz, J., On the Nature of Law (Kobe Lectures of 1994), Archiv fr Rechts- und
Sozialphilosophie 82 (1996): 125.
Raz, J., Postema on Laws Autonomy and Public Practical Reasons: A Critical
Comment, Legal Theory 4 (1988): 120.
Raz, J., Practical Reason and Norms, rev. ed. (Princeton: Princeton University
Press, 1990).
Raz, J., The Practice of Value (Oxford: Clarendon Press, 2003).
Raz, J., Two Views of the Nature of the Theory of Law: A Partial Comparison,
Legal Theory 4 (1998): 249282.
Raz, J., Value, Respect, and Attachment (Cambridge: Cambridge University Press,
2001).
Ross, A., On Law and Justice (Berkeley: University of California Press, 1959).
Schauer, F., Critical Notice of Roger Shiner, Norm and Nature: The Movements
of Legal Thought, Canadian Journal of Philosophy 24 (1994): 495510.
Schauer, F., Fullers Internal Point of View, Law and Philosophy 13 (1994):
285312.
Soper, P., Legal Systems, Normative Systems, and the Paradoxes of Positivism,
Canadian Journal of Law and Jurisprudence 8 (1995): 363376.
Stavropoulos, N., Objectivity in Law (Oxford: Clarendon Press, 1996).
RAZ ON NECESSITY 559
Tamanaha, B. Z., Conceptual Analysis, Continental Social Theory, and CLS:
A Response to Bix, Rubin and Livingston, Rutgers Law Journal 32 (2000):
281306.
Tamanaha, B. Z., The Folly of the Social Scientic Concept of Legal Plur-
alism, Journal of Law and Society 20 (1993): 192217.
Tamanaha, B. Z., Realistic Socio-Legal Theory: Pragmatism and a Social Theory
of Law (Oxford: Clarendon Press, 1997).
Tamanaha, B. Z., Socio-legal Positivism and a General Jurisprudence, Oxford
Journal of Legal Studies 21 (2001): 132.
Waluchow, W. J., Inclusive Legal Positivism (Oxford: Clarendon Press, 1994).
Wittgenstein, L., Philosophical Investigations, 3rd ed. (Oxford: Basil Blackwell,
1958).
University of Minnesota
The Law School
229 19th Avenue South
Minneapolis, MN 55455-0400
USA