PDF - Vol - 12 - No - 02 - Complete Issue Final

GERMAN LAW JOURNAL

Review of Developments in
German, European and International Jurisprudence

Editors–in‐Chief: Russell A. Miller; Peer C. Zumbansen

Editors: Gregor Bachmann; Betsy Baker; Nina Boeger; Gralf‐Peter Calliess;
Matthias Casper; Helge Dedek; Patrycja Dabrowska; Morag Goodwin;
Felix Hanschmann; Hans Michael Heinig; Florian Hoffmann; Karen Kaiser;
Alexandra Kemmerer; Malcolm MacLaren; Stefan Magen; Ralf Michaels;

Christoph Safferling; Frank Schorkopf

www.germanlawjournal.com

© Copyright 2000 ‐ 2011 by German Law Journal GbR. All rights reserved.
ISSN: 2071‐8322 / ISSNL: 2071‐8322
Vol. 12 No. 02 Pages 599‐826 01 February 2011
Table Of Contents
Special Issue
The Many Fates of Legal Positivism

Otto Pfersmann, András Jakab & Jürgen Busch

Preface—The Many Fates of Legal Positivism 599‐600

TABLE OF CONTENTS PAGE I
Table Of Contents
N.E. Simmonds

The Nature of Law: Three Problems with One
Solution 601‐624

Mátyás Bódig

Comment on Simmonds—Legal Positivism and the
Limits of the Contemporary Legal Theoretical
Discourse 625‐662

Karen Petroski

Is Post‐Positivism Possible? 663‐692

Thomas Bustamante

Comment on Petroski—MacCormick’s Post‐
Positivism 693‐728

Alexander Somek

The Spirit of Legal Positivism 729‐756

András Jakab

Seven Role Models of Legal Scholars 757‐784

TABLE OF CONTENTS PAGE II
Table Of Contents
Christoph Kletzer

Kelsen, Sander, and the Gegenstandsproblem of
Legal Science 785‐810

Luís Duarte d’Almeida

Comment on Kletzer—Positive Law and the
“Cognitivity Thesis” 811‐826

TABLE OF CONTENTS PAGE III
Special Issue
Preface—The Many Fates of Legal Positivism

By Otto Pfersmann, András Jakab & Jürgen Busch *

Probably the most interesting debate in legal theory of the 20th century, the debate about
legal positivism, is appearing to fade away. The contributions to this special issue aim to
analyze the question as to why this is happening. We asked the authors to consider the
following hypotheses, which are partly contradicting:

(1) Positivism is (or was) only an answer to the historical challenges of industrial
societies and the nation state. In the 21st century, its plausibility has seriously
diminished.

(2) The theoretical landscape has become so diversified (there are so many
different strands of positivism) that the mere denotation of a legal theorist as
positivist does not say much about him or her. A positivist and a natural lawyer
can be nearer to each other in most of the jurisprudential questions, than two
legal positivists of different strands of legal positivism.

(3) Every argument has already been stated in the debate, so we keep repeating
ourselves (and our respective theoretical ancestors). So, it is not irrelevant, but
simply boring.

(4) Positivism has proved to be a plausible explanation on the structure of law, but
it cannot explain satisfactorily the phenomenon of adjudication. So, though not
false, it is just irrelevant to most of the questions that “really” matter for
jurisprudence. Its explanatory force is too limited.

(5) Legal positivism is based on epistemological assumptions that are outdated
today (cf. esp. Putnam). Objective truth about the law is not possible, as our
knowledge is necessarily based on improvable epistemological presumptions
(Quine).

*
Otto Pfersmann is Professor in Comparative Constitutional Law and Legal Theory, University of Paris I Panthéon‐
Sorbonne. E‐mail: otto.pfersmann@univ‐paris1.fr. András Jakab is Associate Professor in Constitutional Law,
Pázmány Péter Catholic University Budapest. E‐mail: ajakab@jak.ppke.hu. Jürgen Busch is Research Fellow in
Legal Theory, University of Vienna. E‐mail: juergen.busch@univie.ac.at.
600 G e r m a n L a w J o u r n a l [Vol. 12 No. 02
(6) Legal positivism has lost much of its attractiveness because the illusion of its
political/moral neutrality has been destroyed. The separation of law and morality
is a moral choice itself.

The above list was never meant to be a questionnaire, and the authors indeed took a
rather liberal approach in answering our original questions and they did so—we believe—
to the advantage of the final result. Partly because of this, and partly because it is usual
with topics of legal theory anyway, we did not reach any final conclusions as to whether
the above statements are just false commonplaces (and positivists conceptualizations have
thus a good explanatory force on general legal questions), or whether the above
statements are actually real and unsolvable problems, so legal positivism belongs only to
the history of legal theory. But we definitely think that the reader will be able to form a
more founded opinion after having read the contributions.

The contributions to this special issue are mostly based on papers presented at a special
workshop of the 2009 IVR World Congress in Beijing organized by the present editors.
Four colleagues were so kind to offer their participation ex post facto (Nigel Simmonds,
Mátyás Bódig, Thomas Bustamante, Alexander Somek), and thus made our electronic
conference volume more comprehensive even though we know that the topic we have
chosen offers endless issues to discuss.

The contributions follow in pairs from which the second is always a comment on the first
one. We are grateful to the editors of the German Law Journal for accepting our proposal.
We thank Lisa Giles for her help in editing the contributions.
Special Issue
The Nature of Law: Three Problems with One Solution

By N.E. Simmonds ∗

A. Introduction

Whether or not we have ourselves studied the philosophy of law, most of us are familiar
with the fact that philosophical debate concerning the nature of law has been around since
Ancient Greece. In much the same way, there have been long‐running philosophical
debates concerning justice, truth, reason and a host of other issues. The debate
concerning law is in some respects different, however. For it is not too difficult to see how
the nature of justice or truth or reason could give rise to a specifically philosophical debate,
while it is far from clear why the nature of law should generate any philosophical
puzzlement at all.

For present purposes, we might say that the hallmark of a philosophical question is to be
found in an initial sense of bewilderment that overcomes us when we first encounter the
question. It is not simply that we have been asked a question to which we do not know
the answer. Rather, we have no idea of how one might start looking for an answer, or
indeed of what sort of thing an answer might be.

Take the question of the nature of justice. We think of justice as an abstract standard that
can be invoked in the criticism or evaluation of human actions and institutions. The
content of justice is thought of as independent of all such actions and institutions: the fact
that we punish criminals does not, in itself, show that it is just to punish criminals, for
example. Nor can justice be regarded as simply a matter of conventional belief: the fact
that people believe that it is just to punish criminals does not, in itself, make it just
1
(someone who thinks otherwise has clearly failed to understand what justice is). But, if
the content of justice cannot be inferred from our knowledge of any human action,
practice or widely‐accepted belief, from whence can it be derived? We are faced by a
question (what is justice?) but have no idea how one could set about answering it, or even
what an answer would look like.
* Reader in Jurisprudence in the University of Cambridge; Fellow of Corpus Christi College.
1
Some theories might wish to claim that justice does depend upon conventional beliefs. Such theories will either
derive that conclusion from deeper non‐conventional standards or they will be skeptical theories that are really
denying that there is such a thing as justice, claiming that there are only groundless beliefs about justice. See
THOMAS HOBBES, LEVIATHAN Chapters 14–15 (1651) (illustrating the way that Hobbes' conclusions depend upon his
"Laws of Nature").

Distinctively philosophical problems tend to concern our most fundamental intellectual
and moral criteria. They may often be bound up with issues of reflexivity. For example,
they may involve questions about the status or content of a criterion, while the questions
can be answered only by reliance upon (some version of) the criterion in question.
Suppose that we want to know what truth (or reason) is. What we want to know is the
truth about truth. We want a reason for choosing this account of reason rather than that
one. Sometimes, philosophical inquiries can be clear about their own status only when the
2
principal problem of the inquiry has been successfully resolved.

When, by contrast with these debates, we turn to the philosophical debate concerning the
nature of law, everything seems at first to be different. The terrain can appear to be flat
and dull, devoid of reflexivity or any other interesting problems. Far from not knowing
how we might proceed with our inquiry, the path ahead seems all too obvious: so obvious,
indeed, that we wonder how the debate could have endured so long. For law, we assume,
is a social institution that is open to straightforward observation and description. Scholars
such as doctrinal experts, historians and social scientists know about its various features. If
you are puzzled about its nature, why not ask them? What can the philosopher have to
contribute here?

Our confidence that the philosophy of law addresses some genuinely difficult problems, on
a par with those explored in other areas of philosophy, will not be strengthened if we
examine some influential modern accounts of the core issues. Take, for example, H.L.A.
3
Hart’s account in the opening chapter of his classic work The Concept of Law. There, Hart
tells us that the long‐running debate concerning the nature of law has been driven by
three recurring issues. The first two of these issues concern the relationships between law
and morality (on the one hand) and law and “orders backed by threats” (on the other).
Hart explains that in some respects law resembles morality (it consists of a body of
prescriptions and employs a similar vocabulary to morality) and in some other respects it
resembles orders backed by threats. What we are looking for (in Hart’s view) is a clear
understanding of these resemblances and differences.

As an account of either the fundamental goal or the starting point of our inquiry this leaves
a lot to be desired. Indeed, it does not really succeed in identifying a philosophical
4
problem at all. Explaining how one thing resembles but also differs from two other things
2
See NIGEL SIMMONDS, LAW AS A MORAL IDEA 1 (2007) ("Consequently, jurisprudence can fully understand its own
status only when it has solved its central problem, by answering the question ‘What is law?’").
3
H.L.A. HART, THE CONCEPT OF LAW (2d ed. 1994).
4
This is, of course, not to deny that there is a philosophical problem in the vicinity of Hart’s remarks, but Hart has
failed to put his finger upon it in a way that can be expected to enlighten the reader.
2011] The Nature of Law 603
is not, without more, a philosophical enterprise. Until we are told more (about, for
example, the special character of the objects to be compared) the task seems to call for
careful description rather than philosophical reflection. If, therefore, we set out not
understanding how the philosophy of law can call for more than careful description of the
observable phenomena of law, Hart’s account will not help us.

In philosophy of law as in every other field of philosophical inquiry, it is both easy and
common for philosophers to conflate the intractable problems that they are meant to be
addressing with much easier problems that they know how to solve. When this happens a
lot, the inquiry can start to founder. People begin to lose any sense of what the questions
really are; or they start to think that the questions have now been solved; or they doubt
that there were really any questions there in the first place. Some brief reminders of what
those questions are can always be useful. I will offer some such reminders in the next
section of this essay.

B. Three Problems

Suppose that we begin from our initial assumption that law is a social institution. This
seems to entail the thought that laws are made by human beings. Indeed, this latter
thought seems at first so appealing that we may wonder how anyone could doubt it. Yet
we also tend to assume that human actions can create law only if they are authorised so to
do by the law itself. And here is the difficulty. For, if all laws are made by human actions,
and human actions can make law only when legally authorised so to do, we seem to be
faced by an infinite regress.

Austin tried to address this problem by abandoning the assumption that all law‐making
acts must be legally authorised: according to him, the law‐making acts of the sovereign are
5
grounded in the facts of power, not in any authorising law. This response to the problem
sounds promising but proves, in the end, not to be very successful. 6 Another response
abandons the assumption that all laws are made by human actions: it postulates “natural
laws” that are made by no one but that ground the authority of the supreme law‐maker.
At first this sounds deeply implausible, for we find ourselves asking “where could such
natural laws come from?” Once we scratch the surface of the position and start to
examine it closely, however, it begins to look more convincing.

Kant, for example, tells us that one might have a system of wholly posited laws (laws
established by a lawmaker) but a natural law (not made by any such lawmaker) would still
5
JOHN AUSTIN, THE PROVINCE OF JURISPRUDENCE DETERMINED (1832); H.L.A. Hart, Introduction to JOHN AUSTIN, THE
PROVINCE OF JURISPRUDENCE DETERMINED (1954).
6
HART, THE CONCEPT OF LAW, supra note 3; see also H.L.A. HART, THE CONCEPT OF LAW (1961).
have to precede the system, by establishing the authority of the lawmaker. 7 What Kant
has in mind is a good deal less mysterious than one might at first think. His thinking might
be explained as follows. Positive laws are invoked (by, for example, judges) as a
justification for the imposition of coercive sanctions. Laws must therefore be the kind of
thing that is capable of justifying coercion: decrees that do not in fact justify coercion
would be mere simulacra of law. If coercion is sometimes justified, there must be some
principles that determine when it is justified. These are (in Kant’s view) the principles that
8
determine when “the choice of one can be united with the choice of another.” The
totality of such principles then forms the “natural law” grounding the authority of the
supreme lawmaker. As Kant puts it, the principles “establish the basis for any possible
giving of positive laws.” 9

Kant does not think of the relevant principles as something like a complete code of human
conduct that the law merely has to reflect. Rather, legislative decision‐making is essential
to establish the full set of conditions within which “the choice of one can be united with
the choice of another.” Thus, we might say that, for Kant, the supreme act of positive law‐
making derives its legal authority from its success in realising the value (joint possibility of
choices) that is, in his view, central to the nature of law.

The first problem we have identified concerns the basis of fundamental law‐making
authority. It exhibits some of the features that are commonly found in philosophical
problems. Thus, some attempted solutions that might at first seem tempting tend to
founder in familiar philosophical dead ends, such as reductionism (e.g. Austin’s attempt to
reduce legislative authority to the facts of power) or infinite regress (if every law‐making
act is authorised by a higher law that itself was created by a law‐making act) or circularity
(if, somehow, lawmakers can authorise themselves). Meanwhile, other (potentially
superior) solutions turn out not to be self‐contained but to be tangled up with otherwise
distinct philosophical problems. This is the case with Kant’s solution. For Kant’s proposal
is intimately bound up with another puzzle that must be confronted by the philosophy of
law: the problem of law’s justificatory force.

Laws are typically invoked by judges as a reason or justification for imposing a sanction.
Any theory of law needs to accommodate this fact. For example, one of Hart’s numerous
reasons for rejecting Austin’s theory of law concerns Austin’s analysis of “legal obligation.”
Seeking clearly to distinguish law from morality (the latter being, for Austin, embodied in
the principle of utility) Austin viewed statements of legal obligation as statements
concerning the likelihood of suffering a sanction in certain circumstances. Hart rejected
7
Immanuel Kant, The Metaphysics of Morals, in PRACTICAL PHILOSOPHY 353, 379 (Mary J. Gregor ed., 1996).
8
Kant, supra note 7, at 387.
9
Id.
this analysis on the ground that it would render unintelligible the judge’s invocation of
legal obligations as a reason for imposing a sanction. 10 Hart thereby acknowledges that a
theory of law must explain how laws can intelligibly be invoked as a justification for
imposing sanctions.

This gives rise to the question of what “justification” means in this context. Is justification
here a matter of moral justification? Or can there be some form of justification that is
“technical” and “confined” in so far as it consists in the subsumability of a case under a rule
without regard to the question of whether the application of that rule to this case is
11
morally appropriate or permissible? The great philosophers of law, such as Hobbes and
Kant, tended to assume that what was in question here was some form of moral
justification. Thus, their accounts of the nature of law formed an integral part of broader
political philosophies. They asked, in effect, what must law be that it can provide a
justification for the use of coercive force? Any attempt to answer this question is of
necessity bound up with other philosophical inquiries, into justice and the authority of
states.

We have identified two philosophical problems of the nature of law: the problem of
fundamental law‐making authority, and the problem of law’s justificatory force. Now let
us note the existence of a third, which I shall call (for want of a better label) the problem of
the ideality of law. The problem arises from a simple yet pervasive feature of legal
discourse: lawyers frequently offer differing accounts of the law’s rules and principles even
when they are fully and equally familiar with all of the statutes, cases and other relevant
12
materials. Lawyers therefore seem to be working with a conception of law that cannot
straightforwardly be equated with the totality of rules explicitly established in statutes and
cases. We may therefore ask what conception of law that is, and how does it relate to the
authoritative pronouncements of the lawmakers?

10
See HART, THE CONCEPT OF LAW, supra note 3, at 84 (“[W]here rules exist, deviations from them are not merely
grounds for a prediction that hostile reactions will follow or that a court will apply sanctions to those who break
them, but are also a reason or justification for such reaction and for applying the sanctions.”).
11
H.L.A. HART, ESSAYS ON BENTHAM 266 (1982).
12
Cf. RONALD DWORKIN, LAW’S EMPIRE 3–6 (1986) (explaining the disagreement about law).
C. Hart’s Solutions

In The Concept of Law 13 and subsequent writings, H.L.A. Hart offered a set of elegant
solutions to the three problems identified above (in spite of failing, in his opening chapter,
to identify the problems in a fully enlightening way). We will briefly summarise his
solutions before turning to their deficiencies.

Let us begin with the problem of the basis of supreme law‐making authority. The problem
arises, it will be recalled, if we assume that all laws are made by human acts and that
human acts can create law only if legally authorised so to do. This seems to generate an
infinite regress. We may be unwilling to abandon the thought that all laws are made by
human acts, and may therefore be tempted to follow Austin in a reductionist approach
that grounds ultimate law‐making authority directly in the facts of coercive political power.
However (as Hart amply demonstrates) such an approach is unable to offer an adequate
analysis of law‐making authority, or of the meaning of propositions of law. Adequately to
grasp these notions we need to see law‐making authority as an authority conferred by
rules. What then is the status of these authority‐conferring rules?

Hart’s solution is to say that the rules in question are rules accepted by officials. The rules
exist in so far as the officials accept them and act in accordance with them: they are in that
sense rules created by human action (not natural laws). But the rules are not enacted by
anyone, and their creation therefore does not require legal authorisation. The solution
depends upon Hart’s notion of the “rule of recognition”: a rule, accepted by officials, that
identifies other rules (such as rules enacted by lawmakers) as “valid” rules of law. On this
account, legal rules generally exist in the sense of being “valid”: which is to say that they
satisfy whatever criteria are set out in the rule of recognition of the system in question.
The rule of recognition itself, however, is neither valid nor invalid: it is the ultimate
criterion of legal validity. It exists in so far as the conduct of officials exhibits a regular
pattern of conformity (in, presumably, deciding disputes by reference to certain rules and
sources of rules) that Hart calls the “external aspect” and the adoption of an attitude
(whereby the regular pattern of conforming conduct is viewed as a standard that one
ought to comply with) that Hart calls the “internal aspect.”

Thus, for Hart, all laws are created by human actions, but not all laws are deliberately
14
enacted. Acts of deliberate lawmaking require legal authorisation before they can
successfully create law. But the rule of recognition provides the necessary authorisation
without itself requiring authorisation.
13
HART, THE CONCEPT OF LAW, supra note 3.
14
There are perhaps some exceptions that do not affect the present argument. See HART, THE CONCEPT OF LAW,
supra note 3, at 153 (discussing the contexts where judges create law and then get their power to create it
accepted after the event). As Hart puts it, “all that succeeds is success." Id.

Hart’s solution to the first problem brings us to his solution to the second problem: that is,
the problem of the law’s justificatory force. For, in order for the rule of recognition to
exist, the officials must take “the internal point of view” towards the regular pattern of
conforming official conduct. The internal point of view incorporates the belief that the
regular pattern of conforming conduct ought to be followed, and that deviation from it
should be criticised and the criticism viewed as justified. The “ought” here need not,
according to Hart, be any sort of moral “ought”: the officials may take the view that they
“ought” to follow the rule of recognition for purely prudential reasons, such as a desire to
maintain the power of an exploitative regime from which they benefit. When officials
apply the laws to individual citizens, and regard the laws as “justifying” the imposition of
sanctions, the “justification” that is involved here can be purely “technical”: simply a
matter of showing that the individual case falls under the legal rule, without regard to the
question of whether application of the legal rule is morally right or permissible. In this
way, Hart believes that his theory captures the normative, action‐guiding, justificatory
aspect of law without needing to base law upon morality in any way.

What of the third problem, which I have called the problem of the ideality of law? Hart’s
writings contain two different, although mutually compatible, responses to the problem.
15
His principal response suggests that the assumption that law somehow goes beyond the
authoritative materials of the law is a kind of illusion that arises if we ignore the “open
texture” of rules and assume that every case can be resolved by applying the rules of the
system. In fact, as Hart points out, rules always give rise to “penumbral cases” where the
rules give no determinate answer. Cases where the lawyers appear to be disagreeing
about what the law requires are therefore, very often, cases where the rules are
indeterminate in their application to this particular case: the case cannot be resolved one
way or the other by applying the rules, and the disagreement between the lawyers is, in
substance though not in form, a disagreement not about what the law is but a moral or
political argument about how a gap in the law should be filled.

D. Where Hart Goes Wrong

Amongst Hart’s critics, the writer who has attracted the most attention is Ronald Dworkin.
Dworkin focuses his attack primarily on Hart’s account of penumbral cases. He points out
that, in such cases, lawyers think of themselves as offering legal arguments, not as offering
extra‐legal moral or policy considerations. We should try to find a conception of law that
makes sense of the lawyers’ assumptions, and only adopt Hart’s view if the lawyers’
ordinary assumptions prove to be indefensible. Dworkin believes that he has constructed
15
I set on one side for present purposes the “inclusive” aspect of Hart’s legal positivism. See HART, THE CONCEPT OF
LAW, supra note 3, at 250 (explaining the inclusive aspect of Hart's legal positivism).
a suitable account of law that captures the juridical character of doctrinal arguments in
contentious cases, and is in all other respects more enlightening than Hart’s theory.

My own theory also aims to contest Hart’s account of penumbral cases, but it opens its
attack on Hart at a different point in the argument. Instead of examining Hart’s response
to the problem of the law’s ideality, let us examine his response to the problem of the
law’s justificatory force.

Hart attempted to steer a middle course between reductionism and the rejection of legal
positivism. John Austin’s version of legal positivism was essentially reductionist: it
analysed propositions of law (such as propositions concerning legal obligation) as
statements regarding the likelihood of suffering a sanction. Reductionism of this sort, Hart
felt, was unable to capture the prescriptive, action‐guiding character of propositions of
law. Yet the alternative to reductionism, Hart felt, need not be a rejection of legal
positivism: a theory can accommodate the prescriptive or normative character of law
without grounding law in morality.

The rule of recognition is the key to Hart’s solution. For the rule of recognition to exist
(and, therefore, for a legal system to exist) the officials must adopt the internal point of
view towards their shared pattern of convergent conduct: they must regard that pattern as
a standard that “ought” to be complied with. But the “ought” need not be a moral
“ought”: officials may believe that they should follow the rule of recognition for non‐moral
reasons of self‐interest, for example. Given the rule of recognition, we can grasp the
nature of law’s normativity while setting on one side all questions about why the officials
adopt the internal attitude. Their reasons may be moral or non‐moral: it makes no
difference.

Consider the analogy of a game. I may play a game for many different reasons: to keep a
promise, to pass the time, to please a friend, to earn money, to humiliate an enemy.
Within the context of the game, however, my actions must be guided by the rules. The
rules give me reasons for action that are partially independent of my reasons for playing
the game in the first place. Thus, we do not need to say whether reasons‐within‐the‐game
are moral reasons or prudential reasons: the game is a domain of reasons that enjoys a
degree of autonomy vis a vis our moral or prudential reasons for playing the game. In the
same way, the normativity of law is a matter of the law offering reasons for action that
need not be reduced to moral or prudential reasons.

This provides us with Hart’s response to the problem of justification. When the judge sets
forth, in his judgment, the propositions of law (propositions concerning, for example, the
defendant’s legal obligations) that justify his decision, he is offering a justification that is
“technical” in so far as it invokes neither moral nor prudential reasons for the decision: it
invokes legal reasons that are not directly reducible to either of the two more familiar
categories of reason. Hart’s theory in this way hopes to enable us to understand what
these legal reasons are and how a fully intelligible human activity gives rise to them.

We may say that Hart regards law as a partially autonomous domain of normative
reasoning. It is a domain of normative reasoning, in so far as propositions of law do not
(for example) predict the incidence of sanctions or the probable reaction of courts, nor do
they describe events such as the issuing of commands or states of affairs such as the
existence of regular patterns of human conduct. Rather, propositions of law have their
sense within the realm of reasons and prescriptions, not predictions and descriptions. To
say that someone has a legal obligation is to express a conclusion about the applicability of
a rule that prescribes his or her conduct, and gives a judge a reason for imposing a
sanction.

This domain of normative reasoning is autonomous to the extent that it is not reducible to
either moral or prudential reasoning. But the autonomy is only partial for a combination
of two different reasons. In the first place, legal reasons can only possess practical force in
so far as one has reason to be guided by the law (just as the reasons internal to a game
have practical force only if one has reason for playing the game): such reasons could be
either moral or prudential. Secondly, penumbral cases cannot be resolved by exclusive
reference to the law, but only by reliance upon considerations (of morality or policy, for
example) that lie beyond the legal rules.

At this point we must call to mind one of Hart’s criticisms of Austin. Austin had claimed
that a statement to the effect that a certain person has this or that legal obligation is really
asserting that the person in question is likely to suffer a sanction if they act in a certain way
(that is, a way of acting that violates the obligation). Hart pointed out that Austin’s
analysis is unable to explain the way in which judges invoke legal obligations as a
justification for imposing sanctions. When the judge orders X to pay damages, and cites
X’s legal obligations as a justification for so ordering him, the judge can scarcely mean, by
his invocation of X’s obligations, to assert that X is likely to suffer a sanction: the likelihood
16
of suffering a sanction can hardly be offered as a justification for imposing a sanction. By
offering this argument, Hart in effect concedes that he regards it as a requirement of an
adequate theory of law that it should be able to explain how law (e.g. the defendant’s legal
obligations) can intelligibly be invoked as a justification for the imposition of a sanction.
The question that we must ask is whether Hart is really any more successful in satisfying
this requirement than Austin was.

16
See id. at 84 (“[W]here rules exist, deviations from them are not merely grounds for a prediction that hostile
reactions will follow or that a court will apply sanctions to those who break them, but are also a reason or
justification for such reaction and for applying the sanctions.”).
Let us begin with a few reminders of our ordinary understandings. We ordinarily assume
that the judicial judgment is a setting forth of the justification for the judge’s decision (a
decision which may involve ordering the deployment of the state’s coercive force against
an individual). We also assume that, within the judgment, the law will play a key role: we
assume, in fact, that the judge will cite certain propositions of law as central to the
justification for his or her decision. These rules of law are cited not because they are just
or wise or otherwise desirable, nor because the judge approves of and endorses them:
they are cited precisely because they are law within the relevant jurisdiction. Now let us
see how successful Hart’s theory is at capturing these familiar features of the judicial
judgment.

According to Hart, the justification offered by the judge for his or her decision is
17
“technical” and “confined.” When the judge invokes the defendant’s legal obligations as
a justification for imposing a sanction, the judge is operating with reasons that are internal
to the domain of law: the judge has a legal reason for imposing a sanction. Of course, for
this legal reason to possess real practical force (and so be a genuine reason for action) the
judge must have reason to guide his or her conduct by law. But such judicial reasons for
being guided by law could, Hart tells us, be either moral reasons or prudential reasons.
The judge might have purely prudential reasons of self‐interest for following the rule of
recognition, yet these would nevertheless be sufficient to give a practical force to the
propositions of law invoked in the judgment.

Suppose that a judge sentences me to prison, citing a certain rule as the justification for my
punishment. I protest and demand to know how the existence of the rule serves to justify
sending me to prison. The judge explains that the rule is a valid rule in so far as it is
derivable from the rule of recognition, a rule that he and his fellow judges accept. I
continue to protest: why should I care about what rule he and his colleagues accept?
What does that have to do with me?

Now the coherence and intelligibility of the judge’s justification will scarcely be increased if
he goes on to add further explanation along Hartian lines, explaining that he and his
colleagues accept the rule of recognition for self‐interested reasons: clearly I will feel that
the judge’s self‐interest cannot possibly provide an intelligible justification for sending me
to prison. Let us suppose then that the judge accepts a friendly suggestion from Neil
MacCormick, even if the suggestion comes at the cost of strict adherence to Hart’s legal
positivism. MacCormick pointed out that the justification provided by invocation of the
18
rule of recognition could be complete and transparent only if framed in terms of moral
17
HART, ESSAYS ON BENTHAM, supra note 11, at 266.
18
NEIL MACCORMICK, LEGAL REASONING AND LEGAL THEORY 139–40 (1994). I phrase MacCormick’s point in my own
words here, which seem to me to capture his point better than his own formulations (which leave something to
be desired).
considerations supporting the judge's acceptance of the rule of recognition (MacCormick
called these “underpinning reasons”).

The judge would now be saying “I am justified in sending you to prison because that is
required by a rule that stems from a rule of recognition I and my fellow judges accept for
good moral reasons.” The justification has now become fully intelligible. But it still fails to
capture one of the settled understandings mentioned above: the understanding that the
judgment is a justification of the decision by reference to the law and not simply by
reference to the rules that the judge considers wise or desirable or just.

At this point you may think that I have gone astray. For, surely, if the rule invoked by the
judge is derivable from the rule of recognition, it must indeed be a law. But this is not so.
For one can have a system that successfully maintains order by enforcing rules derived
from a basic rule of recognition, and yet it does not constitute a system of law. Suppose,
for example, that the rules of the system are all kept secret; or that they change on an
hourly basis; or that all of the rules, rather than prescribing duties for citizens, simply
confer upon officials extensive powers to employ coercive force whenever the officials
19
think best. We would not, I think, regard these systems as systems of law.
Consequently, we cannot say that all rules derived from the rule of recognition (in a system
that is effectively in force) are necessarily legal rules. For they will only be legal rules if the
rule of recognition from which they are derived is the rule of recognition of a system of
law. Since one of our settled understandings of the judicial judgment is that it justifies the
imposition of the sanction by invoking the law, we must say that judges can never justify
their decisions simply by invoking the rule of recognition without addressing, or at least
assuming an answer to, the question of whether the system as a whole is a system of law.

E. Reflexivity in Law

In Hart’s view, the rule of recognition represents an outer bounding limit upon the juridical
domain. Adjudicative legal thought is at its core a matter of applying the rule of
recognition and the valid rules that stem from it (beyond the core, adjudicative legal
thought is a matter of deciding, on extra‐legal grounds, matters that fall within the
penumbra of uncertainty of the rules and so are left undetermined by the existing law).
Once we reach the rule of recognition itself, we have reached the ultimate basis of legal
validity and legal justification. Of the rule of recognition itself one can ask factual
questions (e.g. how secure is its acceptance by officials?) or moral questions (e.g. is it a
good rule?) but one cannot ask juridical questions. To ask, for example, whether the rule
of recognition is “legally valid” is to ask a meaningless question: being the ultimate basis of
20
legal validity, the rule of recognition is itself neither valid nor invalid.
19
LON FULLER, THE MORALITY OF LAW (Revised ed. 1969).
20
HART, THE CONCEPT OF LAW, supra note 3, at 100–23.

Can we not then ask whether the rule of recognition is the rule of recognition of a system
of law? Such a question does not seem to lack sense for, as we saw above, one could have
a system that failed to amount to law in spite of possessing a rule of recognition. Hartian
legal positivists would in all probability agree that such a question (unlike a question
concerning the validity of the rule of recognition) is perfectly meaningful; but they would
also insist that it is essentially a classificatory question the answer to which forms no part
of the justificatory reasoning underpinning a judicial judgment. Raz, for example, has
insisted that the judge’s duty is simply one to apply the rules of the system under which he
or she was appointed, without regard to the question of whether or not the system is a
21
system of law.

The insistence that the rule of recognition bounds the limit of the juridical domain in this
way is intended, amongst other things, to police a distinction between legal doctrinal
questions concerning the content of the existing law (on the one hand) and jurisprudential
questions concerning the nature of law (on the other). Jurisprudence, in Hart’s view, does
not contribute to the justification of judicial decisions nor offer guidance to the judge: it is
an essentially descriptive enterprise concerned to note the resemblances and differences
between different phenomena (as he explains in the opening chapter of The Concept of
Law). As explained above, we might well feel a sense of dissatisfaction with this account of
the nature of jurisprudence, since it fails to identify an intellectual project of an
unambiguously philosophical character.

We now have a further reason for questioning Hart’s account of the nature of
jurisprudence. For we have seen that, if the judicial judgment is to be understood in a
manner compatible with our settled understandings, it cannot be construed as simply
relying upon the rule of recognition and its valid progeny without explicitly or implicitly
addressing the question of the status as law of the system as a whole. We ordinarily take
the judgment to be a setting forth of the justification for the judge’s decision, addressed to
the litigants amongst others; and we assume that, within such justification, the status of
certain rules as law will play a key part. Indeed, these assumptions seem to be quite
central to our understanding of the justification that is offered. A judge cannot intelligibly
justify a decision (perhaps involving the ordering of coercive force against the defendant)
by citing a rule that he accepts for self‐interested reasons. While the invocation of moral
reasons (McCormick’s “underpinning reasons”) here might reduce our sense of
bewilderment, they would still not capture our normal understanding of the judicial
judgment. For we ordinarily assume that judges do not justify the imposition of sanctions
by reference to their own moral or political views, but by reference to what they claim to
21
JOSEPH RAZ, BETWEEN AUTHORITY AND INTERPRETATION: ON THE THEORY OF LAW AND PRACTICAL REASON 84–85 (2009). For
criticism, see N.E. Simmonds, Reflexivity and the Idea of Law, 1 JURISPRUDENCE 1 (2010).
be the law. Making sense of this claim has always been one of the central tasks for
jurisprudence. 22

What this suggests is that the philosophical problem of law’s nature does not spring, as
Hart seems to think, from a need for the careful description of resemblances and
differences between different social phenomena, but from the reflexivity of legal thought.
That is to say, legal thought is always guided and informed by reflection upon the idea of
law, and it is the task of jurisprudence to investigate that idea.

F. The Idea of Law

Legal thought is guided by the idea of law. As Lon Fuller expressed the point, the law is
23
always “in quest of itself.” Many lawyers will react to this claim with incredulity. Surely,
they will say, whatever its faults, Hart’s theory that lawyers are guided by a basic rule of
recognition is more plausible than this! After all, lawyers do not think of themselves as
engaged in a philosophical inquiry, nor do they daily reflect upon such an abstraction as
“the idea of law.”

To dispel the incredulity, we should first notice that being guided by reflection upon the
idea of law is not incompatible with being guided by a basic rule of recognition. What it
does exclude is the thought that the basic rule of recognition is an outer bounding limit on
juridical thought, beyond which lie factual and moral questions, but no juridical questions.
In fact, we can only capture our settled understandings of the nature of judicial judgments
(i.e. the understandings that (i) the judgment sets forth a justification of the decision,
addressed amongst others to the litigants; and (ii) that, in that justification, the status of
certain rules as law plays a central role) if we treat the judgment as assuming that the rule
of recognition in question is the rule of recognition of a system of law. The rule of
recognition can acquire an appropriate justificatory relevance only on the assumption that
it grounds a system of law; and, in certain circumstances, this assumption could be
rendered questionable and problematic.

Once we have seen this point we start to appreciate the extent to which the concept of
“law” is not one that simply describes a familiar social phenomenon: the concept also plays
22
Might one object that, since Hart’s project is a purely descriptive one, he need not explain how the imposition
of sanctions is to be justified? Such an objection would be misguided. Hart’s theory “is concerned with the
clarification of the general framework of legal thought.” H.L.A. Hart, Preface to H.L.A. HART, THE CONCEPT OF LAW, at
vi (2d ed. 1994). Hart would readily concede that this requires him to be able to explain how laws can intelligibly
be invoked as a justification for the judicial decision, even if they do not in fact succeed in justifying the sanction.
After all, it is on precisely this basis that Hart rejects Austin’s analysis of “legal obligation” in terms of the
likelihood of suffering a sanction: one cannot intelligibly invoke the likelihood of suffering a sanction as a reason
for imposing a sanction.
23
LON FULLER, THE LAW IN QUEST OF ITSELF (1940).
an indispensable role within that phenomenon. Judges do not justify their decisions by
reference to the rules that they accept for whatever reasons, or the rules that they
consider to be wise or just: they justify their decisions by reference to the law, and the
status of the relevant rules as law is regarded as essential to the justification of the
decision. Similarly, legislators do not issue commands: they enact laws, and the fact that
their enactments are law is regarded as essential to the claim that they make upon the
citizen’s conduct. The practices of law are practices oriented towards an idea of law.

This is not to say, of course, that lawyers are constantly pondering on the idea of law. The
daily reality of law is a matter of unreflective conformity structured by a range of settled
and unquestioned assumptions: philosophical problems arise when we try to assemble
these understandings into a coherent picture of law’s nature, as we occasionally need to
do even to address the practical issues that the law confronts.

How then are we to proceed in investigating the idea of law? The philosophy of law is a
battlefield where almost no ground seems uncontested. From what starting point, then,
can our inquiry proceed?

In his Essay on Philosophical Method, Collingwood offered the following description of “a
method repeatedly used throughout the history of philosophy”:

To define a philosophical concept . . . it is necessary
first to think of that concept as specifying itself in a
form so rudimentary that anything less would fail to
embody the concept at all. This will be the minimum
specification of the concept, the lower end of the
scale; and the first phase of the definition will consist
in stating this. Later phases will modify this minimum
definition by adding new determinations, each implied
in what went before, but each introducing into it
qualitative changes as well as additions and
24
complications.

Collingwood described philosophical reflection as an activity that we are always “trying to
bring into conformity with an idea of what it ought to be.” 25 Something similar is true of
law. For legal thought, like philosophical reflection, is also always “trying to bring itself into
conformity with an idea of what it ought to be.” But a philosophy of law should not be a
product of the utopian imagination: it should be a reconstruction of the ideas and
conceptions structuring the form of association that we think of as a legal order. The
24
R.G. COLLINGWOOD, AN ESSAY ON PHILOSOPHICAL METHOD 100–01 (1933).
25
Id. at 4.
trouble, of course, is that the ground is contested, and has probably been contested as
long as there has been anything that we could recognize as legal thought. Law exists only
in so far as a great many people share certain understandings and expectations. But which
understandings are essential to the existence of law, and which are peripheral? Which
understandings are soundly based and which are misguided products of intellectual
confusion or wishful thinking? The ground having always been so hotly disputed, there
seem to be no uncontentious points from which an argument can proceed, at least if the
argument aspires to pass beyond purely banal and unhelpful propositions. Collingwood’s
observation offers us a possible procedure for addressing the problem. Can we begin by
identifying a “minimum specification” of the concept of law, in the form of a set of
conditions without which nothing could count as law at all? From such a minimum
specification, can we proceed in some orderly and intellectually defensible way towards
“later phases” of the concept, with their “qualitative changes as well as additions and
complications”? The possibility is worth exploring, however unfashionable may be the
philosophical viewpoint from which it springs.

G. From Minimum Conditions to Guiding Ideal

The philosophical trajectory described by Collingwood can be pursued from a starting point
provided by Fuller. Fuller’s famous story of Rex, and his failed attempts to enact law,
addresses the first stage of Collingwood’s method by identifying the minimum conditions
that something must satisfy in order to count as an instance of law. Through the narrative,
Fuller presents us with a series of bizarre fairy‐tale examples, and invites us to draw clear
conclusions about them. In effect he says to us: “If you encountered a system of
governance that had no rules at all, you would not regard it as a legal system, would you?
Similarly, if you encountered a system where all the rules were retrospective? Or where
they were all kept secret?” and so on. In this way Fuller identifies eight desiderata for law:

(1) there must be rules;
(2) which are published;
(3) prospective;
(4) possible to comply with;
(5) intelligible;
(6) free from contradiction;
(7) reasonably stable through time; and
(8) there must be congruence between the declared rules and the official
action.

Fuller takes these eight requirements to encompass the traditional rule of law requirement
that force should be employed against the citizen only in response to the violation of a law.

It is of great importance that one should notice the wholly uncontentious nature of the
understandings that Fuller seeks to isolate and identify here. This is why his story has the
character of a fairy tale. In the real world, we do not encounter systems where literally all
of the rules are kept secret, or where all of the rules are purely retrospective, or where
officials never act in accordance with the rules. Yet we nevertheless seem to have firm
semantic intuitions concerning such fairy tale imaginings: we clearly would not regard
them as instances of law. In this way Fuller seeks to rise above the faction‐infested
battleground of claims and counter claims concerning the propriety of labeling as law
various evil or otherwise imperfect regimes.

However, the stability of the starting point might at first seem to come at the price of
sterility. For what could be the interest in establishing that we do have stable semantic
intuitions regarding such unreal examples? Fuller’s theory becomes interesting when he
proceeds to demonstrate that the eight desiderata (identified as minimum conditions),
when taken collectively, can be regarded as a guiding ideal for legal thought: the ideal that
26
we usually label “the rule of law.” Much of Fuller’s book is taken up with showing how
legal practices represent, as he puts it, a purposive activity, and how the overall coherence
of the activity is revealed by grasping the way in which it serves the idea of compliance
with the eight desiderata.

Fuller claimed that his eight requirements represent an “inner morality of law.” This claim
has been much discussed and widely rejected. Hart’s critique of Fuller appeared to
27
suggest that the moral value of compliance with the eight requirements is wholly
contingent upon the law’s content, and that the eight requirements are more akin to
precepts of efficiency than to moral standards. This criticism has been widely endorsed as
correct.

In fact, the criticism is not correct. The eight requirements are not principles of efficacy,
but (when taken together) represent a moral ideal for legal systems. Considerations of
efficacy would, at best, provide a good reason to comply with the eight requirements to a
limited extent. If we abstract from the content of governmental objectives, we must say
that, beyond a certain point, compliance with the precepts would significantly restrain the
ability of a government to pursue its objectives, rather than advance that ability. It is not
uncommonly the case that some degree of compliance with moral precepts (such as
principles of justice) will be capable of serving a variety of morally neutral or positively
wicked goals, but this does not demonstrate that such moral precepts are really only
28 29
morally neutral principles of efficacy. Nor does the alleged compatibility with evil of the
26
See MATTHEW KRAMER, OBJECTIVITY AND THE RULE OF LAW 108–09 (2007) (disputing this idea). I think, however, that
Kramer misunderstands the import of Fuller’s observation that the ideal of perfect realisation of the eight
desiderata "is not actually a useful target for guiding the impulse towards legality." SIMMONDS, LAW AS A MORAL
IDEA, supra note 2, at 145.
27
For discussion, see SIMMONDS, LAW AS A MORAL IDEA, supra note 2, at 69–76.
28
N.E. Simmonds, Evil Contingencies and the Rule of Law, 51 AM. J. JURIS. 179 (2006); N.E. Simmonds, Freedom,
Law and Naked Violence, 59 U. TORONTO L.J. 381 (2009).
eight requirements demonstrate that they are of morally neutral character: for legal
systems, as for people, there may be distinct moral virtues, and virtue in one respect may
be compatible with lack of virtue in some other respect.

Nevertheless, it must be conceded that Fuller never really succeeded in giving a clear
explanation of the moral status of his eight requirements. It is here that I feel my own
work clarifies matters and contributes positively to the debate. It is wise to begin from
claims that have traditionally been made for the virtue of the rule of law. Most commonly,
those who have sought to extol the virtues of the rule of law have suggested that it is
intrinsically connected with liberty. But the claim has tended to attract a curt dismissal
from those legal positivists who wish to emphasise (in line with Hart’s critique of Fuller) the
absence of any necessary connection between law and a value such as liberty. They have
pointed out that the law may sometimes protect the liberty of the citizen, but may also
restrict that liberty very severely: everything depends upon the content of the law, and
upon the circumstances. Thus it is perfectly conceivable that someone living under a
regime that scrupulously adheres to the rule of law might enjoy less liberty than some
other person whose government violates the rule of law extensively.

It should be noticed, however, that this familiar argument assumes that liberty is a matter
of the number or extent (and perhaps quality) of the options one has available: to have
more options (or more extensive options, or perhaps better options) is to have more
30
liberty; to have fewer options (or less extensive or worse options) is to have less liberty.
But this well‐established approach to the notion of liberty clearly fails to capture every
aspect of the relevant value. For example, few will wish to deny that slavery is inherently
violative of liberty. Only someone dogmatically committed to a theoretical agenda could
wish to claim that the connection between slavery and lack of freedom is purely
contingent and dependent upon all of the circumstances. Yet that is the logical
consequence of saying that the notion of freedom is fully captured by the idea of “freedom
as available options.” The reason is quite simple: a slave may conceivably, in certain
circumstances, enjoy more options than a free man: consider a slave whose master gives
him very few tasks to perform, by contrast with a free man who has many onerous duties
as part of his work, or his family life, or his role as a citizen.

If we want to preserve our sense that slavery is inherently violative freedom we must look
beyond “freedom as available options” and must acknowledge that there is another aspect
to freedom. We may call it “freedom as independence from the power of another” and it
can best be understood by reference to the contrast between the slave and the free man.
Whether the slave has many options available to him or very few, those options are all
29
HART, THE CONCEPT OF LAW, supra note 3, at 207.
30
This summary description of the position obviously skates over a host of questions concerning, for example, the
kinds of interferences that will prevent an option counting as being “available.”
dependent upon the will of his master. In the case of the free man, by contrast, the
options that he has, and the duties he bears, are never fully dependent upon the will of
someone else. This is so, at least, to the extent that the free man lives under the rule of
law.

H. Freedom as Independence

To the extent that we are governed by institutions approximating to full compliance with
Fuller’s eight precepts, we enjoy a degree of freedom as independence that can be
enjoyed in no other way. Of course, the law imposes duties upon us and, to the extent
that it is effectively enforced, it restricts the options that are factually available to us, as
well as normatively available. But to comply with Fuller’s eight precepts, the rules must be
such that they are possible to comply with. It follows from this that the rules must leave
me in possession of certain domains of optional conduct, however restricted those
domains may be: even if my life is fully occupied with the performance of my legal duties, I
must still have some choices left to me if the duties are to be performable at all (e.g.
Should I whistle while performing my duty? Should I wear a blue tie?). Those available
options are independent of the will of anyone else to an extent that is impossible, within a
human community, in the absence of the institutions of law. The intrinsic moral value of
law, therefore, consists in the way in which law secures a degree of freedom as
independence that can exist in no other way.

There are two different aspects to this, and they are perhaps worth distinguishing. Any
human community will require some procedures or strategies for dealing with disputes,
and it is unlikely that such arrangements can ever be wholly optional and voluntary: even
when they are not supported by the use of coercive force they will be surrounded by
informal mechanisms of social pressure and collective disapproval, so that those who
refuse to submit to the dispute resolution process are conceived to have failed in their
duties to the community. Where the rule of law does not obtain, these dispute resolution
mechanisms subject one to the will of some or all of the other members of one’s
community. Where the rule of law does obtain, however, one to that extent enjoys a
realm that is free from any duty of submission to the wishes of others. This gives us one
aspect of freedom as independence. But we may go further. Any government of any
orderly society will need to place restrictions upon the use of violence by one citizen
31
against another. Such restrictions will inevitably provide a perimeter of protection for
those liberties that stem (as explained above) from the law’s performability. Hence, the
rule of law provides a degree of protection for the individual, not only from the power of
communal dispute resolution and other decision‐making processes, but also from the
actions of fellow citizens.

31
HART, ESSAYS ON BENTHAM, supra note 11, at 171–73.
A variety of objections might be made to my argument at this point. Some of them merit
more discussion than they can receive here, but I will nevertheless address them briefly at
this point.

I. The Will of the Lawmaker

It might be said that my argument ignores the obvious fact that, within a legal system, the
citizen’s options depend upon the will of the lawmaker. Indeed, one might see legal
systems as achieving a particularly thorough subjection of the citizen’s options to the will
of those who enact the law. The objection ignores the fact, however, that compliance with
Fuller’s eight desiderata involves the rules being prospective not retrospective. Therefore,
at any one point in time, my options will depend upon a law that is already in place, and
that law will constrain the powers that the lawmaker, along with everyone else, has in
relation to me. It might be said in response that the lawmaker could change the law for
the next moment, so that my options are pretty insecure. But it must also be remembered
that compliance with the eight desiderata requires that the rules should be reasonably
stable through time: a lawmaker who constantly changes the law departs from the eight
precepts and undermines freedom as independence to that extent.

II. The Will of Everyone

A slightly more subtle argument points out that the existence of a legal system consists in
the adoption of certain expectations, understandings and intentions by a great many
people (lawmakers, judges, other officials, citizens). If they were all, simultaneously, to
abandon their current understandings, expectations and intentions, the law would cease
to exist. Consequently, my “freedom as independence” is always fully dependent upon the
will of others, since they could, if they so chose, take it away at a stroke. This argument
fails for a quite obvious reason. My claim is not that the existence of a legal system
renders one’s options wholly independent of the will of anyone else: such a situation can
never obtain except for the type of pre‐social natural man imagined by Rousseau. Rather,
my claim is that, to the extent that the eight desiderata are satisfied, one enjoys a degree
of freedom as independence that (within the context of a human society) can be enjoyed in
no other way. This claim is not damaged at all by the objection.

III. “Freedom as Independence” Is Reducible to “Freedom as Available Options”

Here it could be argued that what I am calling “freedom as independence” is really just a
matter of the probability or improbability of my being interfered with (and so having my
options reduced) in the future. Such an argument has been offered, for example, against
32
the “republican” conception of liberty defended by Philip Pettit and Quentin Skinner.
32
See IAN CARTER, A MEASURE OF FREEDOM 237–45 (1999); MATTHEW KRAMER, THE QUALITY OF FREEDOM 135–49 (2003).
Now it is true that my conception of “freedom as independence” is quite similar to the
republican account developed by Pettit and Skinner. 33 But I have deliberately avoided
expressly associating my position with those views for two reasons. In the first place,
Pettit and Skinner are inclined at times (quite frequently but not invariably) to conflate the
dependence of a person’s options upon the will of another with that person’s vulnerability
to future interference from another. Secondly, my account of “freedom as independence”
is presented as an aspect of freedom, and an aspect that must be taken into account by
any adequate account of the nature of freedom. To the extent that the standard liberal
account of negative freedom ignores this aspect, it is deficient. But “freedom as
independence” on my account need not be viewed as a self‐standing rival to the liberal
account. In endorsing my account of “freedom as independence” one need not follow the
34
claim of Pettit and Skinner that, when it serves the interests of the governed, the law can
reduce people’s options without reducing their freedom. One could equally say that, when
the law reduces my options, it thereby reduces one aspect of my freedom, even though (to
the extent that it satisfies the eight desiderata) it may also respect another aspect of my
freedom. Liberals can therefore regard my position as a friendly supplement to their
position rather than a rival, thereby reducing the necessity for them to oppose it. But,
more fundamentally, I most certainly do not conflate the status of “freedom as
independence” with a low level of vulnerability to future interference, as Pettit and Skinner
are sometimes inclined to do. On my account, a slave is still a slave even if we can be quite
35
sure that his master will never interfere with him.

IV. Too Thin?

People sometimes find my position unsatisfactory in so far as the moral value to which I
am connecting law seems very thin and impoverished. Certainly “freedom as
independence” is not everything. Some would say it is not much. But they would be
mistaken. This aspect of freedom is real and important, and it grounds the intrinsic value
of law. Moreover, as we shall see in due course, a full and well‐grounded concern for
freedom as independence leads us into a concern for justice and principle that may go
some way towards satisfying the critic’s craving for a richer and more sustaining value at
the heart of juridical thought.

33
For other equally significant connections, see Kant’s position as explained in ARTHUR RIPSTEIN, FORCE AND FREEDOM:
KANT'S LEGAL AND POLITICAL PHILOSOPHY (2009).
34
See PHILIP PETTIT, REPUBLICANISM (1997); QUENTIN SKINNER, LIBERTY BEFORE LIBERALISM (1998).
35
I am sure that Pettit and Skinner would wish to echo this claim; my point is simply that they sometimes appear
to lose sight of it, and as a result, they seem to render themselves vulnerable to the line of attack developed by
Carter and echoed by Kramer.
V. Private Threats to Freedom as Independence?

A more interesting objection to my argument might point out that, even within a
sophisticated legal system that complies with the eight desiderata to a very high degree, I
may find my options depending upon private institutions and the will of fellow citizens. I
may, for example, enter into contracts which have that effect. This is, of course, true. But
notice once again that my argument is not that the existence of law gives one perfect and
absolute freedom as independence (assuming for the moment that such a concept of
perfection is fully intelligible): my argument is that, to the extent that the eight desiderata
are realised, one enjoys a degree of freedom as independence that can be enjoyed in no
other way. Although unsound as an objection, however, the argument points us to an
interesting feature of the rule of law, namely its implications for the regulation of private
power. Fuller was himself very concerned with the need to think about private power and
private institutions in terms of the rule of law, and his insights here have been taken up by
36
others. Indeed, the strong association between “freedom as independence” and the
need for an appropriate structure of private rights is a prominent theme in Kant’s
philosophy of law. 37

I. The Ideality of Law

Since the judicial decision must be justified by reference to the law, and since (as we have
seen) the derivability of a rule from the rule of recognition does not guarantee that the
rule is law, we cannot treat the judge’s duty as fundamentally a duty to follow the rule of
recognition. As Fuller himself emphasised, the fundamental duty of the judge is one of
“fidelity to law.” We can understand this as “fidelity to the idea of law.” The notion of
fidelity to law provides us with a vantage point from which Hart’s analysis of penumbral
cases can effectively be challenged.

Suppose that the law consist of a finite body of rules derived from the rule of recognition.
If such rules are exhaustive of the law, the law will, as Hart points out, give rise to
“penumbral cases” where there is no determinate legal answer. But might the duty of
fidelity to the idea of law have implications for the legal resolution of such supposedly
penumbral cases?

The idea of law is the idea of that set of conditions within which citizens can enjoy a degree
of freedom as independence. What then is the best that a judge can do to sustain such
conditions in cases where the established rules yield no straightforward answer? The
answer is: the judge can try to decide the case justly. To decide the case according to
principles of justice is to decide according to standards that are independent of the judge’s
36
See, e.g., PHILIP SELZNICK, LAW, SOCIETY, AND INDUSTRIAL JUSTICE (1969).
37
See RIPSTEIN, supra note 33.
will or preference (to doubt that this is possible is to say that there is no such thing as
justice).

The distinction between the core case (where the rule can be straightforwardly applied)
and the penumbral case (where it cannot) is a continuous distinction. That is to say, there
is no clear boundary between the core and the penumbra. For that reason, the judge
cannot adopt a discontinuous strategy of adjudication that requires core cases to be
decided one way (by reference to the rules) and penumbral cases to be decided differently
(by reference to justice). The judge must adopt a strategy of adjudication that addresses
all cases in the same way. It follows that the duty of fidelity to law requires the judge, in
every case, to construe the law (so far as possible) as just. The texts of the law must be
read as texts concerning justice, as if they were intended to put forward a series of
propositions concerning justice.

Thus adjudicative judgment must always depend upon the judge’s understanding of
justice, constrained by the texts that the judge must interpret. This captures many of the
features of legal reasoning that are highlighted in Dworkin’s theory of law, but without
some of that theory’s problematic features, such as the need to invoke a special value of
“integrity.” The present argument requires only the familiar values of “legality” and
“justice.”

There are other features of the judge’s duty of fidelity to the idea of law that point in a
similar direction. For example, Fuller’s requirement that the law should exhibit reasonable
stability through time suggests that, even when specific rules and doctrines change, they
should so far as possible respect more fundamental principles that can be thought to
38
underpin the law as a whole. The requirement that the law should be intelligible suggests
that it should, so far as possible, reflect moral and other understandings that will be shared
by the populace as a whole. 39 And the requirement that the law should be performable
suggests that the body of law should amount to a scheme of rules that is compatible with a
tolerable communal existence where the most basic human needs are satisfied. Thus,
what might at first appear to be the disorderly rag‐bag of considerations that inform the
interpretation and development of law 40 can nevertheless be best understood as guided
by the single idea of law. There is no point at which the judge’s duty of fidelity to law can
be said to have exhausted its implications.

38
See N.E. SIMMONDS, CENTRAL ISSUES IN JURISPRUDENCE: JUSTICE, LAW AND RIGHTS 270–74 (3rd ed. 2008).
39
SIMMONDS, LAW AS A MORAL IDEA, supra note 2, at 161. Cf. JOHN GARDNER, OFFENCES AND DEFENCES: SELECTED ESSAYS IN
THE PHILOSOPHY OF CRIMINAL LAW 45 (2007) (explaining the notion of “moral clarity”).
40
See JAMES GORDLEY, FOUNDATIONS OF PRIVATE LAW: PROPERTY, TORT, CONTRACT, UNJUST ENRICHMENT (2006).
J. Conclusion

Underlying the ancient philosophical debate concerning law are three problems. These are
not the “three recurring issues” identified by Hart, for Hart fails clearly to identify a
philosophical problem here, confusing the real issues with the search for a description of
the resemblances and differences between different social phenomena. In spite of his
preliminary misidentification of the problem, however, Hart’s theory can nevertheless be
reconstructed as an attempt to resolve the three key difficulties that I have identified.

Although subtle and ingenious, Hart’s analysis fails. His discussion of the internal point of
view, for example, explains how officials might have self‐interested reasons for following
the rules, but is unable to explain how law can intelligibly be invoked as a justification for
the sanction. Yet this latter question is at the centre of philosophical inquiry into law’s
nature: Hart himself tacitly acknowledges this when he rejects Austin’s analysis of the
notion of “legal obligation” for a similar failure to explain how such obligations can
intelligibly be invoked as a justification for the sanction.

To arrive at a better analysis, we need to reject Hart’s account of the rule of recognition
and to grasp the extent to which legal thought is reflexive: the task of determining the
content of law is ultimately guided, not by a basic rule of recognition (such a rule may play
a part, but is not fundamental), but by reflection upon the nature of law itself.
Philosophical inquiry into law’s nature should be understood as an attempt to deepen our
understanding of the guiding idea. To this end we can take as our starting point the
minimum conditions identified by Fuller, conditions which point towards the ideal of
freedom as independence. Law represents the only set of conditions within which people
can enjoy, within the context of a human society, a degree of freedom as independence.
The value is a distinct, and distinctly precious, one, and its realisation gives to the law its
justificatory force. Practices constitute law to the extent that they realise the necessary
conditions. The ideality of law is to be explained by reference to the intimate connections
between “fidelity to law” and the idea of justice. We have one solution to three problems.
Special Issue
Comment on Simmonds—Legal Positivism and the Limits of the
Contemporary Legal Theoretical Discourse

By Mátyás Bódig *

A. Introduction

The editors called for papers that help us assess the current state of legal positivism.
Although Nigel Simmonds’s paper is not a direct answer to that question, I will use it as a
starting point to bring into focus a few issues concerning the character of contemporary
legal positivism. I will seek to show how Simmonds’s legal theory can be seen as a useful
diagnosis of problems with legal positivism—and methodological positivism in particular. I
will seek to go beyond the immediate implications of Simmonds’s claims but my
conclusions will be limited to the Anglo‐Saxon discourse on legal theory.

My comments on Simmonds’s essay paper are of a slightly odd kind. This is not a critical
comment in the usual sense. I am nowhere near a critic of Simmonds’s efforts. My work
here is very much driven by the conviction that his work is systematically undervalued and
widely misunderstood by many. I will try to highlight certain characteristics of the
contemporary legal theoretical discourse that may show his works in a better light. Part of
the reason for me to think that his is a particularly promising account of law is that he is a
neo‐Fullerian, and I find it important to keep some features of Fuller’s account of law alive
in the legal theoretical discourse. I will put particular emphasis on pointing out why I think
Fuller’s legacy can be a guide for anti‐positivist legal theory.

My paper will be odd in another sense as well. Although I will characterise Simmonds as
offering a good starting point for an anti‐positivist legal theory and a successful critical
attack on legal positivism, I will assume all along that it is not possible to refute legal
positivism by mere legal theoretical arguments. It will be one of my main ambitions to
make sense of this odd position.

My analysis will be organized around three clusters of issues. In section B, I seek to
highlight the main arguments Simmonds uses against legal positivism. Then, in section C, I
assess the prospects of the kind of anti‐positivism highlighted by Simmonds. It will be
based on an account of the discursive characteristics of contemporary legal positivism and,
more generally, conceptual legal theory. In section D, I turn to discussing some of the
*
Mátyás Bódig (LL.B., Ph.D., Dr Habil., Miskolc, Hungary) is Senior Lecturer at the Law School of the University of
Aberdeen, United Kingdom.
underlying methodological and philosophical assumptions of Simmonds’s theory. It will
help me bring into focus the problem of interpretivism. That will lead to a few critical
remarks about Simmonds’s project in section E.

Although my paper is written as a comment on Simmonds’s essay published in the present
volume, I do not restrict myself to reflecting upon it. I will assume that the essay published
here is in line with the theoretical position that Simmonds developed in his recent book,
1
Law as a Moral Idea. I will rely on the book whenever I feel that we need to have a more
complete understanding of his views. But I will refrain from involving in my analysis other
works by Simmonds.

I will not try to provide a full analysis of Simmonds’s legal theory. That would deserve
more than I can provide in a comment like the present one. There are aspects of
Simmonds’s legal theory that I seek to set aside here. For example, Simmonds’s book puts
2
emphasis on the problems of doctrinal legal scholarship. Although this is the aspect of his
theoretical initiative that is the most important to me, and it may even provide
ammunition for an attack on legal positivism, 3 my analysis is not complex enough to treat
that issue adequately here. I also set aside the debate that has been going on between
Simmonds and Matthew Kramer for a number of years. It may seem strange as the debate
directly concerns Simmonds’s quarrel with legal positivism, and many people have learned
about Simmonds’s theoretical initiative from this very debate. 4 But I am afraid that the
quarrel with Kramer has a tendency to overshadow important features of Simmonds’s
account. By setting it aside, we have a better chance of having a fresh look at his anti‐
positivist initiative. I will also try to avoid making extensive comments on Simmonds’s
1
See NIGEL E. SIMMONDS, LAW AS A MORAL IDEA (2007) [hereinafter SIMMONDS, LAW AS A MORAL IDEA].
2
See id. at 164–168. This is also reflected in the essay I am commenting on. See Nigel E. Simmonds, The Nature of
Law: Three Problems with One Solution, 12 GERMAN LAW JOURNAL 2, 604 (2011) [hereinafter Simmonds, The Nature
of Law].
3
Attacks on legal positivism could proceed from the remarkable fact that positivists tend to shy away from raising
issues concerning the role and methodological character of legal doctrinal scholarship. One could say that they
more or less consciously seek to drive a wedge between conceptual legal theory and legal doctrinal scholarship
(as it is actually reflected in Simmonds’s analysis; see Simmonds, The Nature of Law, supra note 2). By the way,
this feature of positivism goes back way beyond the emergence of the Hartian account. An extreme version of
this “detachment” can be found in Kelsen who sought to purge the “pure theory of law” of doctrinal aspirations,
and condemned the work of doctrinal scholars as an illegitimate exercise of political influence. See HANS KELSEN,
INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY 3–4 (1992).
4
The main documents of the debate are the following: MATTHEW KRAMER, IN DEFENSE OF LEGAL POSITIVISM: LAW
WITHOUT TRIMMINGS ch. 2 (1999) [hereinafter KRAMER, IN DEFENSE OF LEGAL POSITIVISM]; Matthew Kramer, On the
Moral Status of the Rule of Law, 63 CAMBRIDGE L.J. 65, 65–97 (2004); Nigel Simmonds, Straightforwardly False: The
Collapse of Kramer’s Positivism, 63 CAMBRIDGE L.J. 98, 98–131 (2004); Nigel Simmonds, Law as a Moral Idea, 55 U.
TORONTO L.J. 61, 61–92 (2005); SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at ch. 3; Matthew Kramer, Once More
Into the Fray: Challenges for Legal Positivism, 58 U. TORONTO L.J. 1, 1–38 (2008); Nigel Simmonds, Freedom, Law
and Naked Violence: A Reply to Kramer, 59 U. TORONTO L.J., 381, 381–404 (2009).
2011]
Comment on Simmonds 627
“archetypal” concept of law. I do not set it aside, as it constitutes the heart of his account
of law. Leaving it unmentioned would result in a terribly distorted analysis of Simmonds’s
views. But I am more interested in focusing on the implications of his theory for the
debate over legal positivism, and that requires me to bring other aspects to the fore. I
want this comment to be mainly about Simmonds’s interpretivism. I do believe that it
gives me a better chance to indicate how Simmonds is relevant to the discussion that the
editors wanted to ignite.

Criticising legal positivism is a complex and daunting task, and not simply because legal
positivism has grown into a complex, multi‐faceted theoretical tradition. Legal positivism
is the dominant approach in contemporary conceptual legal theory. Its dominance is due
to the decisive impact of legal positivists like Herbert Hart on the agenda and the internal
norms of the contemporary discourse on legal theory. (I will have more to say about the
characteristics of the contemporary discourse below—in section C.)

Attacks on legal positivism that are not based on a careful clarification of their specific
target are bound to be defeated by the complexity of positivist theorising. And, if we want
to make good sense of Simmonds’s anti‐positivism, we will have to figure out what
Simmonds’s arguments are directed at, and in what respects they are capable of posing a
challenge against positivists. There are four possibilities of anti‐positivist initiatives that I
reckon with here. The anti‐positivists can seek (1) to refute core positivist claims (like the
“separability thesis”), (2) to undermine methodological assumptions allegedly shared by all
positivists, or (3) to show that some important legal theoretical challenges cannot be
handled by positivist accounts of law. And, of course, (4) anti‐positivists can try to refute
the main positivist conceptions one by one. Simmonds’s essay can be read, first and
foremost, as specifically directed against an individual conception: Hart’s positivist legal
theory. We will have to consider Simmonds first as an anti‐Hartian. But there are other
layers in his anti‐positivism: he can be read as making a more general claim about the
methodological deficiencies of legal positivism. As we dig deeper, we glimpse the outlines
of a more comprehensive attack on legal positivism. In section C, I will consider the
prospects of targeting the core legal positivist claims. That will allow me to consider the
viability of Simmonds’s account as a challenge against methodological assumptions
underlying many versions of contemporary legal positivism. That will help me bring into
focus Simmonds’s vision of jurisprudence as a philosophical inquiry.

B. Simmonds’s Attack on Hartian Legal Positivism and its Implications

As I have indicated, we will first read Simmonds’s arguments as an attack on Hartian legal
theory. Seen from this angle, Simmonds’s anti‐positivism revolves around a claim
concerning the inability of legal positivism to handle some of the issues that are central to
legal theory. Hart and those who follow him are incapable of accounting for the way the
law is used to justify official action. 5 Hartians fail to explain how adjudicative reasoning
can be offered intelligibly as a justification for imposing sanctions on citizens.

I. The Attack on Hart

I take this central claim to be an aspect of the issue of normativity in legal theory—the
issue of law’s ability to guide human action. It is definitely relevant for a critical
assessment of Hartian legal theory. The issue of normativity was particularly important for
6
Hart, and he definitely played a crucial role in clarifying the character of the problem. As
Simmonds puts it, Hart sought to capture the “prescriptive, action‐guiding character of
propositions of law.” 7 Hart was convinced that only legal positivist jurisprudence had the
theoretical resources to tackle this issue. It is because law’s normativity was not to be
grounded in morality. We need to maintain the conceptual distinction between law and
morality if we do not want to misconstrue the conceptual character of law and its
normativity.

The key to Hart’s account of law’s normativity was the idea of a non‐moral “ought,” and
8
the theory of “social normativity” that it gave rise to. The Hartian account of normativity
is manifested in one of his most characteristic conceptual devices: the rule of recognition.
The rule of recognition is a conventional rule (generated by the convergent practice of
officials) that sets out the criteria of legality for a particular legal system. 9 Norms can only
be regarded as being legal (being part of a legal system) if they meet the criteria of legality
set out by the rules of recognition. The insistence upon the conventional character
became the very foundation of the Hartian version of the commitment to the separability
of law and morality (separability thesis). 10 If the rules of recognition are strictly
conventional in character, their existence and content will not be a matter of moral
correctness: it will be a matter of facts—facts of “acceptance” by officials, facts of
5
This was an important theme in Simmonds book. See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 130, 135.
6
This is what makes his analysis of the “gunman situation” so important. See H.L.A. HART, THE CONCEPT OF LAW 20–
25 (2nd ed. 1994).
7
See Simmonds, The Nature of Law, supra note 2, at 608.
8
See, e.g., HART, supra note 6, at 254–57. He made sense of the idea of non‐moral “ought” (pretty successfully) by
pointing to the way in which normative claims figure in conventional practices (like games) that offer reasons for
action not reducible to moral reasons. See id. at 56–57, 140–41. Cf. Simmonds, The Nature of Law, supra note 2,
at 608–9.
9
See HART, supra note 6, at 100–10.
10
See id. at 185–86.
2011]
convergent behaviour in the context of a social practice. 11 We will be bound to conclude
that there are possible legal systems where the rule of recognition does not involve any
moral criterion of legality for any norm. 12

For obvious reasons, the Hartian strategy to establish a positivist account of law can only
work if we are led to believe that the rule of recognition is the sole determinant of the
criteria of legality, and that the legality of the rule of recognition itself cannot be called into
question. As Simmonds puts it, “[Hart’s theory] invites us to regard the normative or
justificatory force of the rules as entirely a function of their derivability from the rule of
13
recognition.”

Simmonds attacks the Hartian account at this very point. He challenges its claims about
the rule of recognition on both counts: legality is not entirely a function of derivability
from the rule of recognition, 14 and one can ask meaningful questions about the legality of
the rule of recognition. 15 It is important that Simmonds’s position does not involve the
rejection of the idea that being guided by a basic rule of recognition plays a vital role in the
operations of a legal system. 16 What Simmonds denies is that “the basic rule of
recognition is an outer bounding limit on juridical thought, beyond which lie factual and
moral questions, but no juridical questions.” 17

But what is wrong with the Hartian account of the rule of recognition? As we have already
indicated above, Simmonds challenges the Hartian account of normativity and the rule of
recognition because he finds it incapable of accounting for the reliance on law in justifying
the application of sanctions to citizens. The basic consideration can be easily identified by
asking the following simple question:

11
For this reason, the so‐called “conventionality thesis,” as formulated by Coleman, seems the best attempt to
capture the core idea of Hartian legal positivism. JULES L. COLEMAN, THE PRACTICE OF PRINCIPLE: IN DEFENCE OF A
PRAGMATIST APPROACH TO LEGAL THEORY 71 (2001).
12
See Jules L. Coleman, Negative and Positive Positivism, in RONALD DWORKIN AND CONTEMPORARY JURISPRUDENCE 28,
30 (Marshall Cohen ed., 1983).
13
SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 129.
14
See Simmonds, The Nature of Law, supra note 2, at 611; see also SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at
50, 128.
15
16
I have to admit that I would be more inclined to try another strategy that claims that the rule of recognition, in
the sense specified by Hart, does not even exist. See MÁTYÁS BÓDIG, JOGELMÉLET ÉS GYAKORLATI FILOZÓFIA 292 (2004).
But I set that issue aside for now.
17
Simmonds, The Nature of Law, supra note 2, at 613.
Suppose that a judge sentences me to prison, citing a

certain rule as the justification for my punishment. I
protest and demand to know how the existence of the
rule serves to justify sending me to prison. The judge
explains that the rule is a valid rule in so far as it is
derivable from the rule of recognition, a rule that he
and his fellow judges accept. I continue to protest:
why should I care about what rule he and his colleagues
18
accept? What does that have to do with me?

This is supposed to make us realize that Hart can offer only a stunningly inadequate
answer to a perfectly appropriate question about the legal practice (an answer that does
not even seem fully intelligible). If the criteria of legality that confer legal normativity upon
the laws applied to the citizens (and, for that matter, the laws that confer legal powers
upon the officials who make the decisions) are to be traced back to a conventional rule
generated by official action, one has to admit that invoking the rule of recognition in
explaining the justificatory force of laws boils down to little more than saying: “That is just
19
how we [officials] do things around here.” That does not really sound like an answer at
all. If we actually get this answer, we may have the impression that the Hartian did not
really understand our question, or she politely indicated her unwillingness to answer it.

I do believe that this is a solid argument. But it does not really come as a surprise: it
belongs to a family of standard anti‐positivist complaints about the Hartian theory. More
or less similar arguments have been made repeatedly against Hartian legal positivism for
over forty years. Perhaps it was Lon Fuller who could claim the credit for levelling a
20 21
version of it against Hartian positivism for the first time, and many have followed him.
The argument can be developed and contextualised in a number of ways: some of them
are more similar to Simmonds’s argument 22 than others. 23
18
Id. at 610.
19
SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 133.
20
Lon L. Fuller accused positivists of perceiving law as a “one‐way projection of authority.” LON L. FULLER, THE
MORALITY OF LAW 215–16 (2nd ed. 1969).
21
See, e.g., Michael Payne, Hart’s Concept of a Legal System, 18 WM. & MARY L. REV. 287, 287–319 (1976); R.A.
Duff, Legal Obligation and the Moral Nature of Law, 25 JURID. REV. 61, 80–82 (1980); Gerald J. Postema, The
Normativity of Law, in ISSUES IN CONTEMPORARY LEGAL PHILOSOPHY 92, 92–93 (Ruth Gavison ed., 1987); Jeffrey D.
Goldsworthy, The Self‐Destruction of Legal Positivism, 10 OXFORD J. OF LEGAL STUD. 449, 456–57 (1990); ROGER
SHINER, NORM AND NATURE: THE MOVEMENTS OF LEGAL THOUGHT 160–61 (1992); T.R.S. ALLAN, CONSTITUTIONAL JUSTICE: A
LIBERAL THEORY OF THE RULE OF LAW 64–66 (2001).
22
“The notion of acceptance [by legal officials] does not enable Hart to distinguish between a legal system based
on power and a legal system based on authority.” Payne, supra note 21, at 318.
2011]

I will not provide here an analysis of the merits of the argument (although I will return to
some of its methodological implications in section C). I have tried it elsewhere, and I have
already concluded that it poses massive problems for Hartian legal positivism. 24 Simmonds
would not be important for us just because he made this argument. It makes more sense
to concentrate our attention on how he uses this argument to develop his own account of
law, giving his anti‐positivism its unique identity.

II. Simmonds’s Anti‐Positivist Initiative

What makes Simmonds’s criticism really significant is his rejection of the standard solution
anti‐positivists tend to offer at this point. At first glance, the question of the justificatory
force of law must be settled by throwing in a moral and political justification of law’s
normative claims. Simmonds refers to this (borrowing a term from MacCormick) as relying
25
on “underpinning reasons.” Its application would look something like this: “Imposing
legal sanctions is justified because the official practice is supported by good moral
reasons.” The legal mechanisms in our society have a plausible claim to moral and political
legitimacy.

The trouble is that we do not have to think very hard to realize that going down this route
is likely to lead us into a theoretical conundrum that never worked in favour of the critics
of legal positivism. If we derive the normative force of law directly from extra‐legal moral
or political principles, it may force us to settle for a lack of independent or, at least,
26
distinctive normative force for law. The law may end up looking like a mere institutional
manifestation of the normativity of morality. And it can have uncomfortable implications.
If the normativity of law is derived directly from extra‐legal principles, it is hard to resist
the conclusion that the pertinent principles set limits to the normative competence of legal
institutions. That may encourage a moral or political “second‐guessing” of legal
23
Many prefer to formulate the arguments against legal positivism in terms of the ability of law to establish
obligations. As legal sanctions are invariably justified by the alleged breach of legal obligations, that issue also
revolves around asking questions about the justificatory use of legal propositions.
24
See Mátyás Bódig, Interpretivism and Conventionalism: Contributions to the Critical Assessment of
Contemporary Methodological Legal Positivism, in LEGAL POSITIVISM: CONCEPTUAL APPROACH 152, 167–70 (Asifa
Begum ed., 2008) [hereinafter Bódig, Interpretivism and Conventionalism]; see also BÓDIG, supra note 16, at 65–
69.
25
See Simmonds, The Nature of Law, supra note 2, at 610–11.
26
The requirement that the law must have distinctive normative force is often formulated as the “practical
difference thesis” by legal positivists, and they think it strengthens their position. See Scott J. Shapiro, On Hart’s
Way Out, in HART’S POSTSCRIPT: ESSAYS ON THE POSTSCRIPT TO THE CONCEPT OF LAW 178, 178–79 (Jules Coleman ed.,
2001); COLEMAN, supra note 11, at 69; see also Scott J. Shapiro, The Difference That Rules Make, in ANALYZING LAW:
NEW ESSAYS IN LEGAL THEORY (Brian Bix ed., 1998).
requirements: we must make sure that the decisions of legal officials are truly supported
by the underpinning reasons before we accept them as binding upon us. This may be
appealing for some but it would make almost impossible to account for some of the
familiar features of our legal practices. Some of our legal mechanisms were specifically
designed to rule out this moral second‐guessing of law. We would end up with an
unattractive account of the authority of legal institutions. 27 We know that the claim that
anti‐positivism is incapable of making sense of the distinctive normativity of law (its ability
to override at least some of our moral reasons), and that it tends towards undermining the
authority of law is a powerful motivation behind positivist jurisprudence. 28 If anti‐
positivism pins its hopes on the reliance on “underpinning reasons,” it is more than likely
to generate a positivist backlash.

The challenge is to provide an anti‐positivist alternative that avoids such pitfalls.
Simmonds attempts to meet the challenge by trying to figure out how the rules of law can
have a justificatory force “not because they are just or wise or otherwise desirable” but
29
“precisely because they are law within the relevant jurisdiction.” He seeks to make sense
of the ordinary assumption that “judges do not justify the imposition of sanctions by
reference to their own moral or political views, but by reference to what they claim to be
the law.” 30

The solution Simmonds offers is manifested in the idea highlighted in the very title of his
book: we must treat law as a moral idea. If we set out to understand the conceptual
features of law, our inquiry will give us access to a moral ideal that a society can only
realise by maintaining legal mechanisms. As I have indicated, I have no wish to provide an
extensive analysis of Simmonds’s own account of law—I would like to focus on his
philosophical vision and its implications for the prospects of anti‐positivism. But I need to
highlight here three elements of Simmonds’s account before I can move on. First, he
argues that the conceptual features of law are to be understood in light of an “intellectual
27
I do believe that that is what happened with Moore who insists that “only morality can obligate.” Michael S.
Moore, Law as a Functional Kind, in NATURAL LAW THEORY: CONTEMPORARY ESSAYS 188, 224–25 (Robert P. George ed.,
1992). For his account of authority, see Michael S. Moore, Authority, Law, and Razian Reasons, in EDUCATING
ONESELF IN PUBLIC (2000).
28
This idea is one of the central features of Kelsen’s criticism of the natural law doctrine. Kelsen claimed, pretty
unfairly, that natural lawyers cannot explain the need for a legal system alongside a system of moral norms. See,
e.g., Hans Kelsen, The Natural‐law Doctrine before the Tribunal of Science, in WHAT IS JUSTICE? JUSTICE, LAW, AND
POLITICS IN THE MIRROR OF SCIENCE 142–44 (1960).
29
30
Id. at 612–13.
2011]
archetype.” 31 Secondly, he is convinced that the moral core of the ideal the law embodies
is an aspect of freedom: “freedom as independence from the power of another.” 32
Thirdly, he links up his methodological commitment to the “archetypal view” with the
moral ideal of “freedom as independence” by way of an analysis of the eight desiderata of
the rule of law that Fuller identified in his The Morality of Law. 33 Those desiderata (or
principles) help us understand the depth of the ideal of law. 34 They facilitate an
understanding of the moral significance of law. 35 Simmonds’s anti‐positivist legal theory
claims an inherent connection between the concept of law and the ideal of the rule of law.

There is, therefore, no sense of “law” in which law can
be detached from the value that we call “the rule of
law”, or in which legality is reduced to a simple matter
of derivability from a rule of recognition. For, if it were
so detached or reduced, legality (the status of a rule of
law) would be incapable of intelligibly being offered as
36
a justification for a judicial decision.

Why would this be all that different from most other forms of anti‐positivism? Well, the
key claim that makes Simmonds’s initiative distinctive is that the law is reflexive. Although
I will have more to say about the implications of this claim in section D, a brief clarification
must be provided here. It is impossible to appreciate the character and significance of
Simmonds’s theoretical initiative without seeing clearly what he means by reflexivity. The
underlying idea is rooted in Simmonds’s vision of philosophical inquiry. He believes that
distinctively philosophical problems tend to lead us to ask fundamental questions about
31
See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 51–56. The idea is turned into the following conceptual
claim about law by Simmonds: “The status of rules as law depends on the approximation of the system of rules to
an abstract idea of law.” Id. at 130.
32
See Simmonds, The Nature of Law, supra note 2, at 617; see also SIMMONDS, LAW AS A MORAL IDEA, supra note 1,
at 158–63.
33
See Simmonds, The Nature of Law, supra note 2, at 615–17; see also SIMMONDS, LAW AS A MORAL IDEA, supra note
1, at 64–68. The desiderata are the following: (i) there must be rules; (ii) which are published; (iii) prospective;
(iv) possible to comply with; (v) intelligible; (vi) free from contradiction; (vii) reasonably stable through time; and
(viii) there must be congruence between the declared rules and the official action.
34
“Fuller’s theory becomes interesting when he proceeds to demonstrate that the eight desiderata (identified as
minimum conditions), when taken collectively, be regarded as a guiding ideal for legal thought: the ideal that we
usually label ‘the rule of law.’” Simmonds, The Nature of Law, supra note 2, at 616.
35
“The eight requirements are not principles of efficacy, but (when taken together) represent a moral ideal for
legal systems. Considerations of efficacy would, at best, give one good reason to comply with the eight
requirements to a limited extent.” Simmonds, The Nature of Law, supra note 2, at 616. This is the “entry point”
of Simmonds’s debate with Kramer.
36
See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 191.
the intellectual and moral criteria we rely on in our understanding of the world around
us. 37 For example, philosophy makes us realise that we may need to get clear about the
idea of truth before we can determine the truth value of a particular claim. Or we may
need to get clear about the idea of reason before we can choose between reasons. If law
is reflexive in this sense, it would imply that, at least in a number of strategically important
cases, we cannot determine what the law requires of us without getting clear about the
idea of law. Therefore, legal thought must be seen as guided by reflection upon the idea of
law (and the methodological character of legal theory, as a philosophical inquiry, will be
38
determined by this insight). A proper understanding of law involves more than just
becoming familiar with the mere facts of a social practice: it must be organised around an
ideal that is not detached from our moral understanding. 39 (And it makes the
philosophical character of law similar to that of justice.) 40

It is because of the reflexivity of law that anti‐positivist jurisprudence does not need to
derive the normativity of law directly from extra‐legal principles. We do not need to
impose on the law external moral and political principles to clarify the normative
significance and justificatory force of legal institutions. By reflecting on what the law is and
the practical conditions under which it can become a working institutional mechanism, we
gain insights about a moral ideal inherent to law.

III. Fuller’s Legacy

We have to get a bit clearer about the methodological “profile” of this account of law. It is
an anti‐positivist conception with obvious sympathies towards a form of natural law
41
doctrine. But Simmonds is nowhere near an orthodox naturalist or a standard anti‐
positivist. His account concentrates on the reflexivity of law in a way that one can hardly
find in the works of most naturalists. One could say that it is largely due to his reliance on
Fuller’s legal theory that becomes particularly explicit in Simmonds’s analysis of the
principles of legality.

It may not sound very promising. One could hardly say that Simmonds follows a popular
trend when turning to Fuller for inspiration. The Fullerian account of law has not been at
the forefront of the debate over legal positivism for a long time. What can we hope from
37
38
See id. at 613.
39
Cf. SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 4.
40
See id. at 1.
41
“Another response . . . postulates “natural laws” that are made by no one but that ground the authority of the
supreme law‐maker. . . . Once we scratch the surface of the position and start to examine it closely, however, it
begins to look more convincing.” Simmonds, The Nature of Law, supra note 2, at 603.
2011]
an attempt to renew it? An answer to this question could start from highlighting what
made Fuller’s account rather unpopular among anti‐positivists. Fuller was one of the first
to appreciate the significance of Hart’s legal theory, and the critical edge of his anti‐
positivism (in the last couple of decades of his career) came to be specifically directed
against Hartian positivism. 42 As is well known, Fuller’s central claim is that there is an
inherent connection between the desiderata of the rule of law (or the “principles of
legality”) and the concept of law. Fuller argued that the principles of legality constitute the
“internal morality of law,” and that this rules out the separation of law and morality that
positivists like Hart argue for. 43

Unfortunately for Fuller, positivists have never been particularly worried about his
criticism. 44 Hart developed an argumentative strategy against the Fullerian challenge in
the 1960s, and positivists are happy to resort to it whenever Fullerian considerations are
raised against them. Hart’s response is built on the combination of two ideas. The first is
that, although we must grant that that the “principles of legality” are inherently relevant
for any rule‐based mechanism for guiding human action, it does not make them inherently
moral. They are “unfortunately compatible with very great iniquity.” 45 As Simmonds aptly
put the Hartian claim, the moral quality of law is wholly contingent upon the law’s
content. 46

The second idea that positivists tend to rely on against Fuller comes from the realization
that the Fullerian account can only make sense in the context of a teleological approach to
law. It is a functional account: it seeks to characterise the law in light of morally relevant
human purposes. 47 Famously, Hart made a mockery of this feature of Fuller’s account in
his book review of The Morality of Law, 48 and it became an article of faith for most legal
42
See FULLER, supra note at 20, at 133–51; see also Lon L. Fuller, Positivism and the Fidelity to Law, 71 HARV. L. REV.
630, 630–72 (1958).
43
See FULLER, supra note at 20, at 33–34.
44
For a recent example of positivist confidence about handling the Fullerian challenge, see KRAMER, IN DEFENSE OF
LEGAL POSITIVISM, supra note 4, at 37–77 (2003).
45
See HART, supra note 6, at 207; see also H. L. A. Hart, Book Review—The Morality of Law, 78 HARV. L. REV. 1281,
1284–87 (1965) [hereinafter Hart, Book Review].
46
47
“Unlike most modern theories of law, this view treats law as an activity and regards a legal system as the
product of sustained purposive effort.” FULLER, supra note 20, at 106. Cf. Lon L. Fuller, Human Purpose and
Natural Law, 53 J. OF PHIL. 697, 697–705 (1956).
48
“The author has all his life been in love with the notion of purpose and this passion, like any other, can both
inspire and blind a man. I have tried to show how it has done both to the author. The inspiration is so
considerable that I would not wish him to terminate his longstanding union with this idée maîtresse. But I wish
that the high romance would settle down to some cooler form of regard. When this happens, the author’s many
positivists that the conceptual features of law cannot be revealed in light of a particular
function, point or purpose. The law serves to tackle a wide variety of social problems and
challenges—it is characterised by an irreducible functional plurality. It does not have a
specific purpose. Instead, it has distinctive ways to serve human purposes: there may be
distinctively “legal” ways for tackling practical problems. 49 Therefore, if we want to clarify
the conceptual character of law, we have to concentrate on the structural features of a
social phenomenon. 50 The main elements of the Hartian account (the difference between
primary and secondary rules, the rule of recognition, etc.) must all be seen as elements of
such a “structural” account of law.

Hart’s answer to Fuller is one of the best expressions of the spirit of methodological legal
positivism. 51 It is driven by the conviction that one needs to be a positivist for
methodological reasons—to have an adequate grasp of the nature of law. Positivists are
the ones who got clear about the methodological requirements for an adequate
conceptual account of law. They are the ones who do not mystify the law. 52 They know
that one of the characteristic ways of messing up conceptual inquiries can be found in
functional or teleological accounts of law. When we try to elucidate the law in light of its
moral point or function, we tend to overlook the fact that the central problem for
conceptual legal theory (“what is law?”) is descriptive in character. We mystify the law
instead of clarifying it. Legal positivists are there to warn us against this mistake.

Of course, the positivist counter‐attack did not scare off the anti‐positivists who still
preferred to provide accounts of law in light of morally relevant functions. But it convinced
most of them that they should distance themselves from Fuller’s strategy. Nowadays, they
are more likely to follow Dworkin, and insist that law is to be characterised in light of its
readers will feel the drop in temperature; but they will be amply compensated by an increase in light.” Hart, Book
Review, supra note 45, at 1296.
49
See Leslie Green, The Political Content of Legal Theory, in 17 PHILOSOPHY OF THE SOCIAL SCIENCES 1, 12–13 (1987);
Jules L. Coleman, Methodology, in THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW 311, 337–38
(Jules Coleman & Scott Shapiro eds., 2002).
50
This is what came to the fore in Hart’s replies to Dworkin. He insisted upon the fundamental importance of the
theoretical task of answering “important questions to which the existence of law has always given rise, and which
are not questions of moral or political justification but concern the structure or constitution and interrelationship
of legal phenomena.” H.L.A. Hart, Comment on Dworkin, in ISSUES IN CONTEMPORARY LEGAL PHILOSOPHY 37 (Ruth
Gavison ed., 1987) (emphasis added); see also HART, supra note 6, at 240.
51
I took the term from Stephen R. Perry, see Stephen R. Perry, The Varieties of Legal Positivism, 9 CANADIAN J. OF L.
& JURIS. 2, 361, 361–81 (1996); see also Stephen R. Perry, Hart’s Methodological Positivism, 4 LEGAL THEORY, 427,
427–67 (1998). For my account of methodological positivism, see BÓDIG, supra note 24, at 158–61.
52
This commitment to demystification is what Hart admired in Bentham. See H.L.A. HART, ESSAYS ON BENTHAM:
JURISPRUDENCE AND POLITICAL THEORY 21–39 (1982); see also H. L. A. Hart, Bentham, Lecture on a Master Mind, in
MORE ESSAYS IN LEGAL PHILOSOPHY: GENERAL ASSESSMENT OF LEGAL PHILOSOPHIES 18, 27–33 (R. S. Summers ed., 1971).
2011]
point because law is an interpretive concept. 53 Or they may follow Finnis in trying to revive
the classical tradition of the natural law doctrine, and claim that a moral function makes it
practically reasonable for a community to have a legal system. 54 Or they may put their
faith in Michael Moore’s claim that law is a functional concept by virtue of its semantic
character. 55

However, there is a stubborn problem with all these attempts to make the anti‐positivist
position more resilient (and it happens to be closely related to the difficulty involved in
relying on “underpinning reasons” in explaining the normative force of law). All these
accounts end up being distanced from an important aspect of Fuller’s effort: Fuller sought
to establish a connection between law and morality by remaining within the realm of
ordinary legal experiences. Fuller tried to reveal the internal morality of law by reflecting
on uncontroversial features of governance by rules in general, and the alternative
strategies gave up this aspiration. By contrast, Dworkin requires us to impose upon law
political values that are formulated by moral and political reflection—in the process of
56
developing one’s own moral and political theory. It is even more obvious in the case of
Finnis whose key claim is that the legal theorist cannot find the proper point of view to
grasp the conceptual characteristics of law without figuring out the principles of practical
reasonableness. 57 His methodology requires us to leave the field of jurisprudential
reflection right after clarifying the methodological character of legal theory. We have to
lay the foundations for a whole practical philosophy (including a complete theory of
values), and arrive at an account of law at a later stage: by way of a specification of our
moral and political philosophy.

The trouble is that such a long detour towards moral and political philosophy is likely to
commit any account of law to a particular political perspective, thereby generating a
suspicion of political partisanship. It gives rise to the recurrent complaint that mainstream
58
versions of anti‐positivism tend to be terribly ideological. That can be unappealing for
53
See, e.g., RONALD DWORKIN, LAW’S EMPIRE 87–89 (1986).
54
See JOHN FINNIS, NATURAL LAW AND NATURAL RIGHTS 16 (1982) [hereinafter FINNIS, NATURAL LAW]; see also John Finnis,
Law as Co‐ordination, 2 RATIO JURIS 97, 97–104 (1989).
55
See supra note 27; see also Michael S. Moore, Moral Reality Revisited, 90 MICH. L. REV. 2424, 2424–533 (1992).
56
See, e.g., RONALD DWORKIN, TAKING RIGHTS SERIOUSLY 105–07 (1978). The tendency of imposing political values on
law is visible in the way the analysis of “integrity” leads Dworkin into a political philosophical discussion. See id. at
176–224.
57
See FINNIS, NATURAL LAW, supra note 54, 14–15.
58
Finnis and his disciples (like Robert George) are the most obvious targets of such criticism. For them, a
conceptual account of law is inherently linked to laying the foundations for arguing about controversial political
issues (like abortion, euthanasia, or homosexual emancipation). See, e.g., John M. Finnis, Law, Morality, and
Sexual Orientation, 9 NOTRE DAME J.L. ETHICS & PUB. POL’Y 11, 11–39 (1995); Robert P. George, Public Reason and
Political Conflict: Abortion and Homosexuality, 106 YALE L.J. 2475, 2475–504 (1997). But Dworkin can also be
legal theorists who are worried (like me) that, by channelling conceptual inquiries through
a conception of substantive moral and political principles, we turn conceptual legal theory
(and legal theory in general) into just another battleground for fighting out our moral and
political differences.

Once again, it gives a considerable advantage to legal positivists. Although legal positivism
(especially in its early forms) was often formulated in support of partisan politics (Bentham
is an obvious example), it later gained much of its credibility as an intellectual movement
by seeking to resist external ideological pressures. It is damaging for anti‐positivism if it is
59
seen as facilitating unwanted ideological influences on legal theory.

This is why one may think that the Fullerian initiative deserves another chance. It may turn
out to be an advantage at the end of the day, that it was not an attempt to clarify how the
constitutive principles of one’s moral and political outlook are reflected in law. Instead, it
was an attempt to reveal the morality that is internal to law. If we can renew the Fullerian
initiative in some way, we have a chance to undermine the common misconception that
the only effective guarantee of the integrity of law and legal thought is the positivist
separation of law and morality.

Of course, the Fullerian account must be strengthened and reconstructed in many
respects. We have to acknowledge the problem that, in terms of the quality of his
philosophical reflection, Fuller was constantly on the back foot in his debate with Hart.
There is a need here to ground the Fulleran ideas in a more appealing philosophical
methodology, and to relate them to a better understanding of morality and moral
reflection. It is this particular “overhaul” of the Fullerian account that Simmonds seeks to
60
achieve.

I will return to analysing some of the philosophical characteristics of Simmonds’s legal
theory in section D. But, before that, we must come to terms with some other aspects of
the quarrel between Simmonds and legal positivists.
vulnerable to accusations of political partisanship; see Richard Posner, The Problematics of Moral and Legal
Theory, 111 HARV. L. REV. 1637, 1637–717 (1998).
59
Of course, the issue for legal theory is not finding ways to make accounts of law non‐ideological. That is
unrealistic as legal practices (just like any other human practice) have their ideological features. But it can be
realistic (even important) to try to avoid imposing on law ideologies that are external to legal practices—that are
not inherently connected to the professional culture of lawyers. Giving up resistance on this point would imply
giving up the claim of the legal profession and legal scholarship to integrity. And the claim to professional
integrity is dear to the heart of lawyers and legal scholars. They are unlikely to sympathise with theories of law
that deliberately give it up.
60
“Nevertheless, it must be conceded that Fuller never really succeeded in giving a clear explanation of the moral
status of his eight requirements. It is here that I feel my own work clarifies matters and contributes positively to
the debate.” Simmonds, The Nature of Law, supra note 2, at 617.
2011]

C. Legal Positivism and the Contemporary Legal Theoretical Discourse

In the previous section, we considered Simmonds’s legal theory as a challenge against
Hartian legal theory. It emerged as an attempt to show that Hartian legal theory is
incapable of handling a vital aspect of law: its normativity, or, more specifically, its
justificatory potential. Of course, Simmonds’s central argument about justificatory
potential is limited in its scope in a number of ways. It is designed to challenge a form of
methodological positivism, and it puts heavy emphasis on the Hartian concept of the rule
of recognition which happens to be based on a form of jurisprudential conventionalism.
For such reasons, it may not be effective against normative positivists (like Waldron and
61
Tom Campbell). It may not work against Kelsenian normativists who do not rely on
conventionalism. Raz may also be ruled out as a legitimate target: he is one of those who
called into question the ability of the Hartian account to clarify the normativity of law. Raz
criticised Hart for not realising that officials must be able to think that their decisions are
supported by a justification that generates practical reasons for the ordinary citizens. 62

Shall we accept that Simmonds can lay claim only to attacking a particular group of
positivists? Or is there an aspect to his anti‐positivism that can be turned against at least
some other positivists who did not follow Hart’s path? In order to figure that out, we will
need to consider whether some of Simmonds’s arguments work on another level, as a
different type of challenge against legal positivism. I do believe that Simmonds has the
resources to pose a more general challenge to legal positivism. And we glimpse that
aspect to his theory if we begin to think about the way contemporary positivists tend to
react to critical challenges.

When it comes to defending legal positivism in general, positivists tend to resort to a
distinctive argumentative strategy: they determine the identity of legal positivism by
formulating core conceptual claims that the positivists are supposed to agree upon (like
63 64
the “separability thesis,” the “sources thesis,” or the “limited domain thesis” ).

61
See Jeremy Waldron, Normative (or Ethical) Positivism, in HART’S POSTSCRIPT, ESSAYS ON THE POSTSCRIPT TO THE
CONCEPT OF LAW 410 (Jules Coleman ed., 2001); TOM D. CAMPBELL, THE LEGAL THEORY OF ETHICAL POSITIVISM (1996); see
also Tamás Győrfi, Normatív pozitivizmus, in NATURA IURIS: TERMÉSZETJOGTAN ÉS JOGPOZITIVIZMUS ÉS MAGYAR JOGELMÉLET
(Miklós Szabó ed., 2002).
62
See JOSEPH RAZ, PRACTICAL REASON AND NORMS 56–57 (2nd ed., 1990); Joseph Raz, Hart on Moral Rights and Legal
Duties, 4 OXFORD J. OF LEGAL STUD. 123, 130 (1984). For Raz’s ideas on how this limits the force of certain anti‐
positivist arguments against his account, see Joseph Raz, Postema on Law’s Autonomy and Public Practical
Reasons: A Critical Comment, 4 LEGAL THEORY 1, 19 (1998).
63
See, e.g., JOSEPH RAZ, ETHICS IN THE PUBLIC DOMAIN 194 (1994).
64
See, e.g., Frederick Schauer, The Limited Domain of the Law, 90 VA. L. REV. 1909, 1914–15 (2004).
The question for us may seem to be whether Simmonds can be successful in undermining
those core positivist claims. But that is not really the case. As I will try to show, no one can
be successful in demolishing the positivists’ core conceptual claims. But one can try to call
into question the argumentative strategies that generate them, and the discursive
constructions that facilitate the reliance on them in the legal theoretical discourse. It may
be that arguing against legal positivism has become particularly tricky in the contemporary
legal theoretical discourse but it is also possible that it is not exactly a sign of the strength
of the positivist position. In some respects, it may be a sign of serious discursive
dysfunctions. I believe that Simmonds has a claim that deserves particular attention in this
context: the point he makes about the “fragmentation of inquiry.” So the questions that
help us assess the anti‐positivist challenge Simmonds raises are like this: How can the
strategy of defending core conceptual claims make sense in a discourse? What does it tell
us about the discourse of legal theory? Do we have a chance to challenge the discursive
constructions that facilitate such argumentative strategies? These are the issues to be
pursued in this section.

I. Core Conceptual Claims and the Defence of Legal Positivism

If we want to assess the significance of Simmonds’s anti‐positivist initiative, we need a
deeper understanding of the positivist strategy of defending legal positivism by insisting
upon the plausibility of core conceptual claims. It is no surprise that the strategy is really
popular amongst legal positivists: it simplifies the task of the apologist who does not have
to defend legal positivism as a complete account of the conceptual character of law or a
complete conception of law. More importantly, the core claims are so modest that they
are bound to be true in some sense. They also prove to be frustratingly elusive for the
critic. The critic is forced to go against particular formulations of the core claims (to
provide a serious analysis with properly set objectives) but, naturally, they do not have
canonical formulations. Even if the arguments are successful against certain formulations
of the core claims, it is hard to imagine that the positivists cannot salvage them in some
revised and qualified form.

The way this legal positivist strategy works can be seen particularly clearly in Coleman’s
classic rejoinder to Dworkin, “Negative and Positive Positivism.” Coleman’s negative
65
positivism revolves around a core conceptual claim: the separability of law and morality.
It is formulated as a very modest claim that is very hard to dispute (in fact, Coleman
believes that its truth is more or less beyond doubt). 66 It is designed to make us accept
65
See Coleman, supra note 12, at 30–31.
66
Here is what Coleman defends as the specification of the separability thesis: “[T]here exists at least one
conceivable legal system in which the rule of recognition does not specify being a principle of morality among the
truth conditions for any propositions of law.” Id. at 31. Although I am not sure that that this particular
formulation is beyond doubt, I am ready to admit that the separability thesis is true in some (pretty unimportant)
sense.
2011]
that the positivist has a plausible point of departure—unlike the anti‐positivist who will
have to struggle with the burden of making sense of a claim (the denial of separability) that
looks pretty implausible from the outset. It gives the impression that, whatever they say in
the end, positivists remained faithful to an otherwise true claim all along. Of course,
negative positivism is not even close to a full account of law. It becomes an important
thesis for legal theory only if we see it as a component in a broader set of claims—in
versions of what Coleman calls “positive positivism.” 67 Naturally, a decent positivist will
seek to develop a form of positive positivism. But, when it comes to defending legal
positivism against the critics, she can always rely on the “safety net” of negative
positivism. 68

The positivist strategy tells us a lot about the character of a legal theoretical discourse that
was very much shaped by legal positivism. I will try to highlight some of its implications by
providing an illustrative analysis of the separability thesis. (I opt for the separability thesis
because it will allow me to make use of what I have said in the previous section.) We know
that the thesis concerns the conceptual relationship between law and morality. Of course,
one can speak of the separation of the two in many senses: the positivist has a wide
variety of ways in which she can specify the thesis for legal theoretical use. And the
specification will fall somewhere between an (indefensible) strong and an (uninformative)
weak understanding.

The indefensibly strong understanding would make the claim imply that the law has
nothing to do with morality: we can explain every conceptual element of law without ever
making a reference to morality, and we can practice law without ever taking moral
69
considerations into account. Positivists are unlikely to make this strong clam. They are
70
happy to admit that law and morality “meet” at several points. At the other end of the
spectrum, we find a thesis that would only imply that law and morality are not the same:
they both have conceptual features that the other lacks. Obviously, the claim in this form
67
See id. at 32.
68
The confidence in this strategy often generates a form of triumphalism among positivists. Many of them think
they won the debates with anti‐positivists long ago. It justifies the dominating position of legal positivism in legal
theory. The current problem of legal positivism is that it is becoming the victim of its own success: it needs to
find a new function for itself. Brian Tamanaha provides a good example for this attitude. He quotes, with
genuine sympathy, a speech by Brian Bix: “[L]egal positivism is orthodoxy in desperate need of dissent.” Brian Z.
Tamanaha, The Contemporary Relevance of Legal Positivism, 32 AUSTRALIAN J. OF LEGAL PHIL. 1, 1–2 (2007).
69
It is not that positivists are never tempted to make a strong claim of separation. Kelsen went pretty close to it
when he denied the possibility of any real conflict between law and morals. The perspectives of morality and law
mutually exclude one another. “[J]udging from one point of view excludes the other . . . . It is evident to any
jurist—that is, when the cognition of legal norms is involved—he must disregard the moral aspect. No moralist
would think of letting considerations of positive law interfere with the validity of norms which he has recognized
from his point of view.” HANS KELSEN, GENERAL THEORY OF LAW AND STATE 410 (1949).
70
Cf. MATTHEW H. KRAMER, WHERE LAW AND MORALITY MEET (2004).
is way too weak: it could not confer upon legal positivism a theoretical identity. It would
be impossible to present this claim as something that anti‐positivists are bound to deny.
The thesis in this form is eminently true as law (in all its incarnations) is a social artefact, 71
and is built around institutions (legislatures, courts, regulative agencies, law firms) in a way
that is alien to morality (in all its incarnations). Everyone in her right mind would accept
this, including the classics of the natural law tradition. Hence, legal positivists are equally
unlikely to formulate the thesis along these lines.

The positivists need a specification of the thesis that can be taken as more or less
straightforwardly true but not without important substantive implications. They will insist
72
on the separability of law and morality in a more specific way. Typically (at least in the
Hartian camp), they tie the claim about separability to the concept of legal validity. They
are likely to claim, in some form, that it is a conceptual truth that the moral properties of a
norm are not necessarily responsible for its status as a legal norm. 73

At first, it sounds like an important claim that is capable of conferring upon legal positivism
a distinctive theoretical identity. It says something important about law, and it seems to
define the challenge for the anti‐positivist critic very clearly: she must argue that the
criteria of legality must have something to do with morality in any legal system. But it
cannot be the end of the story. If we leave the thesis like this, it will be a relatively easy
prey for the anti‐positivist critic. In fact, Simmonds’s argument concerning a justificatory
aspect of law (that we have analysed in section B), can fulfil that task pretty neatly. The
critic may point to the inherent connection between the validity of law and its justificatory
force. The fact that a legal pronouncement is valid is relevant for the officials exactly
because it is a prerequisite for its use in justifying legal decisions. Valid law is supposed to
gain practical weight (to become able to guide human action), and that will not happen if
the valid law cannot interact with the citizens’ own reasons for action—moral reasons
amongst them.

The positivist must neutralise this argument. Luckily for her, she can try to exploit the
elusiveness of her core claim. She may add further specifications that meddle with the
conceptual connection between validity and justificatory force. An intelligent positivist will
see clearly that her problem is rooted in our settled intuition that the practical relevance of
the criteria of legal validity lie in their ability to identify for us normative propositions with
legal force (that is, the force to justify official action and legal claims upon fellow
71
See Eleni Mitrophanous, Soft Positivism, 17 OXFORD J. OF LEGAL STUD. 621, 622 (1997).
72
See John Gardner, Legal Positivism: 5½ Myths, 46 AM. J. JURIS. 199, 222–24 (2001).
73
See Coleman, supra note 12, at 30.
2011]
citizens). 74 Then, the obvious solution seems to be severing the tie between validity and
normative force. And the positivist can try to achieve that by redefining her own
theoretical ambition, setting for herself a more modest task concerning the clarification of
law as a normative phenomenon. She may claim that the issue of normativity has several
aspects, and only some of them fall within the remit of conceptual legal theory that legal
positivism should care about. She can argue for a need to separate out the features of
normativity that cause problems for positivism, and declare them issues to be handled by
substantive moral and political philosophy.

Once again, Herbert Hart leads the way in developing this strategy. A cunning way of
severing the ties between the validity of law and its practical weight is right there in the
75
arguments he deployed against Fuller and Radbruch. Legal validity identifies for us the
rules of the law—the substantive normative claims that law makes on its addressees. But
validity is not inherently linked to obedience, justificatory force or practical weight. One
can meaningfully say that “I recognize this provision as valid law but I refuse to obey it.” It
seems to indicate that that the issue of determining the actual practical weight of law is
completely separate from the issue of its validity. And, to complete the argument, the
positivist only needs to show that conceptual legal theory must concentrate on the issue
76
validity rather than the issue of reasons for compliance (and to imply that consideration
in the separability thesis).

The way this develops into a coherent strategy in support of legal positivism can be seen
pretty clearly in Coleman’s works. He distinguishes three steps in a complete account of
the normativity of law. 77 First, one must point out what normative guidance actually
consists in. It requires no more than an abstract characterisation of the “meaning” of
normativity. Secondly, one needs to figure out how the law in particular seeks to guide
behaviour (how the law seeks to make a practical difference to our behaviour). It is still a
matter of specifying the “meaning” of normativity, and it does not require us to say that
the law actually manages to guide human action. Thirdly, one must explain how the law
can motivate people to obey it. It is only in this third step that we encounter the issue of
the actual practical weight of law. If we accept that the explanation of normativity involves
taking those three steps, and that the first two can be independent of the third, we can
74
By the way, this is the very intuition that makes Simmonds’s rejection of identifying the criteria of legality with
derivability from a rule of recognition work.
75
See HART, supra note 6, at 207–12; H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 HARV. L. REV.
593, 615–21 (1957) [hereinafter Hart, Positivism and the Separation]; see also Frederick Schauer, Positivism as
Pariah, in THE AUTONOMY OF LAW: ESSAYS ON LEGAL POSITIVISM 31 (Robert P. George ed., 1996).
76
And this may not sound implausible at all. For a legal theory, the issue of validity seems more important.
Lawyers and legal scholars, in the context of professional discourses, are more likely to raise issues of validity than
issues of obedience.
77
See COLEMAN, supra note 11, at 69–73.
hardly object to accounts of law that restrict themselves to answering the first two
questions—leaving the third one to moral and political philosophical reflection. And then,
the positivist can say to a critic like Simmonds that his question about the law’s ability to
justify official action or legal claims made upon fellow citizens should not be addressed to
the legal theorist—it should be addressed to the political philosopher. 78 As a matter of
conceptual legal theory, the positivist account is adequate and plausible. The separability
thesis is saved.

II. What Is Wrong with the Reliance on Core Conceptual Claims?

I emphasise that I have provided only an illustration. But it may be enough to indicate how
the positivist strategy of relying on core conceptual claims works. And I believe that one
really has to admire it. Against this strategy, the critic faces an uphill battle. It would be
ineffective (and possibly disingenuous) to insist that the positivist has chosen the easy path
by setting aside aspects of conceptual issues that are difficult to handle within her account.
Everybody relies on similar ways of regimenting issues to her own favour. And the
complaint that the positivist deliberately made her account of normativity incomplete is
also unlikely to cut much ice. Nothing prevents a theorist from picking out aspects of more
79
complex problems for closer analysis. Conceptual legal theory is about clarifying
preliminary issues anyway. We can hardly think that there is something in the character of
law that excludes the positivist strategy of separating issues of validity and actual
normative force on the one hand, and issues of conceptual legal theory and substantive
practical philosophy on the other.

So why not admit that this is not simply a cunning strategy but a proof that legal positivists
are right? Why not accept that they can defend their core claims successfully because
those core claims are actually true? It may take some time before they find the best way
to formulate them, but positivists are definitely on to something, and that explains the
staying power of their theoretical outlook.

Well, I am ready to admit that, as a matter of propositional truth, legal positivism (if it is
defined in terms of core conceptual claims) cannot be refuted. But it is not enough to
accept that we should give in to legal positivism. What follows from the apparent success
of the positivist strategy is that we should not carry on struggling with the core conceptual
78
Cf. id. at 118. I had a personal experience that serves as a nice illustration of this point. A few years ago, I had a
conversation with a prominent positivist theorist. I outlined for him a few familiar arguments about the inability
of legal positivism to account for the ability of law to establish obligations. He answered that he agreed with me
as a political philosopher but not as a legal theorist.
79
Gardner admits that legal positivism is “not a whole theory of law’s nature, after all.” Gardner, supra note 72,
at 210.
2011]
claims. 80 The anti‐positivists have a much better chance of making progress if they try to
point out how hollow the success of positivist defence of core claims really is.

One may think that the best way to do it is by pointing out that the positivist strategy often
resorts to twisting the character of key legal concepts to make the positivist account look
plausible. The concept of validity is a case in point. As we have indicated, the validity of
law figures in the professional discourse on law as an eminently practical concept. The fact
that a statutory provision is valid (or, more exactly, that it is in force) means for the
practitioners that it establishes duties and rights, and that it has direct justificatory
81
implications. There is not much room here for separating the issue of validity from the
practical weight of legal propositions. What lawyers say about validity does not really
allow for dithering about the issue of obedience. 82 The positivist account may provide a
misleading picture of law. It does not seem to be designed to explain certain features law
as we find them—it looks carefully tailored to make a particular theoretical approach look
plausible.

But this is hardly the best way to go about the task of developing an anti‐positivist
argument here. There is no conclusive argument that rules out the positivist accounts of
validity. Perhaps, a bit of twisting of legal concepts is tolerable in the interest of revealing
truths about law. The critics must do better. They should pay more attention to the
discursive characteristics of contemporary conceptual legal theory, and the factors that
allow the legal positivists to play a central role in it. They should try to show that the
cunning strategy of meddling with the issues of conceptual legal theory to the legal
positivists’ favour may serve to maintain the influence of legal positivism in a particular
discourse but it is likely to be detrimental to conceptual legal theory as a scholarly
discourse with a claim to intellectual credibility. The positivist strategy runs the risk of
pushing the conceptual debates about law towards petty quarrels revolving around a
83
handful of elusive claims with very limited explanatory force.
80
I must admit that I have fallen for the temptation to target the core positivist claims several times. See Bódig,
Interpretivism and Conventionalism, supra note 24, at 161; see also Mátyás Bódig, A jog és az erkölcs közötti
viszony, 2(2) MISKOLCI JOGI SZEMLE 5, 5–32 (2007).
81
This understanding of validity is actually reflected in the works of earlier positivists like Kelsen who were less
concerned about anti‐positivist challenges, and did not try to sidestep a whole dimension of the issue of the
normativity of law. “Validity of law means that the legal norms are binding, that men ought to behave as the legal
norms prescribe, that men ought to obey and apply the legal norms.” KELSEN, supra note 69, at 39; see also
Kelsen, supra note 28, at 257. Most anti‐positivists would agree that this is the proper way to formulate the
question of validity: it captures the inherent connection to normative force.
82
And where the issue of obedience actually takes centre stage, like in the case of civil disobedience, it is hardly
decoupled from the issue of validity. Where the law is not thought to be worthy of obedience, there is a
challenge against its claim to validity (or the scope of its validity).
83
It is easy to find scholars (even positivists) who also worry about this. “[T]hese analytical discussions tend to be
flat and repetitive in consequence, revolving in smaller and smaller circles among a diminishing circle of acolytes.”

We must see the core claims from another perspective. We know that legal positivists are
capable of making informative and challenging claims about law. This is what happened
when some of them claimed that all laws are ultimately the commands of a sovereign, or
that the validity of law must be traced back to a hypothetical basic norm. Such claims gave
massive stimulus to the legal theoretical discourse. They were worthy of serious debate
and extensive analysis. But the core claims are very different to them. They occupy a
broad “grey zone” between inconsequential truths and bold (and potentially revealing)
conceptual claims. In that grey zone, we find claims that are true in some sense because
84
they can be associated with uncontroversial truths about law. But they are too general,
too weak and too modest to be able to elucidate sufficiently any conceptual feature of law.
When critical challenges put the positivists under pressure, they tend to react by pushing
those core claims deeper into the grey zone, making them even more elusive.

This is what actually happened with the separability thesis in the past few decades. Under
critical pressure, it drifted away from stronger substantive claims, and became ever more
modest and subtle. But it was actually a drift towards the uninteresting commonplace that
law is not the same as morals because it is institutional in a way that morality is not. The
separability thesis was rescued by formulating the issue of legal validity as a more or less
technical problem about an institutional practice: identifying membership in a set of
85
institutionalised rules. And it became unclear who can make any use of such a diluted
concept of legal validity.

III. Discursive Dysfunctions

Many legal theorists know and readily admit that the debates revolving around the core
conceptual claims have a tremendous capacity to make legal theory boring. But is it not an
exaggeration to talk about serious discursive dysfunctions here? How can the positivist
strategy that exploits the slipperiness of the core claims be detrimental to a discourse?

Well, we may begin to see the problems if we realise that deriving the theoretical identity
of legal positivism from a commitment to core conceptual claims alters the position of the
anti‐positivist critic. When legal positivism is seen as representing bold substantive claims
Jeremy Waldron, Legal and Political Philosophy, in THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW

381, 381 (Jules Coleman & Scott Shapiro eds., 2002).
84
For me the best example for an elusive but somehow true claim is that “we have to distinguish what the law is
from what it ought to be.”
85
No doubt, validity has this technical aspect. And the criteria of validity are always developed in a practice that
has a particular sociological profile. There should be ways of making sense of issues of validity within the context
of a sociological understanding of the practice. But the fruitfulness of such an understanding in throwing light on
normativity is a more dubious matter.
2011]
(like all laws are products of state legislation) the anti‐positivist can stand for a negation of
them. That is nowhere near the case with the core conceptual claims. The anti‐positivist is
likely to be the one who accepts that they are true to a certain extent, but objects to the
way they are formulated, and denies that they should have a foundational role in
conceptual legal theory. 86 That generates a different, more slippery kind of debate
between positivists and their critics—a debate that easily loses its focus, and generates
mutual misunderstanding.

And it is not simply the discursive relations between positivists and anti‐positivists that can
become confused. Similar problems are likely to beset legal positivism itself. The core
claims confer on legal positivism a strange kind of identity. It renders positivism organized
around claims that many great positivists never made explicitly. The elusiveness of the
core conceptual claims inevitably generates a kind of uncertainty around the implications
of being committed to legal positivism. This uncertainty blurs the boundaries of legal
positivism, making self‐identification the sole most important factor in determining who is
positivist. And it ignites all sorts of debates on how different theoretical initiatives fit
together to create an intellectual tradition, and who is capable of capturing the identity of
legal positivism. For positivists, the price of their alleged success in defending their core
commitments lies in generating endless theoretical bickering around petty issues of
87
demarcation. It looks like a massive discursive dysfunction to me.

If discursive dysfunctions plague legal positivism, it is bound to have a pervasive impact on
conceptual legal theory more generally. It is guaranteed by the central role legal positivism
plays in what I like to call the mainstream conceptual discourse on law. Conceptual legal
theory revolves around clarifying basic juridical concepts (like “legal rule,” “legal right,”
“obligation,” “legal validity,” and “legal system”) and some related concepts (“normativity”
and “authority”) that are helpful for the analysis of juridical concepts. But conceptual legal
theory is not simply a set of abstract problems: we can make sense of its agenda only
within the boundaries of an actual discourse. Just like in any other theoretical field, issues
of conceptual legal theory are impossible to handle without contextualising them: without
ordering them by the rules and agenda of a discourse—bringing individual theoretical
88
efforts to a common “hermeneutical horizon.”

86
For an example of trying to accommodate and re‐contextualise core positivist claims, see John Finnis, The Truth
in Legal Positivism, in THE AUTONOMY OF LAW: ESSAYS ON LEGAL POSITIVISM 195 (Robert P. George ed., 1996).
87
Even some positivists worry that legal positivism is losing its ability of turning its attention to issues that matter.
See Frederick Schauer, Legal Positivism and the Contingent Autonomy of Law, in JUDICIAL POWER, DEMOCRACY AND
LEGAL POSITIVISM 216 (Tom Campbell & Jeffrey Goldsworthy eds., 2000); Brian Bix, Patrolling the Boundaries:
Inclusive Positivism and the Nature of Jurisprudential Debate, 12 CANADIAN J. OF L. & JURISPRUDENCE 17, 25 (1999).
88
Cf. HANS‐GEORG GADAMER, TRUTH AND METHOD 198–304 (2004).
Conceptual legal theory has a long history, and it went through a series of discursive
transformations. As it happens, in its present construction, at least in the Anglo‐Saxon
world, the mainstream conceptual discourse was organized around Herbert Hart’s agenda.
It took shape after Herbert Hart’s reconsideration of the fundamental problems of
conceptual legal theory in the 1950s. 89 That put the “problematics” of methodological
positivism at the centre of the discourse—leaving to others (including normative
positivists) the position of “challengers” (or even “insurgents”) who make attempts to
modify the agenda.

I do not want to conceal the fact that my own theoretical efforts are very much driven by
the conviction that, without a reformulation of its agenda, the dysfunctions generated by
the recent developments of methodological positivism has the potential of slowly
disintegrating the mainstream conceptual discourse on law. The question for anti‐
positivism is not how we can refute familiar legal positivist claims but how we can find a
way to reshape the conceptual discourse. In my eyes, Simmonds theory must be
measured, first and foremost, by its ability to offer some kind of guidance towards a
reconstruction of the legal theoretical discourse.

IV. Fragmentation of Inquiry

Naturally, it would be foolish to pin our hopes on Simmonds if we could not be sure that he
is aware of the discursive dysfunctions we were talking about. Promisingly, he seems to be
very much aware of them. In fact, one of the best available diagnoses of these problems
can be found in his works. As I have indicated above, he criticises the positivists for their
90
“fragmentation of inquiry.” “Many theorists have sought to detach inquiry into the
nature of law from the broader domain of political philosophy, where we might reflect
upon justifications for the use of force.” 91 He also notes the positivist strategy of
“disentangling” the criteria of the general characteristics of legal systems, the specific
criteria of validity for individual legal systems and the moral considerations concerning the
justifiability of enforcing the law. 92 “According to the positivists, the nature of law and
legality cannot be understood so long as we search for a single set of criteria that both
confers upon a rule the quality of being law and gives that rule its justificatory force.” 93
89
The symbolic discursive event that indicated the emergence of the new discourse is the Hart‐Fuller debate in
1957–1958. See Hart, Positivism and the Separation, supra note 75; Lon Fuller, Positivism and the Fidelity to Law,
71 HARV. L. REV. 630, 630–72 (1958). The other decisive discursive event is the publication of Hart’s The Concept
of Law in 1961.
90
See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 21–25.
91
Id. at 124.
92
See id. at 170–71.
93
Id. at 170.
2011]

I do not claim that Simmonds’s diagnosis is the same as mine but I do believe that they are
compatible. And Simmonds’s diagnosis has the potential to point out how pervasive the
discursive dysfunction is in a discourse driven by the Hartian theoretical initiative. It is
nowhere near limited to generating a problematic contrast between validity and “ought to
be obeyed.” 94 Separating out certain conceptual issues and transferring some of them to
different disciplines is a direct consequence of the methodological positivist vision for
conceptual legal theory. It is manifested in many ways. Simmonds glimpses it in the
contrast that is drawn between the external viewpoint of the observer and the internal
point of view of the participants, 95 between questions about the nature of law and its
content, and between legal doctrinal questions and jurisprudential questions. 96

Simmonds also realises the positivists’ undeniable success in fending off critical challenges,
and the futility of embarking on a direct denial of core positivist claims. 97 Anti‐positivists
have fallen too often for the invitation to step up to the level of meta‐theoretical debates
about the character of legal theory. 98 That meta‐theoretical debate was shaped by the
positivist picture of legal theory, and works systematically in the positivists’ favour by
sidelining substantive issues about the content of law and about the way legal mechanisms
actually work. The “fragmentation of inquiry” serves to secure the dominance of meta‐
theoretical considerations over substantive inquiries in the conceptual discourse.

This leaves us with the real question that my analysis must address: Does his diagnosis of
the deeper problems with legal positivism enable Simmonds to go beyond the positivist
agenda for legal theory? Does he provide an attractive vision for revising the agenda for
the conceptual discourse? This is the issue that I will turn to in the following section.

D. Interpretivism

Simmonds’s solution to all the problems we have identified in the previous sections is a
revised philosophical vision for legal theory. As we have seen in section B, he is looking for
a philosophical vision that does not lose contact with ordinary legal experiences.
Simmonds is well aware that that cannot be achieved if we allow for a rapid collapse of
inquiries concerning the nature of law into meta‐theoretical debates about the nature of
94
Id. at 126.
95
See id. at 22.
96
97
98
See id. at 43.
legal theory. 99 That would lead to a drift towards methodological and meta‐theoretical
debates that work against anti‐positivism (and that is a characteristic feature of the
contemporary conceptual discourse).

This may not sound very promising. How can we reach the level of genuine philosophical
reflection if we are committed to remaining tied to the ground provided by legal
experiences? As we have seen in section B, Simmonds’s answer to this worry is his
insistence on the reflexivity of law. Even in the context of our ordinary experiences, the
concept of law has an intriguing reflexive depth. “Once we have seen this point we start to
appreciate the extent to which the concept of ‘law’ is not one that simply describes a
familiar social phenomenon: the concept also plays an indispensable role within that
100
phenomenon.” The reflexivity of law invites philosophical reflection because, as a
consequence of it, we simply cannot account for our ordinary legal experiences without
getting to grips with an abstract idea (with complex moral implications). “That is to say,
legal thought is always guided and informed by reflection upon the idea of law, and it is the
task of jurisprudence to investigate that idea.” 101

I. Hart and the Denial of the Reflexivity of Law

This link between the claim about the reflexivity of law and a philosophical vision for legal
theory is supposed to provide an opportunity to recontextualise the critical attack on legal
positivism. We may glimpse that the real reason behind all the problems with
methodological positivism may be a deep misunderstanding concerning the philosophical
issues that legal theory is supposed to address. Hartian legal positivism failed to provide a
proper philosophical vision for legal theory.

As we have already seen in section C, Hart was committed to an account of law that is
restricted to elucidating the common structural features of legal systems and their
conceptual implications. He thought of this as a major breakthrough in clarifying the
methodological conditions for a plausible conceptual theory of law. And this led him to an
102
attempt to turn all the problems of conceptual legal theory into descriptive issues. This
required him to transfer the issues that did not fit into his descriptivist framework beyond
the boundaries of conceptual legal theory. 103 This is the very origin of the ‘fragmentation
99
See id.
100
See Simmonds, The Nature of Law, supra note 2, at 613–14; cf. SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at
10.
101
102
See HART, supra note 6, at 240.
103
For an example of the Hartian rejection of the relevance of certain issues for conceptual legal theory, see id. at
107–08.
2011]
of inquiry’ that we talked about in the previous section. Other methodological positivists
were happy to follow Hart down that path.

But there is a problem with this strategy. It is not obvious how this vision of a descriptive
account is to be accepted as a philosophy of law at all. It does not seem to identify a
genuine philosophical problem at all. “Describing how one thing resembles but also differs
from two other things is not, without more, a philosophical enterprise.” 104 Hart’s account
turns out to be surprisingly flat as a philosophical inquiry, and Simmonds offers an
explanation of why this is so. The methodological legal positivist initiative is not viable as a
theoretical project if we acknowledge the reflexivity of law. If reflexivity is one of the
conceptual features of law, Hart’s descriptive account can never clarify it for us. His
methodology forced Hart into denying the reflexivity of law, thereby “obscuring, and
leading people to forget” the problem that should be at the heart of the philosophy of
105
law.

What can we make of this strong claim? Simmonds himself insists that it means that Hart
failed to identify any real philosophical issue for jurisprudence. 106 But that is an
exaggeration. The Hartian agenda is continuous with an important philosophical tradition:
analytical philosophy. It seems more reasonable to claim that the Hartian agenda (that is
still the very background to the contemporary conceptual discourse) is based on a
methodological vision that is not adequate for the philosophical analysis of a normative
social practice. I am ready to subscribe to Simmonds’s claim under this interpretation.

II. Simmonds’s Interpretivism

What is the alternative philosophical agenda that Simmonds offers? Simmonds’s own
philosophical vision for legal theory is based on Collingwood’s views on philosophical
reflection. It seeks to bring reflection into “conformity with an idea of what it ought to
107
be.” The reference to Collingwood may give the impression that Simmonds seeks to
establish here a connection with a philosophical tradition that has not been particularly
influential in legal theory. That is not far from the truth but, in fact, the key idea is not as
unfamiliar in legal theory as it might seem. What Simmonds offers can be seen as a
version of what I call jurisprudential interpretivism. If we are to clarify whether Simmonds
can provide a viable alternative to legal positivism, we have to assess the prospects of his
distinctive interpretivism in legal theory.

104
See Simmonds, The Nature of Law, supra note 2, at 602–3.
105
106
See Simmonds, The Nature of Law, supra note 2, at, 602.
107
See id. at 614.
It is useful to start that assessment with a clarification of what interpretivism is. 108 I prefer

to associate the fundamental idea with a methodological challenge facing any legal theory.
Each legal theory needs to rely on a “background epistemology” that concerns the ways in
which the theorist can reveal theoretically relevant facts and ideas that are to be organized
into a viable theoretical account. The background epistemology provides the
methodological basis for the truth‐claims of the theory. This is what makes the theory
intelligible for others—enabling them to figure out what counts as justification of the
theory or as a challenge to it.

Interpretivism is one possible background epistemology for legal theory. Interpretive
theories of law gather the “data” that underlie their conceptual claims from participant
communication about the legal practices they want to understand. Such theories construe
law as a social practice in which participant communication gives rise to juridical concepts.
The theorist seeks to clarify the conceptual features of law by interpreting those concepts
and their characteristic uses in the relevant social practices.

Interestingly, it was Herbert Hart who made interpretivism the characteristic background
109
epistemology for legal theory. It was a pivotal aspect of his methodological vision that
the analysis of legal concepts must be interpretative (“hermeneutic”). 110 For him, the task
of conceptual legal theory was to elucidate the conceptual implications of our familiar
ideas concerning law and our ways of communicating legal claims.

Naturally, Hartian interpretivism was developed as an integral part of a methodological
positivist enterprise, and most contemporary methodological positivists embrace it today.
But interpretivism became attractive far beyond the boundaries of legal positivism. Over
the past few decades, various versions of interpretivism emerged from the legal
theoretical debates, some of which are designed for the purposes of anti‐positivist
111
theories. Dworkin’s interpretivist methodology is probably the best known among
108
For my account of interpretivism, see Bódig, Interpretivism and Conventionalism, supra note 24, at 155–58; see
also BÓDIG, supra note 16, at 446–55.
109
Before Hart, other types of background epistemology were prevalent, even among positivists. Kelsen, for
example, relied on a version of Kantian transcendentalism that seeks to elucidate concepts by reducing them to
their transcendental presupposition. See, for example, the way he characterises “ought” as a transcendental
category in KELSEN, supra note 3, at 23–25. But empirical reductionism, that seeks to reduce legal phenomena to
their empirical grounds, is a more familiar alternative to interpretivism. Scandinavian realists kept it influential
th
even in the first half of the 20 century. For an example, see KARL OLIVECRONA, LAW AS FACT (1st ed. 1939).
110
See H.L.A. HART, ESSAYS IN JURISPRUDENCE AND PHILOSOPHY 13 (1983).
111
Even John Finnis accepts the interpretivist starting point for legal theory. See FINNIS, NATURAL LAW, supra note
54, at 3.
2011]
them. 112 Although there is a committed anti‐interpretivist strand in contemporary legal
theory, 113 I do believe that the spread of interpretivist methodology is a positive
development. 114 This is actually the good thing that came out of the Hartian initiative.
Anti‐positivism must find ways to attack methodological positivism without undermining
interpretivism.

The interpretivist element in Simmonds’s account is pretty conspicuous. He seeks to
provide a theoretical clarification of our “ordinary understandings” of legal practices, 115
and his philosophical vision aspires to “assemble these understandings into a coherent
picture of law’s nature.” 116 Actually, Simmonds is one of those who think that
interpretivism is not simply one of the methodological options for legal theory: the law has
certain features that call for an interpretive approach.

An account of the nature of law must be judged by its
ability to yield insight and make coherent sense of our
more settled understandings and beliefs about law. For
law is not a phenomenon with a nature wholly
independent of our beliefs about it, but one that is
117
constituted by our understandings and expectations.

The interpretivist element determines the way Simmonds unpacks the implications of the
methodological ideal he borrows from Collingwood. The call for bringing reflection into
“conformity with an idea of what it ought to be” could mean a lot of things. But Simmonds
makes it clear that what he has in mind is not a turn towards metaphysics but “a
reconstruction of the ideas and conceptions structuring the form of association that we
118
think of as a legal order.”

112
See DWORKIN, supra note 53, at 12–14. Some argue that interpretivism itself is to be identified with Dworkinian
jurisprudential methodology. See Nicos Stavropoulos, Interpretivist Theories of Law, STANFORD ENCYCLOPAEDIA OF
PHILOSOPHY, available at http://plato.stanford.edu/entries/law‐interpretivist. I think such claims are misleading
and simplify the character and development of interpretivism in legal theory.
113
See Michael S. Moore, The Interpretive Turn in Modern Theory: A Turn for the Worse?, 41 STAN. L. REV. 871,
871–957 (1989); Gerald J. Postema, Jurisprudence as Practical Philosophy, 4 LEGAL THEORY 329, 329–30 (1998).
114
For my commitment to interpretivism, see Bódig, Interpretivism and Conventionalism, supra note 24, at 156–
57.
115
116
See id. at 614; see also SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 23.
117
118
See Simmonds, The Nature of Law, supra note 2, at 614; cf. SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 58.
III. Linking up Legal and Moral Inquiries

What makes Simmonds’s interpretivism distinctively anti‐positivist? Usually the anti‐
positivist element in interpretivist theories is the claim that the interpretation of the
conceptual implications of participant communication necessarily involves handling
justificatory issues or embracing value assumptions. This is what we see in the most
familiar version of anti‐positivist interpretivism, Dworkin’s legal theory. He argues that law
is an interpretive practice, and one can only develop an interpretive theory of practice by
119
attributing a moral point to it. It might seem that Simmonds offers us something very
similar. He says that our social practices must be seen as expressing values, and we can
only understand them by their orientation towards those values. 120

But we should not be misled by this apparent similarity. In fact, Simmonds’s philosophical
vision is quite different from Dworkin’s. 121 An important clue that can help figure out the
character of Simmonds’s philosophical programme is that he is remarkably unsympathetic
to the analytical philosophical tradition. It indicates a deeper layer in his objections to
methodological positivism. Simmonds believes that the Hartian methodological vision
could become dominant partly because it “chimed with the spirit of an age” 122 in which
analytical philosophy set the tone for philosophical reflection.

Simmonds criticises analytical philosophy because he finds it incapable of providing an
adequate account of how values figure in our practices (and generate moral experiences).
In practical philosophy, analytical philosophy was the culmination of a “modern” approach
that seeks to understand our intuitions about values in isolation from the practices that
shaped them. “[W]e are encouraged to believe that our particular moral perceptions must
be interrogated in the light of articulable principles before they can be given any weight in
123
our deliberations.”

I do believe that Simmonds’s analysis of the development of this modern approach 124 and
its impact on the methodological conundrum that we are struggling with is one of the
119
See, e.g., DWORKIN, supra note 53, at 50–52.
120
121
Simmonds puts real emphasis on making that difference clear. See id. at 33–34. In fact, Simmonds implicates
Dworkin in his criticism of “fragmentation of inquiry.” “Dworkin’s view seems to be that there is no general
philosophical question to be asked about the nature of law as a distinct type of social structure.” Id. at 27. This is
one of the reasons behind his insistence that his anti‐positivism is different to Dworkin’s. See Simmonds, The
Nature of Law, supra note 2, at 607–8, 622.
122
See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 173; cf. id. at 151.
123
Id. at 150.
124
See id. at 150–55.
2011]
great strengths of his theory. He is right that the “modern” approach to reflecting upon
moral experience is favourable to the positivist position. If the issue of the conceptual
relationship between law and morality is raised as a matter of abstract moral principles, it
is difficult to resist the positivist conclusions. If we try to measure up the moral
implications of law by resorting to abstract moral values, the relation of law to morality will
always look contingent at best. It is really ironic that modern naturalists like Grotius and
Pufendorf played a pivotal role in laying the foundations for the “modern” approach. 125

Anti‐positivism has a much better chance of regaining the theoretical initiative if it
challenges the idea that philosophy must capture the elements of morality in terms of
abstract principles. The morality that is relevant to law must be seen as immanent in our
practices, “somehow grounded in what we do.” 126 Once again, Simmonds does not
perceive it as a matter of choice: for him, this is the proper way to make philosophical
sense of our own understanding of pivotal moral concepts.

When detached from our ordinary understandings the
notions of justice and morality can quickly come to
seem no more than empty illusions. . . . If our moral
understanding is to be seen as real, it must be bound
up with our grasp of the commonplace world, forming
part of the realm of socially sustained knowledge that
127
renders us mutually intelligible.

It is in his context (in light of this construal of morality) that conceptual legal theory should
ask whether our experiences of law have genuine moral significance: whether we can
learn something about morality by our engagement in legal (“legalistic”) practices.

This is, obviously, interpretivism about morals, not simply about law. And although it still
128
may not sound all that different to Dworkin, it is at this point that the two accounts
become markedly different. The analysis of the relevant philosophical developments leads
Simmonds to relying on a distinctive philosophical tradition that has not really made an
impact on Anglo‐Saxon legal theory, and that is pretty alien to Dworkin. Simmonds
125
See id. at 152–53. The impact of the “modern” approach may be even reflected in the “fragmentation of
inquiry”—“in the attempt to divide jurisprudence into a neutral and a normative inquiry.” Id. at 147.
126
Id. at 7.
127
Id. at 141.
128
Dworkin also takes an interpretive attitude to values in general. See Ronald Dworkin, Objectivity and Truth:
You’d Better Believe It, 25 PHILOSOPHY AND PUBLIC AFFAIRS 87, 87–139 (1996); Ronald Dworkin, Hart’s Postscript and
the Character of Political Philosophy, 24 OXFORD J. OF LEGAL STUD. 1, 5–9 (2004). For Simmonds, however, Dworkin
is not really detached from the “modern” approach. I accept this even though I can imagine how one can
challenge this reading in light of what Dworkin says on “integrated values.” See id. at 14–16.
sympathises with a kind of Aristotelian practical philosophical perspective that revolves
around the practical wisdom inherent in our practices and morally significant
experiences. 129

Simmonds’s philosophical orientation has a far‐reaching implication that is important for
us now. On this account, the connection between interpretivism about law and
interpretivism about morality becomes particularly intimate. It is not that jurisprudential
reflection runs parallel to moral reflection—jurisprudential reflection is actually
incorporated into moral reflection. Interpretivism about law becomes an aspect of our
efforts to come to terms with the values that our lives and practices come to be
associated. “Jurisprudence is, at its best, an integral part of this process of deepening by
130
reflection our grasp of the values implicit in our forms of association.” Without the
effort of coming to terms with values in light of our practices the theorist cannot find the
right perspective to engage in jurisprudential reflection. 131

It may look like an unduly long overview of Simmonds’s distinctive kind of interpretivism.
But I needed to go this far to make an important point. Although what Simmonds provides
is undoubtedly a neo‐Fullerian account, it fundamentally changes the philosophical
character of the Fullerian initiative. It can be seen pretty clearly in the way Simmonds
integrates the analysis of the Fullerian desiderata of the rule of law into his account.

For Simmonds, the principles of the rule of law provide a paradigm for the way principles
can be developed from our experiences of social practices. They provide a starting point
132
that is wholly uncontentious —not even the positivists are likely to challenge it. The
principles are firmly grounded in our experiences and intuitions about what it takes to
make a legal system work. What makes them morally relevant is that their further analysis
leads us to an understanding of how the law needs to have “a certain fit with existing
forms of life, habits of conduct, settled expectations and shared moral sentiments.” 133
That makes us realise how the organizing ideal behind the idea of law must be a moral
one. 134 But we also realize that the attempt to come to terms with the conceptual
features of law is also about (or really about) learning something about moral values.
129
For Simmonds, Gadamer is the guide to the way the Aristotelian perspective can be used in a contemporary
philosophical context. See SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 146–48.
130
See id. at 8.
131
“Jurisprudence can only properly be understood in the light of the possibility that moral insight might be
derived from historically informed reflection upon our practices and institutions . . . .” Id. at 147.
132
133
134
See id. at 100–01, 142, 163.
2011]
When we try to handle the possible conflicts between the principles of the rule of law, we
realize that “an abstract knowledge of principles will not help us.” 135 In the course of
reflecting upon law, we gravitate towards gaining a deeper understanding of the moral
parameters of ways of life in actual, historically constituted communities. And we come to
realize that the experiences and intuitions that ground the seemingly technical principles
of the rule or law are “inchoate moral insights rather than (as they might at first seem)
semantic intuitions.” 136

E. A Few Critical Remarks

This may be enough for me to be able to provide the foundations for an assessment of the
plausibility and the prospects of Simmonds’s theoretical initiative. I have already indicated
that I am sympathetic to his project. I do believe that Simmonds helps us understand at
least some of the discursive dysfunctions that plague the contemporary discourse on
conceptual legal theory. His theoretical initiative offers one possible way to reconstruct
137
the legal theoretical discourse. His diagnosis of the connection between the legal
positivist “fragmentation of inquiry” and the “modern” approach to making philosophical
sense of our moral experiences is a great strength of his account. Simmonds is right that
we have very little chance of moving forward in conceptual legal theory without deepening
the reflection on morality and moral inquiry. In that respect, recent legal theoretical
debates have often proved frustratingly impoverished. They are too impoverished to be
able to clarify the conceptual connection between law and morality.

But I also have some doubts about the direction Simmonds’s account takes. It tends
towards subordinating the agenda of conceptual legal theory to the needs of gaining a
deeper understanding of the values that our lives and experiences are associated. It seems
to turn conceptual legal theory into an exercise in moral inquiry. That may prove to be too
narrow as an account of the concerns of conceptual legal theory. It may underestimate
the complexity of the ways in which conceptual legal theory can and must respond to
138
genuine epistemic concerns. This is an important point but I will not pursue it here. It
would be a mistake to jump to a hasty judgment on this issue. Simmonds’s account is
sufficiently complex to allow for several readings. What he argues for can boil down to a
135
See id. at 145.
136
See id. at 68.
137
There are other attempts to reconstruct the discourse. Waldron’s can be very important as it comes from
within legal positivism. It is based on reconsidering what we should mean by “general jurisprudence.” See
Jeremy Waldron, Can There Be a Democratic Jurisprudence?, 58 EMORY L.J. 675, 675–712 (2009). Twining also has
a proposal that would be based on assessing how globalisation changes the agenda. WILLIAM TWINING, GENERAL
JURISPRUDENCE: UNDERSTANDING LAW FROM A GLOBAL PERSPECTIVE (2009). I find Twining’s initiative somewhat less
promising than Waldron’s.
138
For my account of the character and functions of conceptual legal theory, see BÓDIG, supra note 16, at 435–42.
version of the claim that conceptual legal theory must be taken as an integral part of
practical philosophy, and I am one of the advocates of that claim. 139 For such reasons, I
cannot substantiate my doubts without an extensive analysis, and I cannot provide that
analysis here.

I. The Limits of Interpretivist Theorising

Instead, I would like to say more about a smaller point here. It concerns an aspect of
Simmonds’s interpretivism. As we have seen, he assigns to legal theory the task of
integrating our understanding of law into a coherent whole. It is to provide a “single
140
intelligible picture of law’s nature.” I am worried that this task may be too demanding
for an interpretivist account of law. I tend to believe that it sits very uneasily with the
methodological characteristics of interpretivism. We are not supposed to think that there
is some methodological guarantee that the elements of our ordinary understanding of law
can fit together into a coherent whole.

Even if we admit that “[l]aw exists only in so far as a great many people share certain
141
understandings and expectations,” those common understandings are likely to be
complex and multi‐faceted. And they may not provide a consistent set of ideas that the
theorist can easily organize into a coherent account. The methodological implication of
this simple (perhaps even commonplace) observation is that interpretivism is exposed to
the contingent character of the interpretive data that participant communication provides.
It has to settle for the fact that interpretive analysis will provide, instead of one particular
account, a number of competing, not wholly consistent accounts of law that have different
advantages and disadvantages. A bit surprisingly, this does not seem to be reflected in
Simmonds’s legal theory.

There are a number of ways to unpack the implications of this methodological observation
about the contingency of interpretivist accounts of law. One could point to the fact that
different participants have different practical orientations towards legal practices, and
their communication provides different interpretive data. It may actually mean that
different (but similarly plausible) interpretive theories will be able to account for the
understanding of law by the professional lawyer and the revolutionary critic. I have
142
discussed this issue elsewhere, and I will not pursue it further here. It seems more
reasonable (and accessible) to concentrate on another aspect of the same methodological
challenge.
139
See id. at 478–506.
140
SIMMONDS, LAW AS A MORAL IDEA, supra note 1, at 23; see also id. at 36.
141
142
See BÓDIG, supra note 16, at 486–90.
2011]

No doubt, we can often rely on an implicit or explicit agreement on the set of interpretive
data that legal theory has to work with. (Mainstream legal theory tends to gather its
interpretive data from the way legal professionals communicate about law, and Simmonds
seems to be no exception.) But even then, there are various pathways open to us that lead
to different accounts with different theoretical advantages and disadvantages. The issue
of normativity that we discussed in section B is a nice case in point.

When we talk of normativity, or, more exactly, the normative impact that official decisions
or statutory provisions have upon our behaviour, our intuitions reflect two basic aspects of
normativity. The first concerns the justificatory force of legal propositions that are
communicated to us. This is the aspect of normativity that comes to the fore when we ask
questions about the way rules are capable of guiding our behaviour and shaping the
conditions under which we can justify what we do. The other aspect of normativity
concerns the status of the officials who have the normative competence (authority) to turn
simple normative claims into legal pronouncements—making them elements of an
institutional practice. This is the aspect that we concentrate on when we ask questions
about the source of the distinctively legal authority of judges or legislators.

The two aspects seem to fit together nicely: without institutionalised legal
pronouncements, there would be no normative guidance, and, without officials who have
authoritative competence, we would have no legal pronouncements. But it is not that
simple. Depending on which aspect of normativity is at the forefront of our analysis, we
are likely to go down markedly different pathways towards an account of law’s
normativity. If we start from asking questions about the normative capacity of
pronouncements to guide action, we are very likely to be guided by one of our
fundamental intuitions about law: law is a matter of guidance by rules, and law is to be
taken as a system of rules or norms. We are likely to treat as fundamental a particular
interpretive data: a connection between the concept of rule and the concept of law in the
participants’ understanding. That leads us to an account of law via a clarification of rules
and rule‐following, the preconditions of a working system of rules, and the type of social
control that revolves around formally enacted, institutionally guaranteed rules.

On the other hand, if we start from the issue of the normative competence of officials, our
path is likely to lead us towards noticing that rules may not be all that central to the
conceptual character of law. The normative competence of the institutions (whose
143
practice underlies legal pronouncements) is more fundamental. Substantive rules may
not even be necessary conceptual features of law. The necessary element is the
institutional framework within which someone gains authority over others and makes
143
For an account like this, see Mátyás Bódig, A normativitás és a jog sajátos normativitása, in ÉRTÉKEK ÉS NORMÁK
INTERDISZCIPLINÁRIS MEGKÖZELÍTÉSBEN (Katalin S. Nagy & Annamária Orbán eds., 2008).
decisions. It is a matter of contingent institutional arrangement whether those decisions
are based on rules (or constitute rules). The conceptually necessary elements are what I
would call “political office” and “authoritative decision.” Rule‐based decision‐making
becomes a feature of law in the process of rationalizing the operations of political offices.
This account would put heavier emphasis on another set of interpretive data: participants
see law as a matter of human decisions, manifestations a distinctive form of social power
that is stabilised by institutions. Law is not so much a system of rules but a social
practice—a system of institutions, a web of interrelated political offices.

The two accounts are markedly different. The first puts the concept of “rule” at the very
heart of the concept of law. For the other, the concept or “rule” gains much less
conceptual significance, and figures in at a much less fundamental level. But both accounts
are firmly grounded in established interpretive data. When we talk of the legal system of a
particular community, we sometimes refer to the rules and doctrines that constitute the
content of law there. But sometimes we mean the system of institutions: courts,
legislative bodies, regulative agencies, etc. I do not see a way in which one could decide
which account of law is “more true” than the other.

The choice that one makes at this point and its implications are reflected in Simmonds’s
144
analysis. When he outlines the issue of normativity, both aspects of it are highlighted.
But as he develops his arguments, the issue of the justificatory force of laws comes to the
fore, and we lose sight of the other aspect of the problem. When, at a later stage, he
arrives at formulating the issue of normativity in the context of his attack on positivism,
only the justificatory aspect of laws remains in focus. 145 The question he raises is why
judges think invoking the law justifies their decisions, and not why judges think they are
the ones to decide.

II. Embracing Interpretivism

I do not hold it against Simmonds. All this follows from his emphasis on accounting for law
as a mechanism that revolves around guiding human behaviour by rules. It is perfectly
justified by the interpretive data he relies on. And it provides the proper background to his
reliance on the theoretical implications of the eight principles of the rule of law. But we
have to realise that it makes his account limited as a conceptual inquiry. It leaves open the
possibility of an alternative approach that could yield no less plausible results. In that
alternative approach (that sees law as revolving around the actions of officials who claim
144
Simmonds raises the issue of grounding the authority of decision‐makers (i.e. legislators). See Simmonds, The
Nature of Law, supra note 2, at 604. For highlighting the justificatory aspect of laws, see id. at 605.
145
“Suppose that a judge sentences me to prison, citing a certain rule as the justification for my punishment. I
protest and demand to know how the existence of the rule serves to justify sending me to prison.” Id. at 610
(emphasis added).
2011]
specific normative competence to make decisions), Simmonds’s arguments concerning the
eight principles of the rule of law would not work in the same way, and would look far less
fundamental. They would work only in light of the conceptual relevance of a legitimacy
issue—after establishing a link between the decision‐makers’ needs for legitimacy and the
potential of the eight principles to make the practice better able to provide normative
guidance.

It does not call into question the Simmonds’s truth claims but it definitely narrows down
the scope of his account. To be more exact, his remarks on the problems with the
contemporary discourse (and the “fragmentation of inquiry” in particular) have a more
general appeal and significance, but those constitute only a series of negative, critical
claims. As Simmonds develops his positive arguments, and brings together the elements of
his own account of law, they become answers to very specific epistemic concerns about
law. Simmonds’s account becomes a theoretical reflection of a particular dimension of our
experiences and intuitions about law. It becomes limited in its ability to address other
epistemic concerns. We have seen the signs of this in our analysis. Its reliance on the
problem of normativity turns Simmonds’s account directly against methodological
positivism, but leaves the issues of the viability of a normative positivism out of focus. It
shares the perspective of interpretive legal theories, limiting its ability to address the
issues involved in the agenda of other positivists, like Kelsen. And his account is an
inadequate response to the epistemic concerns of a positivist, like Joseph Raz, who put
authority relations at the centre of his account of law. Of course, that does not leave him
146
without an abundance of legitimate “targets.” But it leaves him with the challenge of
explaining how this theory is capable of capturing the ambitions of conceptual legal theory
on a more general level, and setting the agenda for conceptual inquiries.

I do believe that this not at all a minor problem for Simmonds. In many respects, the
attractiveness of his philosophical vision depends on his ability to handle it. Without a
proper understanding of the methodological background issues, it may seem disappointing
for a theorist facing Simmonds’s arguments that his own account remains narrow in its
epistemic concerns. If we accept Simmonds’s suggestion that we should resist the
fragmentation of inquiry characteristic of contemporary methodological positivism, we
rightly think that the ultimate justification of his initiative lies in offering a viable
alternative to positivist accounts of law. We wonder how a theory that does not
“fragment” the inquiry would look like. Then it may be confusing to get an account that is
146
I do not simply mean Herbert Hart here (who is singled out in the paper I am commenting upon). I do believe
that Simmonds’s arguments work very well against Jules Coleman, Matthew Kramer, Wilfrid Waluchow and
Kenneth Einar Himma as well. These are mostly the prominent figures of what we tend to call “inclusive legal
positivism.” See COLEMAN, supra note 11; KRAMER, IN DEFENSE OF LEGAL POSITIVISM, supra note 4; MATTHEW KRAMER,
WHERE LAW AND MORALITY MEET (2008); WILFRID J. WALUCHOW, INCLUSIVE LEGAL POSITIVISM (1994); Kenneth E. Himma,
Law’s Claim of Legitimate Authority, in HART’S POSTSCRIPT: ESSAYS ON THE POSTSCRIPT TO THE CONCEPT OF LAW 125 (Jules
Coleman ed., 2001).
a reflection of only one dimension of our legal experiences—with limited validity claims
and explanatory power.

I mean it is confusing until we realise that this is the inevitable consequence of a
commitment to interpretivism. All interpretivist theories are reflections of a particular
dimension of our experiences and intuitions about law. And, if interpretivism happens to
fit the epistemic character of law, the limits of theoretical outlooks are the inevitable
consequence of the fact that our theories deal with law. If we clarify this methodological
situation, we are better positioned to understand the significance of Simmonds’s efforts.
We will know that no one is likely to be able to come up with accounts of law that have a
broader scope. What we are involved in is not a competition to provide one legal theory
that clarifies the conceptual character of law once and for all, but an effort to create a legal
theoretical discourse where we accept the need to develop a multiplicity of interpretive
perspectives to be able to account for the different dimensions of our legal experiences.
Simmonds’s legal theory would look even stronger if it was bolstered by further efforts to
clarify the methodological conditions of interpretive theories of law, and their impact on
the agenda of conceptual legal theory.
Special Issue
Is Post‐Positivism Possible?

By Karen Petroski *

A. Introduction

In some of his last published works, Neil MacCormick began to refer to his theoretical
position as “post‐positivist.” 1 In light of the widely perceived limitations of the “positivist”
label, this self‐identification might seem prudent. 2 Was it anything more? Was
MacCormick’s position really post‐positivist? In this paper, I argue that it was not, but that
this need not be viewed as a failing of MacCormick’s work, since there is a sense in which
modern jurisprudence cannot and need not hope to become generally post‐positivist.
More specifically, given the institutional context in which legal scholarship is produced,
positivism is likely to be an inevitable (if not necessarily dominant) mode of theorizing
about law. Yet much informative work remains to be done under the positivist rubric—not
just along the lines suggested by MacCormick, but along others as well.

My ultimate point concerns the institutional possibility and desirability of a shift to a
theoretical position called “post‐positivism.” Addressing this point requires me to take a
position not only on what that phrase means, but also on what legal positivism itself is. As
a result, in this paper, I recapitulate aspects of the debates regarding the nature or identity
of legal positivism, but I also address topics beyond the traditional scope of those debates.

*
Assistant Professor, Saint Louis University School of Law. Thanks to András Jakab, Jürgen Busch, Otto
Pfersmann, and Thomas Bustamante for valuable input. E‐mail: kpetrosk@slu.edu

1
See NEIL MACCORMICK, INSTITUTIONS OF LAW: AN ESSAY IN LEGAL THEORY (2007) [hereinafter MACCORMICK, INSTITUTIONS];
NEIL MACCORMICK, H.L.A. HART (2007) [hereinafter MACCORMICK, HART].

2
See, e.g., Kent Greenawalt, Too Thin and Too Rich: Distinguishing Features of Legal Positivism, in THE AUTONOMY
OF LAW: ESSAYS ON LEGAL POSITIVISM 1, 8 (Robert P. George ed., 1996) (noting use of “positivism” label for “summary
condemnation”); Frederick Schauer, Positivism as Pariah, in THE AUTONOMY OF LAW, ESSAYS ON LEGAL POSITIVISM 31,
35; Wilfrid J. Waluchow, The Many Faces of Legal Positivism, 48 U. TORONTO L.J. 387, 390 (1998) (discussing
“meaningless[ness]” of debates within legal positivism). Much of twentieth‐century positivist legal theory has
sought to redeem the term from its use as an epithet. See, e.g., H.L.A. Hart, Positivism and the Separation of Law
and Morals, 71 HARV. L. REV. 593, 615–21 (1957) (discussing post‐World War II critique of legal positivism by
Radbruch, among others); ANTHONY SEBOK, LEGAL POSITIVISM IN AMERICAN JURISPRUDENCE 2, 23 (1998) (discussing
derogatory use of “positivist” label).

My discussion has important limitations. First, since MacCormick’s work is my point of
departure, I focus on the Anglo‐American legal positivist tradition within which he wrote. 3
The continental tradition of legal positivism associated in the twentieth century with Hans
Kelsen lies largely beyond the scope of my argument. Second, my focus is on the
institutional conditions of production of legal theories, rather than on assessment of their
content. This essay is not a contribution to positivist jurisprudence, although it arguably
builds on a tradition that is positivist in a more general sense. Finally, my thesis is the
modest one that, as long as the current institutional conditions of legal and academic
endeavors persist, some form of legal positivism is also likely to persist, making post‐
positivism difficult to attain. I do not argue that all, or even most, legal theory must be
positivist.

My discussion has three main sections. Section B seeks to clarify the nature of the legal
positivism beyond which MacCormick sought to move and the reasons for his desire to
move beyond it. I am hardly the first person to consider the question of the nature of legal
positivism; in Section B.I, I describe some previous efforts to address it, identifying the
principal areas of overlap of these efforts. (I call these prior attempts to identify the
nature of legal positivism “second‐order” accounts, because they are not accounts of the
nature of law as such, but rather theoretical accounts of other “first‐order” accounts of the
nature of law.) The second‐order accounts I consider in Section B.I suggest that applying
the “positivist” label to a theory of law has been considered appropriate when the theory
characterizes its subject matter in a particular way. I conclude Section B by examining why
the move beyond positivism might seem attractive to a legal theorist, given the context
within which such theorists produce their theories, and the related question of what it
might mean to be “post‐positivist.”

Some of the second‐order accounts I consider note that the first‐order legal theories they
discuss seem to share important features with the subject matter, law, for which the
theories seek to account. In Section C, I consider this parallel from a different perspective.
I argue that given the modern institutional setting of scholarly activity, something
resembling the positivist mode of inquiry, and debates over its propriety and details will
always be part of legal theory. Indeed, the existence of positivist legal theory appears to
be a product of social and institutional forces very similar to those engendering the
features identified by legal positivists as characteristic of legal systems. In this section, I
also briefly explore how this argument clarifies the work of major figures in the Anglo‐
American positivist tradition. I conclude Section C by considering some potential
objections to my argument; Section D addresses another potential objection—that if it is
not possible to move beyond positivism, legal theorists will run out of interesting questions
to investigate. In fact, legal positivism has important contributions to make to other areas
of legal scholarship. In Section D, I examine two areas in which a positivist‐influenced
3
See NEIL MACCORMICK & OTA WEINBERGER, AN INSTITUTIONAL THEORY OF LAW: NEW APPROACHES TO LEGAL POSITIVISM
(1986) (identifying MacCormick’s work, in contrast to Weinberger’s, as in the British tradition of legal positivism).
2011] Post‐Positivism 665
vocabulary could enrich debates proceeding, and perhaps stalling, without the benefit of
that vocabulary.

B. Defining Positivism and Post‐Positivism

Attempts to clarify the nature of legal positivism have multiplied in recent decades. In this
section I examine the general landscape of these attempts at clarification and some of
their limitations.

As suggested above, accounts of legal positivism may be considered examples of second‐
order legal theory, since they theorize about the nature of first‐order theories, those
concerned with the nature of law itself. Section B.I offers something like a third‐order
account of these second‐order theories, describing first their general characteristics and
then their main points of agreement on the nature of legal positivism. From this
agreement, I draw a set of consensus attributes of legal positivism—my own second‐order
account of legal positivism. In Section B.II, I distinguish my account from a few of the
accounts to which it may seem to bear some superficial similarities. In Section B.III, I
consider the possible reasons for seeking to avoid the “positivist” label, as MacCormick
appears to have done.

I. Existing Accounts of Legal Positivism

1. Types of Second‐Order Account

Since the early 1980s, the volume of Anglo‐American commentary on the nature of legal
4
positivism itself has increased significantly. Individual examples of this commentary have
varied aims and characteristics. Some, for example, appear as preambles to new or revised
theories of the nature of law, which need not themselves be positivist. In the second half
of the twentieth century, this type of work has been done by H.L.A. Hart, Ronald Dworkin,
Joseph Raz, and MacCormick, to name only some of the most notable figures. While these
accounts of legal positivism have a critical, normative component, others are more
descriptive, aiming mainly to clarify the landscape of theories of law or to correct
misapprehensions about such theories.
4
These decades were marked, for example, by the publication of monographs on legal positivism and many
symposia and anthologies on the topic. See, e.g., TOM CAMPBELL, THE LEGAL THEORY OF ETHICAL POSITIVISM (1996);
HART’S POSTSCRIPT: ESSAYS ON THE POSTSCRIPT TO THE CONCEPT OF LAW (Jules Coleman ed., 2001) [hereinafter HART’S
POSTSCRIPT]; RUTH GAVISON, ISSUES IN JURISPRUDENCE: THE INFLUENCE OF H.L.A. HART (1987) [hereinafter ISSUES IN
JURISPRUDENCE]; THE AUTONOMY OF LAW: ESSAYS ON LEGAL POSITIVISM (Robert P. George ed., 1996) [hereinafter THE
AUTONOMY OF LAW]; POSITIVISM TODAY (Stephen Guest ed., 1996); MATTHEW KRAMER, IN DEFENSE OF LEGAL POSITIVISM:
LAW WITHOUT TRIMMINGS (1999); GERALD J. POSTEMA, BENTHAM AND THE COMMON LAW TRADITION (1986); SEBOK, supra
note 2; WILFRID WALUCHOW, INCLUSIVE LEGAL POSITIVISM (1994); Symposium: The Hart‐Fuller Debate At Fifty, 83
N.Y.U. L. REV. 993 (2008). Many other instances of second‐order accounts of legal positivism by these authors and
others appeared as stand‐alone articles and are cited in the footnotes to this Section.


One approach common in this work uses the idioms of analytic philosophy, explaining legal
positivism as a matter of conceptual commitment to particular theses. If Hart initiated the
twentieth‐century link between analytic philosophy and legal theory, 5 Dworkin appears to
have initiated this thesis‐based approach in his early critiques of Hart. 6 The objectives of
this kind of work are consistent with the conventions of philosophical writing, but still
diverse; indeed, the heterogeneity of the debates in this area is often cited as a good
reason to try to move beyond positivism. 7 Thus, some work of this kind draws on the
theses purportedly underlying the legal positivist position in order to demonstrate their
incoherence, 8 their identity with natural law theses, 9 or, from a less critical perspective,
their logical entailments. 10 Other work stresses either the superior descriptive power of
the theses 11 or their absence of descriptive power. 12
5
See Nicola Lacey, Analytical Jurisprudence Versus Descriptive Sociology Revisited, 84 TEX. L. REV. 945, 947–50
(2006).

6
See Ronald M. Dworkin, The Model of Rules, 35 U. CHI. L. REV. 14, 17–18 (1967) describing as the “skeleton” of
legal positivism as three “central and organizing propositions.”

7
See infra notes 44–45 and accompanying text.

8
See, e.g., Jules Coleman, Incorporationism, Conventionality, and the Practical Difference Thesis, in HART’S
POSTSCRIPT, supra note 4, at 99; David Dyzenhaus, Positivism’s Stagnant Research Programme, 20 OXFORD J. LEGAL
STUD. 703 (2000); John Finnis, On the Incoherence of Legal Positivism, 75 NOTRE DAME L. REV. 1597 (2000); Klaus
Fueßer, Farewell to ‘Legal Positivism’: The Separation Thesis Unravelling, in THE AUTONOMY OF LAW, supra note 4,
at 119, 152; Gerald J. Postema, Coordination and Convention at the Foundations of Law, 11 J. LEGAL STUD. 165
(1982).

9
See, e.g., Brian Bix, On the Dividing Line Between Natural Law and Legal Positivism, 75 NOTRE DAME L. REV. 1613,
1624 (2000) (suggesting that positivists differ from one another as much as they differ from natural‐law
theorists); Timothy A.O. Endicott, Herbert Hart and the Semantic Sting, 4 LEGAL THEORY 283 (1998) (describing
overlap of assumptions between Hart and Dworkin); Philip Soper, Searching for Positivism, 94 MICH. L. REV. 1739
(1996) (review of WALUCHOW, supra note 4); Jeremy Waldron, Normative (or Ethical) Positivism, in HART’S
POSTSCRIPT, supra note 4, 410.

10
See, e.g., Matthew D. Adler, Constitutional Fidelity, the Rule of Recognition, and the Communitarian Turn in
Contemporary Positivism, 75 FORDHAM L. REV. 1671 (2006) (exploring nature of the conventions regarding a rule of
recognition that must exist in the U.S. constitutional order); Jules Coleman, Negative and Positive Positivism, 11 J.
LEGAL STUD. 139 (1982); KRAMER, supra note 4; Andrei Marmor, The Separation Thesis and the Limits of
Interpretation, 12 CAN. J.L. & JURIS. 135 (1999).

11
See, e.g., Frederick Schauer & Virginia J. Wise, Legal Positivism as Legal Information, 82 CORNELL L. REV. 1080,
1087–88, 1092 (1997); Frederick Schauer, The Limited Domain of the Law, 90 VA. L. REV. 1909, 1951–54 (2004).

12
See, e.g., David Dyzenhaus, The Demise of Legal Positivism?, 119 HARV. L. REV. F. 112 (2006); Robert P. George,
Natural Law and Positive Law, in THE AUTONOMY OF LAW, supra note 4, at 321; Jeffrey D. Goldsworthy, The Self‐
Destruction of Legal Positivism, 10 OXFORD J. LEGAL STUD. 449 (1990); Leslie Green, Positivism and Conventionalism,
12 CAN. J. L. & JURISPRUDENCE 35 (1999); Gerald J. Postema, Jurisprudence as Practical Philosophy, 4 LEGAL THEORY
329 (1998); Philip Soper, Law’s Normative Claims, in The AUTONOMY OF LAW, supra note 4, at 215.


The other set of prominent academic conventions followed by accounts of the nature of
legal positivism are those associated with the work of legal and social historians. Some
genealogical accounts of this kind describe a coherent legal positivist tradition, resting on
an essential continuity of attitude, method, vocabulary, or conceptual commitment among
those self‐identified or described as legal positivists. 13 Others point to the incoherence of
the positions espoused by the same figures—either to deny the absence of any such
tradition, a way of confirming claims about the meaninglessness of the “positivist” label, 14
or to point out particular theorists’ misinterpretations of their predecessors’ positions, a
way of redescribing the tradition. 15 While some of these accounts focus on aspects of the
institutional setting within which particular legal theorists have written, 16 none relates that
context to the debates about legal positivism itself in any sustained way.

2. Where Second‐Order Accounts Agree

As the above discussion suggests, theoretical accounts of legal positivism disagree on many
points. They do not only disagree about the academic conventions according to which
legal theory should be done, or the other disciplinary resources it should draw on. They
disagree about the nature of legal theory itself, and about whether positivism is a
13
See, e.g., Brian Bix, Patrolling the Boundaries: Inclusive Legal Positivism and the Nature of Jurisprudential
Debate, 12 CAN. J.L. & JURIS. 17 (1999) (tracing legal positivism to Hobbes); Owen M. Fiss, The Varieties of
Positivism, 90 YALE L.J. 1007 (1981); John Gardner, Legal Positivism: 5 1/2 Myths, 46 AM. J. JURIS. 199 (2001); Neil
MacCormick, A Moralistic Case for A‐Moralistic Law?, 20 VAL. U. L. REV. 1 (1985); Stephen Perry, Hart’s
Methodological Positivism, in HART’S POSTSCRIPT, supra note 4, at 311; Kevin Toh, Hart’s Expressivism and His
Benthamite Project, 11 LEGAL THEORY 75 (2005); William Twining, General and Particular Jurisprudence—Three
Chapters in a Story, in POSITIVISM TODAY, supra note 4, at 119; Jeremy Waldron, All We Like Sheep, 12 CAN. J. L. &
JURIS. 169 (1999).

14
See, e.g., Deryck Beyleveld & Roger Brownsword, Normative Positivism: The Mirage of the Middle Way, 9
OXFORD J. LEGAL STUD. 462, 462 (1989) (concluding that modern “normative” positivism is “incoherent”); David
Dyzenhaus, The Genealogy of Legal Positivism, 24 OXFORD J. LEGAL STUD. 39 (2004); Andrew Halpin, The
Methodology of Jurisprudence: Thirty Years off the Point, 19 CAN. J. LAW & JURIS. 67 (2006).

15
See, e.g., James Allan, A Modest Proposal, 23 OXFORD J. LEGAL STUD. 197 (2003); Brian Leiter, Beyond the
Hart/Dworkin Debate: The Methodology Problem in Jurisprudence, 48 AM. J. JURIS. 17 (2003); David Lyons, Review,
Principles, Positivism, and Legal Theory, 87 YALE L.J. 415, 424–25 (1977); Michael S. Moore, Hart’s Concluding
Scientific Postscript, 4 LEGAL THEORY 301 (1998); Schauer, supra note 2, at 31; Brian Z. Tamanaha, The
Contemporary Relevance of Legal Positivism, 32 AUST. J. LEG. PHIL. 1 (2007).

16
See, e.g., Martin Krygier, The Concept of Law and Social Theory, 2 OXFORD J. LEGAL STUD. 155, 155–67 (1982);
Lacey, supra note 5, at 947–49; Andrew Lewis, Legal Positivism—Some Lessons from Legal History, in POSITIVISM
TODAY, supra note 4, at 65, 67–73; Philip Schofield, Jeremy Bentham and Nineteenth‐Century English
Jurisprudence, 12 J. LEGAL HIST. 58, 61–75 (1991); Twining, supra note 13, at 119, 123–31.

descriptive or normative theory. 17 They also disagree on the precise boundaries of legal
positivism (such as whether Ronald Dworkin may be considered a positivist). 18 And, of
course, they disagree on whether legal positivism, however understood, is descriptively
true or theoretically coherent or useful. Despite all of this disagreement, there is also
implicit agreement among these accounts about certain features of legal positivism. These
areas of overlap are, however, different from those, such as the “separability thesis” 19 or
the “sources thesis,” often associated with legal positivism. 20

First, all seem to agree that to the legal positivist (and perhaps to any legal theorist,
positivist or not), law can be differentiated from other things that exist. 21 (Of course,
certain legal theorists nearly universally understood not to be positivists also subscribe to
this view. 22 I return to the question of whether these points uniquely distinguish legal
positivists below.) Second, positivists are described as agreeing that an important aspect
of law is its character as a social phenomenon made up of interrelated communicative acts
involving the identification and provision of reasons for action. In other words, law is a
discursive and normative system. 23 Third, as a result of these first two premises, second‐
order accounts seem to agree that to the legal positivist, the system of law has limits; it
does not coincide with the universe of experience or the full range of ordered behavior.
Different accounts put this point in different ways. Some focus on the phenomenon
17
See, e.g., Deryck Beyleveld & Roger Brownsword, The Practical Difference Between Natural‐Law Theory and
Legal Positivism, 5 OXFORD J. LEGAL STUD. 1, 31 (1985) (arguing that “the ultimate reason for choosing between
natural‐law theory and positivism is not the moral reason but the reason of theoretical superiority”).

18
Dworkin characterized his theory as a “general attack on positivism.” Dworkin, supra note 6, at 20; see also
Dyzenhaus, supra note 8, at 712, 716 (noting that a “Dworkinian” judge would not be a legal positivist). But
others have argued that Dworkin is nevertheless a legal positivist in some sense. See, e.g., Leiter, supra note 15,
at 27; Perry, supra note 13, at 317; Stephen Perry, The Varieties of Legal Positivism, 9 CAN. J.L. & JURIS. 361, 361
(1996).

19
See, e.g., Leslie Green, Positivism and the Inseparability of Law and Morals, 83 N.Y.U. L. REV. 1035, 1035–36
(2008); Gerald Postema, Law’s Autonomy and Public Practical Reason, in THE AUTONOMY OF LAW, supra note 4, at
79, 80.

20
See, e.g., Coleman, supra note 10, at 140; Gardner, supra note 13, at 199.

21
This point is related to but distinct from the so‐called separability thesis. Voluminous citations would be
required to support this assertion fully and specifically; in identifying these points of agreement, I have drawn on
the sources in surrounding footnotes.

22
See JOHN FINNIS, NATURAL LAW AND NATURAL RIGHTS 14 (1980) (describing law as “a significantly differentiated type
of social order”).

23
Commitment to this position does not imply a normative or descriptive aim, a particular mode of argument or
description, or any particular characterization of the acts in question. Since Hart, this aspect of the positivist
position has probably most often been discussed in terms of the following of rules, but work on rule following is
best understood as a particular way of discussing the reason‐seeking function of law.

identified by Hart in his 1957 essay on the separation of law and morals. 24 Others focus on
the distinction between norms that are intentionally generated and others that are not.
Still others focus on the “exclusionary” 25 or “limited” 26 nature of legal reasons for action.
All of these are different ways of making the same point. Finally, second‐order accounts of
legal positivism in the Anglo‐American tradition agree that for the positivist, law
necessarily involves some form of higher‐order discourse that both concerns and carries
out the acts of delimitation constitutive of legal systems, what Hart called “secondary
rules.” 27 (This position is commonly associated with Hart; below, I discuss how pre‐Hart
theorists also described law as necessarily involving some such higher‐order content.) 28

In addition, a point increasingly common to these accounts of legal theory is the
observation that all positivist theories of law are necessarily partial accounts of their
subject matter, so that any given theorist can only hope to capture one or a few aspects of
the nature of law. Both theorists of the nature of law and accounts of these theories have,
more and more, come to acknowledge this point explicitly. 29

II. What This Account Adds

My account is consistent with but extends both the accounts surveyed above and other
more specific positions taken by particular legal theorists. In this section I clarify how my
perspective is related to these other positions.

To recapitulate, the points of agreement among the accounts of legal positivism discussed
above are (1) to the positivist, law is an actually existing social phenomenon necessarily
including a discursive system including practices of identifying and providing reasons, and
it is usefully considered as such; (2) to the positivist, this system is limited; and (3) to the
24
Hart, supra note 2, at 599, 601 n.25.

25
See, e.g., JOSEPH RAZ, PRACTICAL REASON AND NORMS 35–48 (1990).

26
See Schauer, supra note 11, at 1915–16.

27
H.L.A. HART, THE CONCEPT OF LAW 94 (1994).

28
See infra notes 99–107 and accompanying text.

29
See H.L.A. Hart, Comment, in ISSUES IN JURISPRUDENCE, supra note 4, at 35, 39; but see Stephen R. Perry,
Interpretation and Methodology in Legal Theory, in LAW AND INTERPRETATION: ESSAYS IN LEGAL PHILOSOPHY 97, 120
(Andrei Marmor ed., 1995) (arguing that Hart saw his theory of law as the only “possible conceptual analysis of
law”). Joseph Raz makes this point often. See, e.g., The Problem About the Nature of Law, 21 U. W. ONTARIO L.
REV. 203, 218 (1982); On the Nature of Law [hereinafter Raz, On the Nature], in BETWEEN AUTHORITY AND
INTERPRETATION: ON THE THEORY OF LAW AND PRACTICAL REASON [hereinafter RAZ, BETWEEN] 91, 97–98 (2009) (“While the
law has many essential features we are not aware of all of them. . . . The study of jurisprudence is never‐ending,
for the list of the essential properties of law is indefinite.”). See also FINNIS, supra note 22, at 17; KRAMER, supra
note 4, at 216–27.
positivist, legal systems include mechanisms for sustaining their own boundaries, or
effectuating these limits on the system, in the form of communications referring to the
system and the reasons it offers. Further, although this point is not universally
acknowledged, no theorist has captured all of the features of law in a single account.

I will take these points to constitute the core of “legal positivism” as it is understood by
contemporary Anglo‐American legal theorists. It is possible that none of these points
distinguish legal positivists from legal theorists in general, a question to which I return in
Section C.II.3 below. In addition, the understanding of legal positivism proposed here is
not entirely derived from these earlier accounts, which share two related blind spots. With
a few exceptions, these theories of legal theory have not explored the social contexts in
which positivist legal theory has been produced; the exceptions have explored only the
30
contexts within which specific theorists have worked. These second‐order accounts have
also largely failed to explore the curious parallel between the characteristics of law, on the
positivist account, and the characteristics of theoretical discourse about law. 31 The
discussion in Section C suggests that considering these issues—one a positivistic concern,
the other a formal one—indicates that post‐positivism is not necessarily a realistic goal.

Features of the account presented here also closely resemble, but depart from, specific
positions taken by particular legal theorists. My distinction between first‐ and second‐
order theories, for instance, recalls Stephen Perry’s distinction between methodological
32
and substantive positivism, although my conclusions differ from his. Perry pointed out
that the term “positivism” has been used to refer to two types of position, one a
substantive position about the nature of law (the position that there is no necessary link
between moral normativity and legal normativity), the other a methodological position
about the nature of legal theory (the position that legal theory should be purely
descriptive, rather than normative). 33 This distinction is more specific than my distinction
between first‐ and second‐order theory. Perry’s main point in making it is that it is
conceptually inconsistent to defend a substantively positivist position using
methodological positivism, since this amounts to treating a normative subject in a way that
cannot account for its normativity. 34 In contrast to Perry, I argue below that the conditions

30
See supra note 16.

31
Some second‐order accounts of positivism have explored the relationship between “external” and “internal”
legal theory, and the relationship between this distinction and the necessarily partial nature of legal theory. See,
e.g., Perry, supra note 29. Others have explored the contexts in which legal theory is produced. See, e.g., Lacey,
supra note 5, at 947–48; Schofield, supra note 16, at 69–70, 82, 85. But the link between these two topics has
been much less scrutinized.

32
See Perry, supra note 13; Perry, supra note 18; Perry, supra note 29.

33
Perry, supra note 13, at 311.

encouraging the production of what he calls methodological positivism may actually
encourage (without requiring) substantively positivist positions.

Finally, my position may also at times seem reminiscent of the claim, often attributed to
Hart, that a satisfactory account of the concept of law must proceed from an “internal
point of view.” 35 On one account, this is the claim that, to give a full account of law, a
theorist must grasp the way law is experienced by legal professionals. This requirement
arguably leads legal theory to develop formal or systematic characteristics mirroring those
of its subject matter, although it need not do so. 36 I seek to make the related but distinct
point that because law and academic legal theory are practices that historically developed
in parallel, we should not be surprised to find that they share certain characteristics (rather
than that legal theory must account for a particular subset of these characteristics in order
to give an informative account of law).

In coming to this conclusion, I am indebted on the one hand to works that have looked at
37
the institutional context within which legal positivist theory has been produced and at
the connection between this context and the social and institutional history of law itself, 38
and on the other to a literature on the sociology of academic and legal expertise that is
seldom linked to metatheoretical debates in jurisprudence. 39 This work indicates that the
partiality of legal theory, and some of its specific characteristics, are usefully considered as
functions of the social and institutional context within which it is created.

34
Id. at 347, 354.

35
See Perry, supra note 29; HART, supra note 27, at 91 (opining that “all our criticisms of the prediction theory of
obligation may be best summarized as” the criticism that they “define . . . out of existence” “the internal aspect of
obligatory rules”).

36
See Perry, supra note 29. See also HART, supra note 27, at 91, 243–44.

37
See Krygier, supra note 16; Lacey, supra note 5; Frederick Schauer, (Re)Taking Hart, 119 HARV. L. REV. 852 (2006)
(reviewing NICOLA LACEY, A LIFE OF H.L.A. HART: THE NIGHTMARE AND THE NOBLE DREAM (2004)); Schofield, supra note
16.

38
See, e.g., Harold J. Berman, The Origins of Western Legal Science, 90 HARV. L. REV. 894 (1977); Lewis, supra note
16; David Sugarman, Legal Theory, the Common Law Mind and the Making of the Textbook Tradition, in LEGAL
THEORY AND COMMON LAW 26 (William Twining ed., 1986).

39
See, e.g., ANDREW ABBOTT, THE SYSTEM OF PROFESSIONS: AN ESSAY ON THE DIVISION OF EXPERT LABOR (1988) [hereinafter
ABBOTT, SYSTEM OF PROFESSIONS]; ANDREW ABBOTT, CHAOS OF DISCIPLINES (2001) [hereinafter ABBOTT, CHAOS]; TONY
BECHER, ACADEMIC TRIBES AND TERRITORIES: INTELLECTUAL ENQUIRY AND THE CULTURES OF DISCIPLINES (1989); HARRY COLLINS &
ROBERT EVANS, RETHINKING EXPERTISE (2007); RANDALL COLLINS, THE SOCIOLOGY OF PHILOSOPHIES: A GLOBAL THEORY OF
INTELLECTUAL CHANGE (1989).

III. Why Seek Post‐Positivism?

Having clarified what I take to be the core of “legal positivism” as it is understood by Anglo‐
American legal theorists, and the scope of my objectives, I turn in this section to a final
background question: the reasons a legal theorist might seek to move beyond positivism,
to become a “post‐positivist.” So far, Neil MacCormick is most visible Anglophone legal
theorist to have applied this label to himself repeatedly and to have explained in detail
why he considers himself a “post‐positivist.” In this section, I briefly identify several
possible reasons for the impulse to label oneself “post‐positivist,” not all of them cited by
MacCormick. A number of these reasons are intimately linked to the phenomenon
described above—the staggering proliferation of first‐ and second‐order legal theories over
the past several decades, the precise period within which MacCormick wrote. This section
provides a foundation for my examination, in Section C, of how, if at all, MacCormick’s
position differs from the legal positivist commitments discussed above.

As it has been used from the mid‐twentieth century on, first in literary criticism and then in
many other areas, the “post‐” label signifies the desire for a fresh start, based on
40
disillusionment with existing tradition. Use of the prefix implies that the root tradition is
no longer viable in its original form. This is how MacCormick uses the term. 41 A legal
theorist might conclude that the root tradition—legal positivism—is no longer viable for
either conceptual or prudential reasons. Conceptual reasons for seeking to move beyond
legal positivism could include (1) dissatisfaction with the increasingly narrow questions
addressed by positivists and their apparently increasingly trivial conclusions 42 or (2) a
conclusion that one or more of the premises or methods of legal positivism have been
discredited or are irreconcilable. (Note that this second reason, while it is a common
critical approach in legal theory, 43 cannot be a sound reason for rejecting legal positivism
unless the theorist has exhaustively identified the premises or methods of legal positivism.)
Prudential reasons for identifying a position as post‐positivism might include (3) a desire to
free the theorist’s work from the pejorative connotations associated with the “legal
40
See Malcolm Bradbury, What Was Post‐Modernism? The Arts in and after the Cold War, 71 INT’L AFFAIRS 763,
767–74 (1995) (discussing Irving Howe’s initial use of the prefix in 1959 and its subsequent shifts in meaning in
different contexts and periods).

41
See MACCORMICK, INSTITUTIONS, supra note 1, at v (explaining abandonment of “positivist” label as a result of
debates over the issue).

42
See, e.g., Twining, supra note 13, at 123–25 (suggesting that John Austin’s work initiated this narrowing of the
issues addressed by legal theory); R. George Wright, Does Positivism Matter?, in THE AUTONOMY OF LAW, supra note
4, at 57, 57, 68 (arguing that legal positivist positions are trivial).

43
Many legal theorists have taken this position, but prior to MacCormick they described themselves, and would
most likely have been described as, anti‐positivist rather than post‐positivist. See, e.g., WILLIAM E. CONKLIN, THE
INVISIBLE ORIGINS OF LEGAL POSITIVISM: A RE‐READING OF A TRADITION 3, 68, 304 (2002); Fueßer, supra note 8, at 119,
120; George, supra note 12, at 321, 330; Tamanaha, supra note 15, at 35–36.
positivism” label; 44 (4) exhaustion with the volume of material produced under the
positivist rubric; and/or (5) a desire to mark the theorist’s work as original rather than
derivative, that is, a strategy of reputational or theoretical branding. 45

The distinction between conceptual and prudential reasons for using the post‐positivist
label is important because the theory in question is more likely to be genuinely post‐
positivist—transcending or fundamentally departing from the core positions outlined
above—if the use of “post‐positivist” label is supported by conceptual, rather than just
prudential, reasons. As I suggest below, it seems that MacCormick’s chief reasons for
applying the label to himself were probably, in fact, prudential, and that powerful
incentives will continue to deter any widespread rejection or denunciation of the core
preoccupations of legal positivism.

C. The Possibility of Post‐Positivism

In this section, I argue that legal positivism is both a defensible mode of theorizing about
law and an inescapable one. We cannot hope to eliminate it entirely from the landscape of
legal theories, as long as that theorizing proceeds within an institutional context similar to
its present one. This is because modern legal theorists, positivist and non‐positivist, are
always simultaneously concerned with two domains: (1) that of law, about which they
theorize, and (2) that of academic theoretical inquiry, within which they produce their
work. Both of these domains affect the scope of what it is possible for legal theorists to
think and write. All legal theorists write from within the second domain; legal positivist
theorists happen to produce theoretical accounts from that second domain that bear some
of the same characteristics they also ascribe to the first domain, law. This parallel is
sometimes represented as a deliberate theoretical choice, but deliberate or not, it is
encouraged by the parallels between the social and institutional histories of the two
domains, parallels that in turn generate parallel incentives for the production of higher‐
order commentary.

My discussion in this section has three parts. In Section C.I, I argue that the reasons
suggested by MacCormick for moving beyond legal positivism are not compelling. I also
argue that MacCormick did not succeed in this goal. Section C.II examines some of the
factors that make post‐positivism a quixotic goal, through a review of the history of the

44
See, e.g., Greenawalt, supra note 2, at 19 (noting that “the label ‘legal positivism’ may be mainly a matter of
rhetorical force, now usually negative”); SEBOK, supra note 2, at 2 (noting that the “positivist” term has “in recent
years . . . become a pejorative in modern American legal circles”); Schauer, supra note 2, at 32–35 (discussing
pejorative uses of “positivist” label).

45
Cf. BECHER, supra note 39, at 70.

legal positivist negotiation of the two domains identified just above. In Section C.III, I begin
to address possible objections to my argument, a project continued in Section D.

I. MacCormick’s Post‐Positivism

As noted, it seems that for MacCormick, prudential motivations for resort to the “post‐
positivist” label might have been paramount. 46 He does not take the position that the
premises of legal positivism have been discredited; 47 indeed, he consistently identified his
work as “in the same tradition” as that of Hart. 48 Nor does he seem troubled by the nit‐
picking details of legal positivist controversies, since he engages in and acknowledges the
relevance of many of those quarrels. 49 MacCormick in fact admits that the justifications he
provides for his self‐labeling as post‐positivist do not really distinguish his position from
legal positivism. He writes that he identifies himself as post‐positivist because he believes
that “law is necessarily geared to some conception of justice,” a moral concern. 50 But he
concedes that it is only the “more austere and rigorous forms” of legal positivism that
“absolutely exclude the possibility that there is any moral minimum that is necessary to the
46
MacCormick’s use of the label may also be related to terminological bleeding from one specialty to another.
The term “post‐positivist” is not uncommon in work on international law and international relations, another field
in which MacCormick wrote. See, e.g., INTERNATIONAL THEORY: POSITIVISM AND BEYOND (Steve Smith, Ken Booth &
Marysia Zalewski eds., 1996); Thomas Biersteker, Critical Reflections on Post‐Positivism in International Relations,
33 INT’L STUD. Q. 263 (1989); Stan Gontarek, International Legal Theory: Positivist, Naturalist, and Much More, 1
INT’L LEGAL THEORY 5 (1995). In this subfield, the term is used not to refer to a position beyond or following legal
positivism but in a manner borrowed from its meaning in the philosophy of science, where it refers to post‐logical
positivist theoretical positions. See, e.g., RUTH GROFF, CRITICAL REALISM, POST‐POSITIVISM, AND THE POSSIBILITY OF
KNOWLEDGE (2004); JOHN H. ZAMMITO, A NICE DERANGEMENT OF EPISTEMES: POST‐POSITIVISM IN THE STUDY OF SCIENCE FROM
QUINE TO LATOUR (2004). Larry Laudan is known for having argued that these “post‐positivist” positions share
important assumptions with the positivism they claim to move beyond and are largely an extension of positivism
rather than a true departure from it. See LARRY LAUDAN, BEYOND POSITIVISM AND RELATIVISM: THEORY, METHOD, AND
EVIDENCE (1996).

47
For example, he maintains that “law and morality are conceptually distinct.” MACCORMICK, INSTITUTIONS, supra
note 1, at 261, 264.

48
MACCORMICK, INSTITUTIONS, supra note 1, at 278–79. See also MACCORMICK, HART, supra note 1, at 16, 167;
MACCORMICK & WEINBERGER, supra note 3, at 7.

49
See, e.g., MACCORMICK, INSTITUTIONS, supra note 1, at 25–30 (discussing various aspects of theoretical debates
over rules, norms, exclusionary reasons, discretion, and standards); id. at 56–57 (offering critique of received
understanding of Hart’s Rule of Recognition concept); id. at 62–73 (clarifying distinction between rules and
habits); id. at 161–65 (discussing validity and defeasibility of legal norms).

50
MACCORMICK, INSTITUTIONS, supra note 1, at 264. See also id. at 4 (writing that he believes that “some minimum
of justice is essential” to legal validity). MacCormick also points to his incorporation of positions associated with
legal theorists identified as natural law theorists as well as legal positivists. Id. at 279. But as has often been
discussed, many legal positivists have moderated their position to accommodate criticisms offered by natural law
theorists or non‐positivists, without relinquishing their commitment to the positions described above in Section
B.I.2. See, e.g., Robin Bradley Kar, Hart’s Response to Exclusive Legal Positivism, 95 GEO. L.J. 393 (2007).
existence of law as such.” 51 (I am not the first to note that the theory articulated in
Institutions of Law, which is a restatement of theories MacCormick began developing much
earlier, is best understood as an extension of legal positivism, one that draws on
theoretical sources outside the mainstream tradition of legal positivism, rather than a
movement beyond its basic perspective.) 52

But none of the details of MacCormick’s position in Institutions of Law, his final summation
of his theoretical commitments, distinguishes those commitments from the core legal
positivist commitments described above. MacCormick identifies the heart of his
“institutional” theory of law as the proposition that law is an “institutional normative
order” 53 that is “heteronymous [i.e., providing reasons for action external to agents], as
well as authoritative and institutional,” in contrast to morality, “which is autonomous,
discursive, and controversial.” 54 Thus, MacCormick describes law as a social phenomenon,
systemic in character, involving the communication and suspension of particular
(autonomous or moral) reasons for action. He also understands legal systems and
institutions as necessarily containing self‐referential components; in this aspect of his
account, he draws on the social systems theory of Niklas Luhmann and Gunther Teubner. 55
Indeed, one of the primary ways in which MacCormick moves “beyond” legal positivism in
this final work is by considering the work of theorists outside the legal positivist canon,
such as Luhmann, Teubner, Karl Popper, Michel Foucault, and Sundram Soosay. 56 But the
use of a vocabulary that is partly novel does not cause MacCormick’s theory to differ more
from admittedly positivist positions than those positions already differ from one another. 57

51
MACCORMICK, INSTITUTIONS, supra note 1, at 278.

52
See Huib M. De Jong & Wouter G. Werner, Continuity and Change in Legal Positivism, 17 L. & PHIL. 233, 240, 249
(1998). To be sure, there are some ways in which MacCormick moves beyond the concerns of prior legal
positivists, as outlined above. He suggests, for example, that law might not be reducible to discourse. See
MACCORMICK, INSTITUTIONS, supra note 1, at 62–73. He also suggests that judgments regarding the legal validity of
norms occur on a continuum rather than being binary judgments. Id. at 161–65, 257–58. And as noted below,
he draws on a wider range of theoretical material than the traditional legal positivist does. See infra note 56 and
accompanying text.

53
MACCORMICK, INSTITUTIONS, supra note 1, at 1, 60.

54
Id. at 255. See also Neil MacCormick, The Concept of Law and The Concept of Law, in THE AUTONOMY OF LAW,
supra note 4, at 163, 164–71. MacCormick uses the term “discursive” in a narrower sense than I do, to refer to a
disputational setting rather than to practices of verbal articulation.

55
MACCORMICK, INSTITUTIONS, supra note 1, at 24, 177–78, 184, 289, 302–03. MacCormick claims that his theory is
not “derived from” or
“entailed by” Luhmann’s systems theory, but that the two are “strikingly mutually compatible.” Id. at 302.

56
MACCORMICK, INSTITUTIONS, supra note 1, at 65, 154, 292–93.

It seems, then, that MacCormick can only be using the “post‐positivist” label for one of the
prudential purposes I identified above: as a device for avoiding unwanted negative
connotations or for asserting that his intellectual position is distinguishable from others.
As suggested above, these reasons for using the label cannot themselves establish the
conceptual or functional difference of a particular theoretical position from legal
positivism.

Still, even if MacCormick did not distinguish his position sufficiently from legal positivism to
justify labeling his position “post‐positivist,” it is possible that a differently formulated first‐
order theory might achieve this distinction so conclusively as to make positivist‐style
commitments seem truly obsolete. Below I consider whether it is feasible to expect that
legal theory will pass beyond positivism in this sense.

II. The Possibility of Post‐Positivism

Within legal theory, perhaps especially including the tradition of legal positivism, there is
an established practice of drawing on sociological concepts to flesh out accounts of law,
even where the primary conceptual vocabulary comes from other analytical or
58
philosophical work. As noted above, there is also a limited tradition of reflection on the
relationship between the discourse of legal theory, including legal positivism, and the
discourse of law itself. 59 But most of the work reflecting on this parallel is descriptive,
57
MacCormick comes close to acknowledging this when he acknowledges that scholarship can only ever hope to
provide a partial account of its subject matter. MACCORMICK, INSTITUTIONS, supra note 1, at 301–02 (asserting that
scholarship cannot claim “to have found some ‘Archimedean’ point outside, perhaps ‘above,’ the practical activity
from which it looks down on the activity and sees that it represents nothing real. Rather, as Luhmann points out,
it involves a kind of self‐observation of the legal system from within it, but an observation that is not aimed
directly at the solution of particular current practical problems arising within it”).

58
See, e.g., Lacey, supra note 5, at 950–57; Perry, supra note 29, at 97 (“As a discipline jurisprudence claims . . . to
be both a branch of practical philosophy . . . and a social science of a certain kind.”). But see Roger Cotterrell,
Why Must Legal Ideas Be Interpreted Sociologically?, 25 J.L. & SOC’Y 171 (1998). MacCormick’s incorporation of
Luhmannian systems theory into his account of law can be viewed as a logical extension of this tradition. See
MACCORMICK, INSTITUTIONS, supra note 55 and accompanying text.

59
See, e.g., Krygier, supra note 16, at 164, 167 (noting how different disciplinary commitments of analytical
theorists and social scientists studying law lead them to identify different characteristics of legal systems); Lacey,
supra note 5, at 950–57 (providing critique of H.L.A. Hart’s respect for and disciplinary allegiance to analytic
philosophy); Lewis, supra note 16, at 65, 70–71 (presenting descriptive account of parallel histories of valorization
of autonomy in legal practice and legal theory); Schauer, supra note 37, at 858–69 (discussing disciplinary
commitments and dimensions of twentieth‐century positivism in the wake of Hart); Twining, supra note 13, at
129–30 (describing history of particularizing conclusions of post‐Bentham legal theorists, and parallels between
particularization in legal practice and in jurisprudence). Some work, to be sure, has rejected the thesis that the
legal domain involves the kind of partiality or suspension identified by legal positivism, while acknowledging that
legal theory is necessarily partial, and draws from these premises the conclusion that legal positivism in one or
more of its varieties is intellectually incoherent. But this work argues that the positivist description of law is
inaccurate; it does not seek to refute the idea that legal positivism is an inevitable form of legal theory in the
world we inhabit. See, e.g., Goldsworthy, supra note 12, at 451–52 (arguing that the only plausible legal position
simply identifying analogies between legal practice and legal theory. In this section, I use
the first of these techniques to extend the analysis begun by the second, looking at the
social and institutional context within which legal theory is produced and the factors
encouraging producers of theory to give their work positivist traits.

1. Two Domains

In a 1981 article, Owen Fiss explored a question related to my concern here: whether
commitment to the enterprise of generating a theoretical account of law (or “cognitive
positivism,” identified by Fiss with pure description) logically entails commitment to legal
60
positivist conclusions, chiefly the so‐called “separation thesis” (or “ethical” positivism).
61
Fiss concluded (as others have since ) that no such relation is necessary, but suggested
that the lack of a well‐developed critical literature on the practices of legal scholarship
made it impossible for him to venture any firm conclusions about the necessary shape of
legal theory. 62

In the years since Fiss wrote, the literature on practices of legal scholarship and academic
expertise more generally have grown significantly. 63 (The consensus emerging from this
literature is reflected in the increasingly common position, noted above, regarding the
necessary partiality or incompleteness of legal theory. 64 But that observation is just one of
a larger set of common observations about the Western academic domain.) This literature
agrees that, starting in the early modern period, the academic domain has been more and
more characterized by the policing of membership through extensive discursive
indoctrination 65 and the evaluation of various forms of discursive performance. 66
from the “internal point of view” is a natural law position); Perry, supra note 13, at 347 (arguing that
methodological positivism is inconsistent with the normativity of its descriptive object, law); Postema, supra note
8, at 165–66 (arguing that history of legal philosophy is a history of attempts to reconcile irreconcilable theses
regarding legal normativity and social existence of legal institutions); Waldron, supra note 9, at 426, 432–33
(discussing “asymmetry” between legal theory as grasped by actors within the legal system from the internal
point of view and as grasped by academic theorists).

60
Fiss, supra note 13, at 1008.

61
See, e.g., Perry, supra note 13, at 312–14 (discussing Hart’s mixture of methodological and substantive
positivism); Waldron, supra note 9, at 432–33.

62
Fiss, supra note 13, at 1016.

63
See, e.g., sources cited supra note 39; Sugarman, supra note 38; Christopher Tomlins, Framing the Field of Law’s
Disciplinary Encounters: A Historical Narrative, 34 L. & SOC’Y REV. 911 (2000).

64
See supra note 29 and accompanying text.

65
See BECHER, supra note 39, at 32–44; COLLINS & EVANS, supra note 39, at 14, 24; Catherine Kemp, The Uses of
Abstraction: Remarks on the Interdisciplinary Efforts in Law and Philosophy, 74 DENVER U. L. REV. 877, 879–85
According to this literature, the modern university, one of the institutions that has
emerged from this dynamic, has distinctive material and conceptual features. Materially,
those acting within the institution depend to some extent on external support and must
therefore justify their activities to social actors outside the institution. 67 At the same time,
the institutional structure is largely self‐reproducing, both materially and conceptually, and
its self‐reproduction occurs primarily through discourse. 68 These features make the
structure a competitive arena, and the resulting competition encourages academics to
draw distinctions between their positions and those of their forebears and
contemporaries. 69 Sociologist Andrew Abbott calls this characteristic of academic
discourse the drawing of “fractal distinctions.” 70 Although Abbott focuses on the fractal
conceptual structure of the social sciences, others have made similar observations about
academic philosophy 71 and the academic study of law. 72 The nature of these distinctions is
such as to encourage the use of a rhetoric of innovation, even when a theorist is not truly
making a novel point. 73 And these features also encourage an implicit normative hierarchy
of intellectual positions, one in which abstract, conceptual, or higher‐order accounts of a
subject matter attain higher prestige, both within each specialty and across fields, 74
despite the acknowledged impossibility of attaining a “theory of everything.” For the
academic, a high measure of self‐awareness and facility with abstract self‐referential
discourse is a prized and rewarded trait.
(1997); WALTER J. ONG, RAMUS, METHOD, AND THE DECAY OF DIALOGUE: FROM THE ART OF DIALOGUE TO THE ART OF REASON
306 (2nd ed., 2004).

66
See ABBOTT, CHAOS, supra note 39, at 130–31, 137–38, 140–42; COLLINS & EVANS, supra note 39, at 45–76; COLLINS,
supra note 39, at 25–36.

67
ABBOTT, CHAOS, supra note 39, at 141; COLLINS & EVANS, supra note 39, at 9.

68
See ABBOTT, CHAOS, supra note 39, at 130, 140, 147–49; COLLINS & EVANS, supra note 39, at 7, 24–27, 30–31, 39,
86; COLLINS, supra note 39, at 25–36.

69
ABBOTT, CHAOS, supra note 39, at 137 (noting that “[k]knowledge experts compete with one another through
redefinition of each other’s work”); COLLINS, supra note 39, at 31, 71, 80.

70
ABBOTT, CHAOS, supra note 39, at 11–13, 138, 148.

71
See, e.g., COLLINS, supra note 39, at 31, 76, 80–81; HANS‐JOHANN GLOCK, WHAT IS ANALYTIC PHILOSOPHY? 245–46
(2008).

72
FIONA COWNIE, LEGAL ACADEMICS: CULTURES AND IDENTITIES 134, 198 (2004); Sugarman, supra note 38, at 26, 29, 34;
Tomlins, supra note 63, at 926–64.

73
ABBOTT, CHAOS, supra note 39, at 140–42, 148; BECHER, supra note 39, at 70.

74
See, e.g., ABBOTT, CHAOS, supra note 39, at 145–47; ABBOTT, SYSTEM OF PROFESSIONS, supra note 39, at 52–57, 79–
85, 102–104, 110–11, 118–21 (discussing phenomenon of “professional regression” into high‐status positions of
pure reflection on abstract knowledge linked with professional group); BECHER, supra note 39, at 57; COWNIE,
supra note 72, at 198.

Similar features characterize the modern legal domain as a set of social institutions.
Attaining expertise in this domain involves indoctrination into discursive practices. 75 Like
the academy, this set of institutions is both materially and conceptually self‐perpetuating.
Materially, lawyers’ expertise in the discourse used to navigate the institutions regulating
state power ensures their continued support by non‐lawyers. Andrew Lewis has argued,
along these lines, that lawyers have an inherent interest in taking a theoretical perspective
on their activities, since such a perspective allows them to “exclude the practical
consequences” of their judgment from consideration and thus insulate themselves from
76
criticism by outsiders and withdrawal of support. In these ways, lawyers’ maintenance
of their discursive expertise shapes legal discourse, ensuring not only that it requires
certain reasons for action to be set aside, 77 but also that it contains higher‐order discourse
about law itself. 78 As in the academic domain, legal discourse thus takes a “fractal” form,
even at the level of what Hart called primary rules. 79 And as in the academic domain, the
proliferation of higher‐order discourse reflects and creates a status hierarchy. Those who
work with such higher‐order legal discourse—such as appellate and constitutional lawyers
and judges—are widely accorded a higher status than their primary‐rule counterparts. 80
At the same time, legal discourse contains a significant element of what Harry Collins and
Robert Evans call “interactional expertise,” that is, fluency in the discursive conventions of
other specialties. 81 Indeed, legal practice seems to be one of the types of expertise that
Collins and Evans describe as “almost entirely devoted to gaining interactional expertise in
other specialisms,” since it borrows its content and meaning from responses to and
discourse about other social phenomena. 82

75
COWNIE, supra note 72, at 128–29; ELIZABETH MERTZ, THE LANGUAGE OF LAW SCHOOL: LEARNING TO “THINK LIKE A
LAWYER” 12–30, 207–23 (2007).

76
Lewis, supra note 16, at 66.

77
Id.

78
HART, supra note 27, at 94.

79
See J.M. Balkin, The Crystalline Structure of Legal Thought, 39 RUTGERS L. REV. 1 (1986). Using a different
vocabulary, Niklas Luhmann has explored this phenomenon extensively. See, e.g., NIKLAS LUHMANN, A SOCIOLOGICAL
THEORY OF LAW 41–48, 73–83 (1972).

80
Angela P. Harris & Marjorie M. Schultz, “A(nother) Critique of Pure Reason”: Toward Civic Virtue in Legal
Education, 45 STAN. L. REV. 1773, 1777 n.14 (1993) (noting prestige of appellate judges compared to trial judges in
America, and of certain substantive fields of law involving more secondary‐rule content and valuing “rationality”
over “emotion”); Deborah Jones Merritt, Who Teaches Constitutional Law?, 11 CONST. COMMENT. 145 (1994)
(concluding that constitutional law is high‐prestige specialty among American legal academics).

81
COLLINS & EVANS, supra note 39, at 35–39.


Legal theory generally and legal positivist theory in particular—the subject of Section B.I—
are generated from within the first of these domains, the academic, and seek to describe
or account for the second, the legal. Thus, legal theory is both a manifestation and a
description of what Andrew Abbott calls the “fractal fecundity” of certain cultural and
intellectual phenomena. 83 The characteristic assumptions of legal positivism, outlined in
Section B.I.2, observe these facts about law, but they usually do so without turning those
observations back upon the theoretical enterprise itself. If we engage in such self‐
reflection, we can see that legal positivist theory fixates on qualities of its subject matter
that have analogues in the institutions within which the theorists themselves are
operating—a preoccupation with reason‐giving, the systematic suspension of certain
considerations, communicative self‐reference. My point is not that legal theory inevitably
adopts these features because it is impelled to take on the features of the subject‐matter it
is explaining (either because this is necessary for a complete explanation or because it is a
84
kind of irresistible temptation). It is, rather, that because of the parallel social functions
and historical paths of these two practices—legal practice and theoretical inquiry—they in
fact share certain features, at least when they are regarded from a certain perspective.
Moreover, the institutional and conceptual dynamics within each domain provide strong
incentives for those working in each to fixate on just these features. The assumptions of
legal positivism are therefore over determined; those assumptions are demonstrably
accurate, if partial, descriptions of legal reality, and they are also types of descriptions
85
encouraged by the nature of the institution within which they are formed.

In the next section, I further explore, from a more historical point of view, how the
positions associated with legal positivism started to circulate just as both the modern
enterprise of theoretical inquiry and Western legal systems were starting to take their
modern form. 86 Before proceeding, however, I would like to clarify that I am not arguing
82
COLLINS & EVANS, supra note 39, at 38. See also Douglas W. Vick, Interdisciplinarity and the Discipline of Law, 31
J.L. & SOC’Y 163, 189–90 (2004) (noting that most interdisciplinary legal scholarship is theoretical rather than
empirical, probably because of structural and conceptual similarities between theory and doctrinal work).

83
ABBOTT, CHAOS, supra note 39, at 53–55.

84
Cf. The discussion of Hart’s argument for the necessity of adopting an internal point of view in Perry, supra note
29, at 97, 99–100.

85
Thus, theory mirrors practice not because “to study metaphors, one must do so metaphorically,” the
assumption criticized by Michael Moore in Interpreting Interpretation, in LAW AND INTERPRETATION, supra note 29, at
1, 26, but because, in fact, law does have these features (along with others), and academic theorizing also does, in
fact, have these features (along with others).

86
My position is not inconsistent with Gerald Postema’s argument that jurisprudence is an inherently practical
inquiry, since his sense of “practical”—having to do with normativity and with reasons for action—overlaps with
my sense of “theoretical,” which refers only to the abstract accounting for or explaining of a subject matter,
regardless of whether the account or explanation is normative or descriptive. Postema, supra note 12.
either that the discourse identifiable as legal positivism can be generated only by academic
theoreticians or that all academic legal theory is necessarily positivist. On the first of these
points, I believe that the description of judges as positivist is both meaningful and more
interesting than is usually realized; I return to this issue in Section D.I. I explain how my
argument differs from the position that all legal theory is necessarily positivistic in Section
C.III below.

2. The Legal Positivist Bridge

In this section, I briefly survey how five key figures in the development of the Anglo‐
American tradition of legal positivism—Hobbes, Bentham, Austin, Hart, and Raz—fit into
87
the analysis presented above. (Because I have already discussed MacCormick’s position,
I do not repeat that summary in this section.) In addition, this discussion helps to confirm
the features of legal positivism outlined in Section B.I.

Each of the five figures in this list wrote within the type of competitive intellectual context
described in the previous section, even though that context was still in the process of
formation for the earlier figures in the series. Collectively, the work of these figures has
supplied much of the basic conceptual vocabulary available in the domains of both
academic inquiry and law. For instance, while Hobbes did not write from the academic
domain as we know it today, he was one of the architects of the discourse that has shaped
88
that domain, as well as the Anglo‐American legal one. Bentham, too, was a key figure in
the development of the modern academic and legal domains. 89 Austin devoted much of
his career to the establishment of “the study of positive law as an autonomous scientific
discipline” within the academy. 90 All of the most influential twentieth‐century proponents
of first‐order positivist legal theory—Hart, Raz, and MacCormick—have benefited from
Austin’s efforts in this regard and have been, first and foremost, successful academics in
the modern competitive mold. 91

87
Although some accounts of legal positivism trace its origins to pre‐modern Europe, see, for example, CONKLIN,
supra note 43, at 14–32 (discussing Greek distinction between nomos and physis as analogous to positive law‐
natural law dichotomy), most accounts of legal positivism identify Thomas Hobbes as the first modern legal
positivist. See, e.g., Bix, supra note 13, at 18; Dyzenhaus, supra note 8, at 708; 58–60; Gardner, supra note 13,
204–05; Waldron, supra note 13, at 171.

88
See STEVEN SHAPIN & SIMON SCHAFFER, LEVIATHAN AND THE AIR‐PUMP: HOBBES, BOYLE, AND THE EXPERIMENTAL LIFE 100–02
(1985); Steve Fuller, Disciplinary Boundaries and the Rhetoric of the Social Sciences, 12 POETICS TODAY 301, 320
(1991).

89
See NEGLEY HARTE & JOHN NORTH, THE WORLD OF UCL 1828–2004 (2004).

90
Twining, supra note 13, at 123.

91
On Hart, see LACEY, supra note 37, at 112–208. See also Schauer, supra note 11, at 1951–53 (classing Austin
with Bentham and Kelsen as being concerned primarily with the “demarcation of law from its neighbors”).

All of these figures wrote in part to win authority for their subject‐matter, a goal not fully
distinguishable from winning personal and institutional security for the writers themselves.
Describing one’s subject‐matter as differentiable from other subject‐matters reflects
positively on the describer’s authority to make the description. 92 At the same time, that
differentiation cannot be too fine‐grained; the differentiator has an interest in claiming as
much intellectual territory as possible. 93 The incentives to keep differentiations of subject‐
matter (a focus on law) general are in tension with incentives to differentiate theoretical
positions (a focus on the attributes of law associated with legal positivism, and on each
writer’s differences from his predecessors). This tension has been continuously modulated
by these writers’ consistent description of law as basically discursive, normative, and
systemic. Hobbes explained law as necessitated by the transition into a post‐natural state
of language and as existing only to the extent the sovereign’s commands are articulated. 94
Bentham, similarly, defined “a law” as “an assemblage of signs declarative of a volition
conceived or adopted by the sovereign.” 95 Austin referred to the question of the nature of
law as a question of identifying “the principles, notions, and distinctions which are
common to systems of law.” 96 Hart, of course, differentiated his account from Austin’s in
part by pointing out that Austin’s account paid insufficient attention to the variety and
nuances of legal discourse. 97 And Raz, while critical of aspects of Hart’s account, has
similarly described law as “the authoritative voice of a political community.” 98

All of these figures have also described law as a system of a particular kind: one that
suspends the operation of certain reasons for action and that has self‐referential features.
For Hobbes, the Leviathan’s commands forbid certain actions and justifications as asocial
even though (or because) they are “natural”; 99 these commands contribute to the
maintenance of an artificial body created through language referring to that artificial body

92
See COLLINS, supra note 39, at 31.

93
See ABBOTT, CHAOS, supra note 39, at 147; COLLINS & EVANS, supra note 39, at 70–76.

94
See CONKLIN, supra note 43, at 73, 81–91, 98–88.

95
JEREMY BENTHAM, OF LAWS IN GENERAL 1 (1782). Indeed, Bentham has been described as “anticipat[ing] various
trends in twentieth‐century philosophy of language (including Frege’s and Wittgenstein’s ‘context principle’, some
views of logical positivists, and the development of speech act theory).” See Timothy Endicott, Law and
Language, Stanford Encyclopedia of Philosophy, available at http://plato.stanford.edu/entries/law‐language/.

96
JOHN AUSTIN, THE USES OF THE STUDY OF JURISPRUDENCE 365, 367 (1863).

97
HART, supra note 27, at 82–94.

98
Raz, On the Nature, supra note 29, at 99.

99
See CONKLIN, supra note 43, at 96.
and its operations. 100 Bentham’s expository jurisprudence—which anticipated in some
respects what Perry calls methodological positivism—sought to describe law as a system of
specific reasons, including higher‐order reasons relating to the generation of those
reasons. 101 Austin, too, though faulted by Hart for inadequate attention to the structure of
secondary rules, 102 did consider the constraint of legal systems by the higher‐order
discourse of constitutional law, 103 and suggested that law is inherently self‐referential in its
inability to represent the sovereign’s pure will. 104 Hart described his claim that legal
reasons must be distinguishable from moral reasons as entailed by the practices of
referring to something called “law” 105 and, of course, identified the inclusion of self‐
referring norms as a defining feature of a legal system. 106 And Raz has further advanced
Hart’s discussions of the normative and exclusionary functions of legal discourse. 107

Each of these theorists (among many others not discussed here) built on and modified his
predecessors’ ideas, while continuing to attribute certain consistent characteristics to the
subject‐matter he addressed. The pattern of modification of predecessors’ ideas has as
much to do with the gradually solidifying academic institutional system within which these
theorists functioned as it does with the (inherent) incompleteness of their predecessors’
explanations. Indeed, we could see this pattern of attributing a consistent set of
characteristics to the subject‐matter as a kind of confirmation that this conceptualization
meaningfully corresponds to an identifiable aspect of experience. This perspective offers

100
Id. at 82–86.

101
In the Fragment on Government, for example, Bentham stresses the need to acknowledge higher‐order norms
or reasons when he criticizes Blackstone for excessive focus on the law‐making power of the government (“the
right of Government to make Laws”) to the neglect of the rules governing and restricting that power (“the duty of
the Government to make Laws” and “the British Constitution”). See JEREMY BENTHAM, A FRAGMENT ON GOVERNMENT
6–7 (1776); id. at 53–55 (discussing sovereign’s subjection to law). See also Schofield, supra note 16, at 59–60;
Twining, supra note 13, at 121.

102
HART, supra note 27, at 18–25, 91–99.

103
JOHN AUSTIN, THE PROVINCE OF JURISPRUDENCE DETERMINED 257–58 (1863) (discussing constitutional “principles or
maxims which the sovereign habitually observes,” which it “is bound or constrained to observe,” and which are
known to those who might apply (nonlegal) sanctions to the sovereign in the event of its failure to observe these
principles).

104
CONKLIN, supra note 43, at 143–44.

105
See Hart, supra note 2, at 614–15, 620.

106
HART, supra note 27, at 79.

107
See, e.g., Raz, supra note 25, at 35–48, 141–48, 170–77.
another reason to suspect that the traits attributed to law by positivists are likely to remain
compelling focal points for theoretical accounts of legal phenomena. 108

III. Must Legal Theory Be Positivist?

My goal is to suggest that a broad “post‐positivism” is neither desirable nor truly possible
given the current institutional context of scholarly production in general and legal theory in
particular. In arguing that the theoretical position associated with legal positivism is both
accurate and compelling, however, I do not necessarily mean that it is impossible for an
academic legal theorist to be anything but a legal positivist, or that Anglo‐American legal
theory is trapped within a single paradigm, the implications of which are already or soon to
be exhausted, leaving nothing of interest for legal theorists to do. In this section, I address
the first of these possible concerns; Section D addresses the second.

My argument might be taken to suggest not only that post‐positivism is not desirable but
also that all legal theory is positivistic. Brian Bix has argued that such a position is
untenable because understanding legal theory as a “one‐party state” of this kind is not
109
illuminating. There are two problems with this contention from the perspective
advocated here. First, Bix’s conclusion regarding the desirability of a multi‐party state of
legal theory may be unrealistic and inconsistent with the nature of theoretical discourse. It
is possible that we not only do but must in some sense have a kind of “one‐party state” in
legal theory, if all legal theorists must embrace some or all of the premises identified above
in order to be comprehensible to and recognized by other legal theorists. 110 Second, even
if we assume that Bix is using the term more narrowly and accept his conclusion about the
undesirability of a one‐party state, his objection does not undermine my argument if it is
possible to identify features distinguishing legal positivism as I have described it from other
forms of theoretical inquiry into law. This is possible by virtue of the features I have
identified with legal positivism in Section B.I.2. Conceivable non‐positivist approaches to
the academic study of law include specific recommendations for the reform of particular
legal institutions; accounts denying one or more of the assumptions described in Section
B.I.2, such as the distinctiveness of law, the characterization of legal institutions as
systemic, or the existence of meaningful differences between first‐ and second‐order legal
discourse; and interdisciplinary accounts of law, like sociological and anthropological ones.

108
See, e.g., MACCORMICK, INSTITUTIONS, supra note 1, at 292–93 (discussing law, following Karl Popper’s
terminology, as a “World III” “thought‐object”). See also RAZ, BETWEEN, supra note 29, at 265, 269.

109
Bix, supra note 13, at 29.

110
Thus, John Finnis self‐identifies and is identified by others as critical of legal positivism, yet he embraces a
perspective on the nature of law similar in many ways to that described above. See Finnis, supra note 22. See
also Bix, supra note 9, at 1613, 1624.

All of these are possible positions, and many of them have been advanced by legal
theorists. The very fractal logic identified by Abbot that, I argue, ensures the perpetuation
of debates within what we think of as a legal positivist paradigm also ensures the
proliferation of such competing varieties of legal theory.

In other words, I am proposing not that legal positivism is the only possible mode of
academic discourse about law, or even the dominant one, but that without significant
institutional change in the academy, Anglo‐American legal theory is not likely ever to exit
the debate over whether the best explanation of law involves acknowledgment of its
discursive, normative, systemic, suspensive, and self‐referential characteristics. The nature
of theoretical discourse and the dynamics of the institutions within which it is produced
encourage attention to and efforts to clarify these characteristics of law—not to the
exclusion of other perspectives, but because these characteristics are especially well suited
as topics for the kind of theoretical discourse rewarded in these settings.

D. New Paths for Positivist Inquiry

My contention that legal positivism is probably here to stay does not imply that we must
resign ourselves to the endless recycling of familiar debates. Any number of interesting
theoretical questions and concrete legal phenomena remain largely unexplored from the
perspective I identify as positivist. In this section, I briefly discuss two examples: the
phenomenon of positivist adjudication (Section D.I), and the implications of the exclusivity
of legal and theoretical discourse (Section D.II).

I. Positivist Adjudicators

Ronald Dworkin’s work has been read as a critique of the accuracy of legal positivism as a
111
description of actual practices of adjudication. Yet real‐life adjudicators, including
judges in the United States—the system about which Dworkin wrote—have adopted a
rhetoric in their written opinions that resembles the idioms of legal positivism. 112 Some
111
See, e.g., Coleman, supra note 10, at 145–46; Soper, supra note 12, at 507–08, 512. Compare Dyzenhaus,
supra note 8, at 712, 716 (noting that a “Dworkinian” judge would not be a legal positivist) with Leiter, supra note
15, at 27 (noting that Dworkin might be considered an applied positivist seeking to describe adjudicative process),
Perry, supra note 13, at 317 (noting that Dworkin might be classed as a methodological positivist).

112
I am not the first to make this observation, but I hope in this section to suggest some implications that have
not been noted before. For previous characterizations of particular adjudicators and judicial rhetoric as examples
of legal positivism in practice, see, for example, G. Todd Butler, A Matter of Positivism: Evaluating the Legal
Philosophy of Justice Antonin Scalia Under the Framework Set Forth by H.L.A. Hart, 12 HOLY CROSS J.L. & PUB. POL’Y
47, 48, 59 (2008) (concluding that Scalia is a positivist because his opinions betray his adherence to the social‐
facts thesis and the separation thesis, identified by Butler as the “two fundamental tenets” “share[d]” by “all legal
positivists”); Beau James Brock, Mr. Justice Antonin Scalia: A Renaissance of Positivism and Predictability in
Constitutional Adjudication, 51 LA. L. REV. 623 (1991); Anita J. Allen, Autonomy’s Magic Wand: Abortion and
Constitutional Interpretation, 72 B.U. L. REV. 683, 693–94 (1992) (describing both Justice Scalia and Justice Thomas
commentators have even suggested that judicial rhetoric is by its nature positivistic. 113
More common is the identification of certain judges as positivist. U.S. Supreme Court
Justice Hugo Black has been referred to in this way, 114 but Justice Antonin Scalia is the
adjudicator now most commonly labeled a positivist. 115 If, as I have argued, the legal
positivist characterization of law is intimately tied to the academic institutional context in
which legal theory is produced, can a judge, who necessarily writes from a different
institutional position, truly be a legal positivist? In this section, I first consider the ways in
which a judge such as Justice Scalia may be considered a legal positivist in the sense I have
outlined here. Analysis of this question suggests new ways of connecting currently
disparate areas of legal scholarship to generate interesting new questions.

In the context of the framework presented here, calling a judge like Justice Scalia a
positivist might mean one of two things: that the judge speaks from an academic domain
despite his or her role as a judge, or that the judge adopts the rhetoric of the academic
domain. The latter seems to be the more accurate description of Justice Scalia. To be
sure, Justice Scalia was an academic before becoming a judge, and the application of the
116
legal positivist label to him is due in part to his scholarly work. But references to Justice
Black as a positivist cannot be explained in this way, and Justice Scalia has continued to be
identified as a positivist even after leaving the academy (although the precise sense in
which the term “positivist” is being used is not always clear). At any rate, while Justice
Scalia is more of a public intellectual than many judges are, 117 he does not now work an
academic context. Rather, references to him as a legal positivist, and his self‐identification
as such, appear mainly to refer to certain characteristics of his written opinions, that is, to
his adoption of a rhetoric of positivism. These characteristics include a stress on statutory
and constitutional text and precedent as the only legitimate reasons to offer in support of
as positivists); George Kannar, The Constitutional Catechism of Antonin Scalia, 99 YALE L.J. 1297, 1307, 1310, 1308,
1339 (1990) (referring to Scalia as positivist).

113
See, e.g., Gerald B. Wetlaufer, Rhetoric and its Denial in Legal Discourse, 76 VA. L. REV. 1545, 1555 (1990).

114
Brock, supra note 112, at 632.

115
See sources cited supra note 112.

116
See ANTONIN SCALIA, A MATTER OF INTERPRETATION: FEDERAL COURTS AND THE LAW 25 (1997); Antonin Scalia, Morality,
Pragmatism, and the Legal Order, 9 HARV. J.L. & PUB. POL’Y 123, 125 (1986) (“I have never been able to isolate
obligations of justice, except by defining them as those obligations that the law imposes.”). Similarly, Justice
Scalia’s dictum that “[t]here are times when even a bad rule is better than no rule at all” can be taken as an
articulation of the separability thesis. Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. CHI. L. REV. 1175,
1179 (1989) [hereinafter Scalia, The Rule of Law].

117
See David M. Zlotnick, Jusice Scalia and His Critics: An Exploration of Scalia’s Fidelity to His Constitutional
Methodology, 48 EMORY L.J. 1377, 1427–28 (1999) (arguing that Justice Scalia is more accurately characterized as a
politician than as a judge); Stephen A. Newman, Political Advocacy on the Supreme Court: The Damaging Rhetoric
of Antonin Scalia, 51 N. Y. L. SCH. L. REV. 907 (2006).
a judicial decision and a disavowal of reliance on “policy” (i.e., moral or practical)
considerations 118 and “emotion.” 119 In other words, in his opinions, Justice Scalia
represents the law as a discursive system in which some types of reasons are suspended or
excluded, and he does so through statements referring to the system itself. 120

References to Justice Scalia as a positivist are based on his repeated use of statements of
this kind, rather than on his occupation of a particular institutional role. 121 This does not
mean that Justice Scalia’s opinion‐writing technique offers support for the claim, often
attributed to Dworkin, that every judge is a jurisprude. 122 Rather, it suggests that Justice
Scalia deploys in his opinions what Harry Collins and Robert Evans have called “interactive
expertise” in the academic‐theoretical discourse of legal positivism. Collins and Evans
define interactive expertise as “the ability to master the language of a specialist domain in
the absence of practical competence” (the latter would permit the contribution of novel
propositions in the discourse of the domain—for example, Justice Scalia would exhibit
practical competence if he were to contribute to one of the theoretical debates on legal
123
positivism). Alternatively, because of his academic background, Justice Scalia’s rhetoric
might be understood as a form of “referred expertise,” “the use of an expertise learned in
one domain within another.” 124

118
E.g., Carden v. Arkoma Assocs., 494 U.S. 185, 196 (1990) (“The resolutions we have reached . . . can validly be
characterized as technical, precedent‐bound, and unresponsive to policy considerations . . . but . . . that has been
the character of our jurisprudence in this field.”).

119
E.g., Penry v. Lynaugh, 492 U.S. 302, 359 (1989) (Scalia, J., concurring in part and dissenting in part) (“It is an
unguided, emotional ‘moral response’ [in juries] that the [majority opinion] demands be allowed—an outpouring
of personal reaction . . ., an unfocused sympathy. Not only have we never before said the Constitution requires
this, but [in prior decisions we have] sought to eliminate precisely the unpredictability it produces.”).
Commentators have noted that this disavowal is belied by Justice Scalia’s use of colorful language to characterize
views with which he disagrees. See generally Newman, supra note 117.

120
In line with H.L.A. Hart, Justice Scalia has argued that the exclusion of certain considerations from the domain
of legitimate legal reasons advances certainty and predictability. Compare Scalia, The Rule of Law, supra note
116, with HART, supra note 27, at 42, 252 (arguing that a benefit of a rule of recognition is its enhancement of
certainty and predictability).

121
Because we can only detect the occupation of a legal positivist position through statements like those made by
Justice Scalia, the difference between occupying the institutional position associated with the generation of legal
positivist discourse and the use of positivist rhetoric by one functioning within another institution may not be all
that significant. Cf. COLLINS & EVANS, supra note 39, at 86.

122
RONALD DWORKIN, LAW’S EMPIRE 90 (1986) (“Jurisprudence is the general part of adjudication, silent prologue to
every decision at law.”).

123
COLLINS & EVANS, supra note 39, at 14.

124
Id. at 15.
A judge could borrow discursive expertise from the academic domain for use in the judicial
for many reasons, most of them prudential: to better understand or explain the judicial
domain, to contribute to the expansion of the intellectual territory claimed by that
domain, 125 to facilitate communication with experts in both domains, or to confer some of
the status or authority associated with the academic domain on the judicial one. Indeed,
to a non‐expert audience, “referred” or “interactive” expertise is indistinguishable from
what Collins and Evans call “contributory expertise,” 126 and when the status or legitimacy
of a borrowing discourse is questionable, the use of this kind of “parasitic” expertise may
be virtually the only method of legitimating the discourse in question. 127 In light of other
characteristics of his jurisprudence, it seems most likely that Justice Scalia borrows
positivist idioms to confer status, authority, and legitimacy on his opinions. 128

This understanding of Justice Scalia’s positivism helps to clarify the nature of the parallels
between the academic and legal domains discussed above. Justice Scalia is considered a
positivist judge because of the statements he makes in his opinions about the proper kind
of reasons to provide in support of a legal decision. But the type of discursive self‐
reference involved in these statements is not the same as the discursive self‐reference that
is, on the legal positivist account, inherent in law. Instead, it is a self‐reference that
imports the higher‐order discourse of another institutional setting, that of theories of legal
129
justification, reasoning, and authority.

Considering the phenomenon of positivist adjudication in this light raises several
interesting questions. One is the basic question of how typical Justice Scalia’s method is. If
positivist judging is defined as a matter of the deployment of interactive expertise, or as a
matter of discursive performance, then the content analysis of judicial opinions should be
able to tell us how often it occurs. 130 In itself, information about how often judges use the
language of positivism might not be of great interest. But contextualized properly, such
information would have significant practical and political, not to mention theoretical,

125
Id. at 70–76.

126
Id. at 52–54, 60–63.

127
As Collins and Evans argue, “distance lends enchantment,” that is, “the more distant one is from the locus of
the creation of knowledge in space and time the more certain will the knowledge appear to be.” Id. at 20.

128
See generally Kannar, supra note 112; Zlotnick, supra note 117. See also Matthew Kramer’s argument that
legal officials often act for prudential rather than moral reasons, KRAMER, supra note 4, at 64–77, and Jamal
Greene’s argument, not limited to Justice Scalia, about the reasons for the use of originalist rhetoric in judicial
and popular discourse, Jamal Greene, Selling Originalism, 97 GEO. L.J. 657 (2009).

129
See supra notes 116 & 117.

130
For an overview of work taking this type of approach, see Mark Hall & Ron Wright, Systematic Content Analysis
of Judicial Opinions, 96 CAL. L. REV. 63 (2006).
implications. One question to ask would be whether judges using the language of
positivism tend to decide cases in a way that differs systematically from the decisions of
non‐positivist judges. 131 Without relying on the terminology of positivism and expertise
used here, a number of scholars have already sought to determine whether there are any
correlations between particular types of judicial rhetoric and case outcomes. 132 But so far,
this work has remained largely ad hoc, due to the lack of any consensus on the conceptual
relationship between judicial rhetoric and legal decisionmaking. 133 The existing literature
on this subject would surely benefit from the sophistication that academic legal positivists
and other legal theorists have brought to bear on questions of justification and
authority. 134 Those attracted by the problems circulating in the literature of legal
positivism could likewise preserve the relevance of their inquiry into these issues by
contributing to the development of a theoretical framework for this empirical
investigation. 135

The above discussion of judges’ possible prudential motivations for using positivist rhetoric
in opinions also raises another empirical question. Regardless of whether and how it
relates to case outcomes, does use of this rhetoric serve an effective legitimating function?
Answering this question requires analysis not of the content of opinions, but of popular
reactions to and opinions about the judiciary. In this area, too, there is a long tradition of
136
empirical work lacking a unifying theoretical paradigm, which could benefit from the

131
This is similar to, but not quite the same as, the question of the descriptive accuracy of the positivist account of
law. See, e.g., Bix, supra note 13, at 21 (noting that debate between Dworkin and Raz should be ultimately
empirically resolvable); Benjamin C. Zipursky, The Model of Social Facts, in THE AUTONOMY OF LAW, supra note 4, at
219, 243.

132
See, e.g., James G. Brudney & Corey Ditslear, Canons of Construction and the Elusive Quest for Neutral
Reasoning, 58 VAND. L. REV. 1 (2005); Frank B. Cross, The Significance of Statutory Interpretive Methodologies, 82
NOTRE DAME L. REV. 1971 (2007); Jason J. Czarnezki & William K. Ford, The Phantom Philosophy? An Empirical
Investigation of Legal Interpretation, 65 MD. L. REV. 841 (2006); John B. Gates & Glenn A. Phelps, Intentionalism in
Constitutional Opinions, 49:2 POL. RESEARCH Q. 245 (June 1996); Robert M. Howard & Jeffrey A. Segal, An Original
Look at Originalism, 36 LAW & SOC’Y REV. 113 (2002); Jane S. Schacter, The Confounding Common Law Originalism
in Recent Supreme Court Statutory Intepretation: Implications for the Legislative History Debate and Beyond, 51
STAN. L. REV. 1 (1998); Nicholas S. Zeppos, The Use of Authority in Statutory Interpretation: An Empirical Analysis,
70 TEX. L. REV. 1073 (1992).

133
Compare the assumptions of DENNIS PATTERSON, LAW AND TRUTH (2000), with EINER ELHAUGE, STATUTORY DEFAULT
RULES: HOW TO INTERPRET UNCLEAR LEGISLATION (2008).

134
See, e.g., Moore, supra note 15, at 321–24; Philip Roberts, Observations on Method in Legal Theory and
Linguistics, in POSITIVISM TODAY, supra note 4, at 77, 81–92.

135
Cf. Brian Leiter’s call for a “naturalized jurisprudence” in Rethinking Legal Realism: Toward a Naturalized
Jurisprudence, 76 TEX. L. REV. 266 (1997).

136
See, e.g., Gregory A. Caldeira, Neither the Purse Nor the Sword: Dynamics of Public Confidence in the Supreme
Court, 80 AM. POL. SCI. REV. 1209 (1986); Gregory A. Caldeira & James L. Gibson, The Etiology of Public Support for
conceptual sophistication brought to bear on questions of authority and legitimacy by legal
theorists, and particularly by legal positivists. 137

II. Law, Theory, and Exclusion

Justice Scalia’s opinion‐writing rhetoric often criticizes elitism and touts the greater
populism and democratic value of the putatively positivist approach to justification that he
endorses. 138 As I have described legal positivism in this paper, however, it is in a number
of ways not an inclusive but an exclusionary discourse. For one thing, it is a theoretical
discourse describing its object (correctly, but partially) as effectuating the exclusion of
certain considerations. It is also a theoretical discourse that itself effectuates a parallel
exclusion (of, for example, assumptions contrary to its premises). And those able to
understand and contribute to the discourse are specialists, privy to vocabulary, knowledge,
and know‐how inaccessible to others. Lawyers become lawyers on the basis of their grasp
of what is properly excluded from legal analysis, argument, and reasoning—what count as
139
reasons in legal discourse and what do not—and their ability to use that knowledge.
Experts in a scholarly field, including academics who produce theoretical discourse, are
similarly defined as such by their knowledge of how to put forth a scholarly assertion—
which reasons to invoke, which not to invoke, and where to push the envelope. 140 Thus,
on the legal positivist account, law is an exclusionary discourse. And as a matter of social
fact, Western legal systems and academic institutions require expertise and are in this
sense exclusionary institutions.

the Supreme Court, 35 AM. J. POL. SCI. 635 (1992); Gregory Casey, Popular Perceptions of Supreme Court Rulings, 4
AM. POLITICS Q. 3 (1976); Robert H. Durr, Andrew D. Martin, & Christina Wohlbrecht, Ideological Divergence and
Public Support for the Supreme Court, 44 AM. J. POL. SCI. 768 (2000); Valerie J. Hoekstra, The Supreme Court and
Local Public Opinion, 94 AM. POL. SCI. REV. 89 (2000); Dean Jaros & Robert Roper, The U.S. Supreme Court: Myth,
Diffuse Support, Specific Support, and Legitimacy, 23 AM. POLITICS Q. 85 (1980); Timothy R. Johnson & Andrew D.
Martin, The Public’s Conditional Response to Supreme Court Decisions, 92 AM. POL. SCI. REV. 299 (1998); John
Kessel, Public Perceptions of the Supreme Court, 10 MIDWEST J. POL. SCI. 167 (1966); John M. Scheb II & William
Lyons, Judicial Behavior and Public Opinion: Popular Expectations Regarding the Factors that Influence Supreme
Court Decisions, 23 POL. BEHAVIOR 181 (2001).

137
See, e.g., JOSEPH RAZ, THE AUTHORITY OF LAW: ESSAYS ON LAW AND MORALITY (1983).

138
See, e.g., David A. Strauss, On the Origin of Rules (with Apologies to Darwin): A Comment on Antonin Scalia’s
The Rule of Law as a Law of Rules, 75 U. CHI. L. REV. 997, 998–99, 1002 (2008); Zlotnick, supra note 117, at 1382–
87.

139
This is a broader articulation of the point underlying Hart’s practice conception of the rule of recognition. The
issue is explored from different perspectives in, for example, Adler, supra note 10; Lewis, supra note 16; Andrei
Marmor, Legal Conventionalism, in HART’S POSTSCRIPT, supra note 4, at 193, 209–10.

140
See, e.g., BECHER, supra note 39, at 38, 44; GLOCK, supra note 71, at 205, 223–24; Kemp, supra note 65 (noting
that the evaluation of interdisciplinary work is hampered by use of the same terms in different disciplines in
different senses, incomprehensible to non‐specialists).
Justice Scalia’s linking of his methodological statements to the advancement of populism
and predictability 141 is thus an oversimplification, perhaps even a misrepresentation, in
that it attributes to his legal pronouncements an accessibility they do not truly have. But a
denial of the exclusionary tendencies of law is not necessarily entailed by a positivist
perspective. Among others, Leslie Green and Jeremy Waldron, while critical of positivism,
commended positivists such as Hart for noting how the social practice we call law has
grown increasingly esoteric and exclusionary. 142 Indeed, from the perspective advanced
here, it seems likely that legal practice and the generation of academic theory are high‐
status activities in part because of their exclusivity, as well as their cognitive portability. 143
If non‐experts in both areas did not exist, neither lawyers nor theorists would be able to
make a living.

Justice Scalia’s appeal to populist values is powerful because it is consistent with the
premise, shared by moral philosophers and laypeople alike, that it is desirable for all
members of a society to be on mostly equal footing regarding their capacities and
opportunities for communication and practical reasoning. If this premise is granted, then
the esoteric nature of law and theoretical inquiry appears pernicious—even in the
abstract, regardless of their manifestation in any particular legal systems or theoretical

discourse. Is there a way out of this dynamic? Is there anything a lawyer or theorist can do
to counter the esotericism of their respective enterprises?

This is a different question from the familiar one concerning so‐called normative or ethical
positivism. Normative positivism in its classic form seeks to describe what would be
necessary for a minimally pernicious legal system, but does not seek to eradicate its
exclusionary nature, which is accepted by the positivist as a constituent feature of law. 144
The question asked here is whether, taking this exclusionary nature as a given, the expert
can counter its negative implications. This approach to the question suggests that the
most direct response would be for experts to commit themselves to practices disregarding

141
See supra notes 116, 138 and accompanying text.

142
See Leslie Green, The Concept of Law Revisited, 94 MICH. L. REV. 1687, 1698–700 (1995) (review of HART, supra
note 27) (noting that Hart did not exalt secondary rules or formal legal systems as a better form of law, but only
described them as functionally necessary to sustain certain forms of social life); Green, supra note 19, at 1056–58
(noting that Hart described law as, by nature, prone to decay and vice, chiefly the vice of the alienation of law
from its subjects); Waldron, supra note 13, at 175, 179, 181 (noting that Hart’s account of law implied that the
emergence of a legal system allows the more efficient perpetuation of injustice, and that the development and
elaboration of secondary rules tends to make the population increasingly less familiar with primary rules). In The
Concept of Law, Hart acknowledged that the general public usually does not have access to or subscribe to the
rule of recognition. HART, supra note 27, at 59–60, 110–11.

143
Cf. Lewis, supra note 16, at 66. See also supra notes 69–73, 127–128 and accompanying text.

144
See, e.g., MacCormick, supra note 13; Liam Murphy, The Political Question of The Concept of Law, in HART’S
POSTSCRIPT, supra note 4, at 371; Perry, supra note 13; Waldron, supra note 9.
the boundaries of their expertise—specifically, to help non‐experts seek and obtain
interactional expertise in legal and theoretical discourse.

This kind of activity is already institutionalized in the systems in which both lawyers and
academics participate. But the teaching faces of law and theory are low‐status compared
to conversations among experts. 145 To combat the esotericism of their respective
domains, experts must consciously act contrary to the incentives provided by the status
hierarchies in those domains. While counter to self‐interest, this is not necessarily a self‐
defeating activity, nor is it incompatible with continued engagement in the elaboration of
theoretical and legal discourse. Continuing to examine and describe the shape and
conditions of theoretical discourse, in both law and philosophy, will allow us better to
understand the mechanisms by which the practices about which we theorize suspend
consideration of the non‐legal. This, in turn, will better allow us to understand both the
temptations of the discourse and those parts of it that non‐experts need to know in order
to participate in the conversation.

E. Conclusion

Neil MacCormick’s positing of a post‐positivist legal theory is quixotic but unnecessary.
The conditions of production of legal and theoretical discourse make it inevitable that
something resembling what we now call legal positivism will always be a part of legal
theory, whether we choose to use that term for it or not. Those conditions frustrate
aspirations to have the last word in legal theory, but they also ensure the continued
opening of new inquiries with potentially meaningful practical implications.

145
See, e.g., ABBOTT, SYSTEM OF PROFESSIONS, supra note 39, at 52–57, 79–84; COWNIE, supra note 72, at 58–69;
Deborah Jones Merritt, Research and Teaching on Law Faculties: An Empirical Exploration, 73 CHI.‐KENT L. REV. 765
(1998).
Special Issue
Comment on Petroski—On MacCormick’s Post‐Positivism

By Thomas Bustamante *

A. Introduction

In her thought‐provoking paper Is Post‐Positivism Possible?, Karen Petroski argues that
there are certain institutional conditions in the “modern setting of scholarly activity” that
make positivism the inevitable (or nearly inevitable) form of theoretical thinking about the
law. Furthermore, she also claims that there is nothing in Neil MacCormick’s post‐
positivism that should lead us to believe that his legal theory is qualitatively different from
positivism, which according to Petroski constitutes the mainstream Anglo‐American legal
philosophy. Her argument is that there is a parallel between the “characteristics of law”
1
and the “characteristics of the theoretical discourse about law.” Law, as the object of
legal theory, shares with legal‐theoretical inquiry (or, more broadly, with most academic
discourses in the modern society) some institutional conditions that make legal positivism
not merely a “defensible” mode of theorizing about the law, but an irresistible or
inescapable one. 2

Even though there are other interesting topics in Petroski’s paper, I will concentrate on
discussing these two claims. Although I am well impressed by some of the arguments in
Petroski’s paper, I cannot share her thesis that legal theory is constrained by some sort of
institutional context, which rules out the possibility of post‐positivism (Petroski’s first
thesis), and I also doubt her claim that Neil MacCormick fails to provide a real alternative
to the current versions of legal positivism (Petroski’s second thesis). In the following
sections I argue that the first thesis derives both from too strict an understanding of the
roles of philosophy and legal scholarship and from a questionable description of the
necessary features of law. Moreover, I attempt to highlight some of the features of
MacCormick’s latest legal theory that demonstrate that he moves beyond the limited
horizons of positivism.

*
Adjunct Professor, Federal University of Minas Gerais, Brazil. E‐mail: tbustamante@ufmg.br. The author would
like to thank Mátyás Bódig, Tamás Győrfi, Scott Styles, Zenon Bańkowski, András Jakab, Karen Petroski, and
Antonio Cavalcanti Maia for their valuable comments on previous drafts of this paper.
1
Karen Petroski, Is Post‐positivism Possible?, 12 GERM. L.J. 670 (2011).
2
Id. at 673.
My analysis is structured in the following way. In the first part of the paper, divided in two
sections, I summarize Petroski’s main arguments for the alleged predominance of
positivism in the Anglo‐American context and for the claim that post‐positivism is not a
real alternative to the contemporary efforts in the field of legal theory. In section B.I, I try
to determine the best interpretation of Petroski’s argument to explain the connection
between the features of legal theory and the features of legal practice, which according to
her explanation make positivism the most attractive type of legal theory. In section B.II, I
criticize this view and attempt to elucidate the nature of the arguments that we can find in
legal philosophy, as well as the relationship between such arguments and the practice of
law. In Part C, in turn, I focus on the specific criticisms that Petroski addresses to Neil
MacCormick’s post‐positivism. The first subsection re‐states Petroski’s points against the
late version of MacCormick’s institutional theory of law, and the remaining of the paper is
dedicated to showing what answers in MacCormick’s legal theory we could find to
Petroski’s criticisms.

B. Legal Scholarship and Legal Discourse

I. Petroski’s Argument for the Predominance of Positivism

Petroski starts her argument for the first thesis by referring to an emerging literature
which argues that, since early modern days, “the academic domain has been more and
more characterized by the policing of membership through extensive discursive
3
indoctrination and the evaluation of various forms of discursive performance.” She claims
that the modern university has some distinctive material and formal features that impose a
set of constraints in the forms of knowledge produced within its boundaries. Those acting
within the institution of the modern university are depicted as depending on some
external support which can only be achieved by the recognition of a special expertise that
justifies their activities to social actors outside the institutions. At the same time, the
institutional structure is described as “largely self‐reproducing,” and such self‐reproduction
4
occurs primarily through a special type of discourse. These features “make the structure
a competitive arena, and the resulting competition encourages academics to draw
distinctions between their positions and those of their forebears and contemporaries.” 5
The typical strategy of academics in the social sciences, and in law in particular, would be
to make “fractal distinctions” which encourage rhetorical innovation “even when a theorist
is not truly making a novel point.” 6

3
Id. at 677.
4
Id. at 678.
5
Id. at 678.
6
Id. at 678.
2011]
Comment on Petroski 695
In a similar way, the practical legal discourse is marked by the same type of indoctrination.
“Like the academy,” the argument goes, “the set of legal institutions is both materially and
conceptually self‐perpetuating.” Materially, “lawyers’ expertise in the discourse used to
navigate the institutions regulating state powers ensures their continued support by non‐
lawyers.” 7 On the formal level, in turn, lawyers adopt a theoretical perspective which
helps them to exclude certain reasons for action from the set of materials available in legal
reasoning. As in the academic domain, this self‐referential character of law takes a fractal
form and creates a status hierarchy for the experts that master the nuances of legal
language.

In my reading Petroski argues that both the academic discourse and the modern legal
domain are self‐referential; dependant on some sort of indoctrination and fractal
distinctions; in need of mastering a special discourse which justify their activities to actors
outside their realm; exclusive or certain reasons; and self‐perpetuating. But why are these
common features relevant for the debate over post‐positivism? How do these parallel
features of each of these intellectual domains influence the other?

If we accept that there is more than a simple historical coincidence between the character
of legal theoretical inquiry and the features of legal discourse, then there are two
alternative interpretations available for this parallel between legal practice and academic
scholarship.

The first interpretation explains the dominance of positivism on the basis of the features
that Petroski attributes to the legal practice. At first sight, the author appears to be
following this route in the beginning of the second section of her paper. Nevertheless, if
we look closely enough we can see that any attempt to explain the features of legal
theory’s academic discourse on the basis of the features that Petroski attributes to the
legal practice will be very fragile. This line of argument presupposes from the outset the
correction of positivism, which is precisely what it intends to demonstrate. The argument
would be circular, since the description of the legal domain as necessarily self‐referential
and exclusive of any non‐legal reasons already coincides with the key assumptions of legal
positivism. If we define positivism as the theory according to which the law is “self‐
referential” (in the sense that it regulates the process of its own creation and sets a test for
clearly distinguishing between legal and non‐legal rules) and “exclusionary of certain
reasons” (whether this feature is interpreted in a strict sense according to which the
sources of law are to be found in social facts alone or in a broader sense that would
exclude only the set of reasons which are not incorporated by the rule of recognition), then
it becomes obvious that this interpretation of the relationship between law and legal
scholarship simply begs the question, for the features that are attributed to law or “legal
practice” coincide with some of the theoretical theses about the law held by most
7
Id. at 679.
contemporary positivists. To follow this strategy would be saying something like “legal
theory defines the law as self‐referential and exclusionary because the law is self‐
referential and exclusionary,” which seems like a tautology. It is questionable, thus, to
claim that legal theorists tend to be positivists because of any empirical feature of the legal
systems.

Nevertheless, to refute this first interpretation is not enough to dismantle Petroski’s point.
One can see from her text that Petroski is aware of this problem when she says that she is
not actually advocating that legal theory is logically constrained to be positivist due to the
empirical features she attributes to legal discourse:

My point is not that legal theory inevitably adopts
these features because it is impelled to take on the
features of the subject‐matter it is explaining (either
because this is necessary for the complete explanation
or because it is a kind of irresistible temptation). It is,
rather, that because of the parallel social functions and
historical paths of these two practices—legal practice
and theoretical inquiry—they in fact share certain
features, at least when they are regarded from a
8
certain perspective.

This quotation shows that the best way to read the thesis of the parallel between legal
practice and academic discourse is by accepting not that the features of law influence the
theoretical understanding of legal theorists, but rather that the features of academic
discourse are projected in the way scholars describe the law. There would still be some
sort of “irresistible temptation” for positivism, but its cause would not be any feature of
the object of inquiry of legal theorists (the law), but rather a set of features of the inquiry
itself. This is the second interpretation of the connection between the character of
practical legal discourse and that of legal‐theoretical inquiry. I can find support for this
second reading in Petroski’s own words, when she holds that “the institutional and
conceptual dynamics within each domain provide strong incentives for those working in
9
each to fixate on just these features.” The same phenomenon would thus occur in both
institutional contexts: On the one hand, lawyers tend to interpret the law in a legalistic
way because legal reasoning is embedded in an exclusivist discursive pattern; on the other
hand, theorists tend to describe the nature of law in a positivistic way because of some
institutional restraints over the sphere in which they operate. Does this second
interpretation of the parallel connections between legal discourse and legal practice make
Petroski’s thesis more acceptable than the first one?
8
Id. at 680.
9
Id. at 680 (emphasis added).
2011]

An argument for the second interpretation could be plausible if positivism were the only or
the overwhelmingly dominant way of thinking about the nature of law. Nonetheless, the
impact of such “institutional and conceptual dynamics” in contemporary legal theory is not
that evident. No natural lawyer, pragmatist or non‐positivist in general would define the
law as entirely self‐referential and exclusionary of all non‐legal reasons in a strict sense. 10
No compelling institutional factor in the structure of the modern universities seems to be
requiring such theorists to do so. Furthermore, I believe that Petroski’s ideas about the
social functions of law and legal theory are unconvincing, for at least two correlated
reasons. On the one hand, the only social function one can identify for legal practice in her
paradigm is to exercise an aristocratic domination over the mass of non‐lawyers by means
of an obscure discourse which is closed within itself. This seems to be more a pathology
than a genuine social function. On the other hand, if this description of legal theory were
correct, then the theoretical inquiry about law would suffer from self‐imposed limitations
which would keep it away from any interesting real‐life problem. Legal theory would not
be able to achieve what it purports to do, since it would neither be able to explain the
normativity of law nor contribute to justify relevant decisions about the nature of law. As
we will see in the next section, legal philosophy itself would lose one of its central
functions.

II. The Critical Dimension of Legal Theory

Although Petroski intends to explain positivism in sociological terms, I believe that the
debate between positivists and non‐positivists is a philosophical one, and that it cannot be
resolved without a proper understanding of the nature of legal philosophy. We can see
from Petroski’s description of the character of legal‐theoretical discourses that she sees
legal theorists as (at least in part) strategic actors who operate within a competitive
environment that is closed within itself and excludes all the reasons that fall outside of the
scope of its own discursive pattern. Instead of acting critically against this framework, legal
theorists cultivate it and reproduce it because they can only achieve status, authority and
expertise by sticking to such discourse and differentiating it from the competing discourses
by means of fractal distinctions. To move out of this context would be against the interests
of the theorist herself.

10
A natural lawyer would have to say that the normativity of law is justified by some objective moral principles
which cannot be entirely excluded from practical legal reasoning. See generally JOHN FINNIS, NATURAL LAW AND
NATURAL RIGHTS (1979). A pragmatist would never deny that reasons of policy and arguments from other non‐
strictly institutionalized materials play an important part in legal argumentation, although she would tend to be
skeptical about the role of moral principles in legal practice. See generally RICHARD POSNER, THE PROBLEMS OF
JURISPRUDENCE (1993). A non‐positivist like Dworkin or Alexy, finally, would argue that the key principles of
political morality are in any case part of the materials that lawyers use in their practical activity. See generally
ROBERT ALEXY, THE ARGUMENT FROM INJUSTICE: A REPLY TO LEGAL POSITIVISM (2002); RONALD DWORKIN, LAW’S EMPIRE
(1986).
If Petroski’s analysis of law and of theoretical thinking is correct, both the legal system and
legal philosophy could be described in the same way that Jürgen Habermas described
Niklas Luhmann’s sociology of law.

As so described, the legal system is a recursively closed
circuit of communication that self‐referentially delimits
itself from its environment, with which it has contact
only through observations. . . . [T]he system describes
its own components in legal categories and employs
these self‐thematizations for the purposes of
constituting and reproducing legal acts by its own
11
means.

The problem of Luhmann’s 12 autopoietic theory of law, Habermas argues, is that its
empiricist interpretation of the legal system assumes that it is detached from all internal
relations to morality and politics. “The law is reduced to the special function of the
administration of law,” and the interpreter or the theorist loses sight of the connection
between law and the “constitutional organization of the origin, acquisition, and use of
political power.” 13 If the law is understood as an autopoietic social system, one cannot
account for its social normativity and its proper social functions in the context of a
pluralistic and secular society which can no longer rely on metaphysical arguments from
natural law. Striped of its normative connotations that once were backed by metaphysical
natural law, but which can no longer find any room when the law is observed from the
outside and characterized as an entirely autonomous social system, the law becomes
“narcissistically marginalized” and can only react to its own problems. In short, a mutual
14
indifference between the law and all other social systems is assumed by the interpreter.
This mutual indifference, Habermas would say, has the consequence that the law can no
longer find a proper justification, for its normativity is heavily compromised.

Whoever is right in this point, Habermas or Luhmann, it must be stressed that this debate
is more fruitful as a debate about the “law” than as one about the character of theoretical
inquiry over its nature. In fact, one can make very good arguments to support, against
Habermas, the claim that the law is a closed and self‐referential system which can be
identified by means of a strictly empirical analysis, without any consideration about the
role played by politics or morality in its creation and administration. This is, in fact, how
11
JÜRGEN HABERMAS, BETWEEN FACTS AND NORMS—CONTRIBUTIONS TO A DISCOURSE THEORY OF LAW AND DEMOCRACY 49
(1996).
12
See generally NIKLAS LUHMANN, LAW AS A SOCIAL SYSTEM (2008).
13
HABERMAS, supra note 11, at 50.
14
Id. at 51.
2011]
one of the most powerful legal theories of our time explains the validity of a law.
According to Joseph Raz, the valid law should necessarily be identified in a way that is free
from any political or moral connotations, since its sources come from social facts alone. To
include moral principles or any type of non‐authoritative considerations in the tests used
to distinguish legal rules from other social norms would undermine the authoritative
character of legal norms, which is its distinctive feature. When legal officials are set out to
apply the law, they are guided by “positive authoritative considerations” whose existence
15
and content can be asserted without resort to moral judgment. Law, in this theory,
consists only of such positive authoritative considerations. When state officials identify
and apply the law, their personal action is situated not on the deliberative stage of
practical reasoning, but rather on the executive stage, in which the question what ought to
be done is answered without resort to any type of moral or political argument. 16 As in
Luhmann, we can notice in Raz a stress on an intrinsic form of rationality in legal reasoning,
which is conducted in an entirely autonomous practical context. If that is the case, one
might ask whether this autonomy applies also to the theoretical inquiry about the nature
of law, as Petroski suggests. Can we conceive of legal theory as strictly separated from
other fields of inquiry such as politics and morality? Do legal theorists need to confine
themselves to any particular type of reason?

My view is that both of these questions should be answered in the negative. Even within
the realm of legal positivism, few are the theorists who unequivocally agree with Petroski
on this point. Austin, Kelsen, and Hart are possibly the best examples of this approach, but
they seem to be in a minority (even though they are prominent thinkers within the
positivist tradition), and I do not think that they are right in this. They seem to believe that
theoretical inquiry should be entirely neutral. No moral, ethical or pragmatic argument
should be used to back up a particular theoretical conception of law. In Kelsen’s
perspective, for instance, science is the only form of knowledge worth pursuing, and it
must proceed either by means of empirical observation or by strict logical and
mathematical analysis. The theorist is motivated solely by her faith in science and the
pursuit of the truth. She has no ideological ambitions and is determined to free her
analysis from the prejudices and preconceptions that are usually found in ethics, morality,
politics, psychology, biology, and religion. As Kelsen writes in the opening page of the Pure
Theory of Law, legal theory is a “theory of positive law,” and as such it aims “solely at the
cognition of its subject matter”; that is to say: Legal theory is “legal science, not legal
17
policy.” By the same token, Austin’s imperative theory of law is strongly committed to
an empiricist approach, as is visible in one of his most famous passages, where he argues
15
JOSEPH RAZ, ETHICS IN THE PUBLIC DOMAIN: ESSAYS IN MORALITY OF LAW AND POLITICS 189 (1994).
16
Id. at 190.
17
HANS KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY 7 (1992).
that “the matter of jurisprudence is positive law, simply and strictly so called.” 18 Finally,
Hart’s view of jurisprudence as a form of “descriptive sociology” 19 is even more
characteristic of this neutral methodological point of view.

In my view, Petroski seems to believe that it is feasible to analyze the legal phenomenon
with the kind of neutrality that is advocated by the strand of descriptive positivism
mentioned just above. If that were the case, it would appear reasonable to argue that
legal theory is bound to be positivist because of the methodological attitude of the major
legal theorists, who by large exclude political, pragmatic, and moral reasons from their
theoretical inquiry about the nature of law. This is, in fact, the root of my deepest
disagreement with Petroski. I am not convinced that any serious theoretical account of the
legal phenomenon can free itself from the kind of reasons that are found in politics, ethics
or morality. As a matter of fact, I can think of no Anglo‐American positivist whose
arguments in support of his or her own theory are neither moral nor political in nature. In
these matters, I think that Dworkin is right when he claims that “any theory of law,
20
including positivism, is based in the end on some particular normative political theory.”
Law is a contested political concept which takes its sense from its use, “from the contexts
of debates about what the law is, and from what turns on which view is accepted.” 21
According to Dworkin, positivists who claim that their theories are merely a sort of
linguistic study or descriptive sociology potentially do not understand their own
arguments. 22 Every thesis about the nature of law is partly grounded in an argument of
political morality.

If one looks at the five authors that Petroski quotes in support of her argument that legal
theory is bound to remain positivist because of the key features of jurisprudential inquiry,
one should be able to see that none of them excluded moral and political reasons in
support of his theory. In my opinion, neither Hobbes nor Bentham, Austin, Hart, or Raz has
produced a legal theory with all the features pointed out by Petrosky. None of these
theories is self‐referential or exclusive of all moral and political considerations.

Hobbes, for instance, is as much a natural lawyer as he is one of the founders of positivism.
The “laws of nature” are descried in the Leviathan as “immutable and eternal,” and the
23
science of these laws is “the true and only moral philosophy.” Although he is skeptical
18
JOHN AUSTIN, THE PROVINCE OF JURISPRUDENCE DETERMINED 1 (1832).
19
H. L. A. HART, THE CONCEPT OF LAW (2d ed. 1994).
20
Ronald Dworkin, A Reply, in RONALD DWORKIN AND CONTEMPORARY JURISPRUDENCE 254 (Marshall Cohen ed., 1984).
21
Id. at 256.
22
Id. at 255.
23
THOMAS HOBBES, LEVIATHAN para., I. 15 (1651).
2011]
about any natural inclination of mankind towards these laws, for in the “state of nature”
the “laws of nature” cannot determine human action, he shares with classical natural
lawyers the view that people are able to discover what these laws of nature are by using
their practical reason. When humans do that, Hobbes argues, they are likely to form a
commonwealth in which men are no longer apt to act on the so‐called “right of nature” or
“the liberty each man hath to use his own power as he will himself for the preservation of
his own nature . . . or his own life.” 24 These views are more than enough to classify
Hobbes as a natural lawyer, for he admits both that there is a natural law the existence of
which does not depend on any positive or institutional act and that the positive laws
created by the civil society find their justification in such natural laws. Hence, Hobbes can
only be read as a positivist if one understands this label as not necessarily antagonist to
what natural lawyers have to say. Moreover, Petroski’s belief that Hobbes was under
some sort of constraint because of the “basic conceptual vocabulary” that he uses does
not seem to me entirely accurate from a historical point of view, since it misses the point
about Hobbes’ political motivations for the kind of theory he advocated. Hobbes lived
through one of the most volatile moments in British history, and he wrote with a view to
provide a practical and political justification for absolute Monarchy, which was in his
opinion the only way out of the English Civil War. Hence, although Petroski is right when
she argues that Hobbes and other positivists described the legal system as one which
“suspends the operation of certain reasons for action and that has self‐referential
25
features,” the same cannot be said about the inquiry or the research that these theorists
undertook. Despite the initial appearances, the actual inquiry of Hobbes and his
successors is neither self‐referential nor exclusive.

To illustrate that point, one can think of the case of Bentham. Even though some scholars
as important as Hart interpret his theory as a predominantly descriptive or neutral
theoretical enterprise, 26 the best reading of his ideas is the one that recognizes that the
role of his “analytical jurisprudence” is subservient to what he called the “art of
legislation.” 27 Bentham’s positivism has moral roots. He defines a law in a positivistic way
because he believes that in doing so he is following the principle of utility. As a form of
utilitarian positivism, Bentham’s conceptual theory is based on a particular view about the
proper function of law, a function which cannot be understood in purely legal terms. 28

24
Id., para., I. 14.
25
Petroski, supra note 1, at 682.
26
H. L. A. HART, ESSAYS ON BENTHAM: JURISPRUDENCE AND POLITICAL PHILOSOPHY 21–39 (1982).
27
Philip Schofield, Jeremy Bentham and nineteenth‐century English jurisprudence 12 J. LEGAL HIST. 58, 60 (1991).
28
The best defense of this interpretation of Bentham can be found in GERALD POSTEMA, BENTHAM AND THE COMMON
LAW TRADITION (1986).
Even Austin, who concentrated his legal theory on the so‐called “expository
jurisprudence,” as opposed to the “censorial jurisprudence,” may have hidden moral and
political reasons for his allegedly descriptive positivism. Bentham’s reformative
utilitarianism can be contrasted with Austin’s conservative use of the same principle.
Whilst Bentham sees the principle of utility as a device for criticizing government and
changing the law into something better, Austin adopted a “theological utilitarianism”
which presumed that “it was the will of God that his creatures should be happy, and
29
therefore the theory of utility and the will of God . . . coincided.” Unlike Bentham, Austin
seemed to assume that the principle of utility had in practice commonly guided the
legislator. As Schofield notices, “whereas Bentham’s ‘scientific’ version of utilitarianism
subjected existing practices and institutions to the scrutiny of the principle of utility . . . ,
Austin’s theological utilitarianism tended to see those same practices and institutions as
[already] embodying utility.” 30 We can see, thus, that despite Austin’s claim that his
theory was neutral and purely descriptive, his jurisprudence could not be separated from
the conservative moral‐political context in which it was initially proposed.

Hart, likewise, regardless of his firm conviction that his “account is descriptive in that it is
morally neutral and has no justificatory aims,” 31 has not managed to free his theoretical
inquiry from arguments of political morality. I can think of two of his most central
arguments as genuine examples of the moral‐political commitments of his theory, which
will be analyzed in the following paragraphs.

The first argument which exemplifies Hart’s moral‐political commitments appears in his
reply to Radbruch’s post‐war papers against positivism. In one of his most celebrated
essays, Hart heavily criticizes Radbruch and the German Constitutional Court for the
decisions that applied the so‐called “Radbruch’s Formula” and thus denied legal character
to a set of Nazi Laws which imposed racist measures on people of the Jewish religion. In
particular, Hart was not satisfied with the reasoning provided by the Constitutional Court
to justify, in a set of criminal cases, the conclusion that some statutes are too unjust to
deserve any form of obedience. Instead of saying that the laws which legalized murder
against the Jews lacked legal validity because of their extreme injustice, Hart argues, the
court should have admitted that these statutes had indeed legal character, although the
law in that case was too wicked to be obeyed. In order to correctly justify its decisions, the
court should have recognized the legal character of the old statutes while creating a new
legal rule with retrospective effects. In Hart’s own words:

29
Schofield, supra note 27, at 63.
30
Id.
31
HART, THE CONCEPT OF LAW, supra note 19, at 240 (emphasis in original).
2011]
Odious as retrospective criminal legislation and

punishment may be, to have pursued it openly in this
case would at least have the merits of candour. It
would have made plain that in punishing the woman a
choice had to be made between two evils, that of
leaving her unpunished and that of sacrificing a very
precious principle of morality endorsed by most legal
systems. Surely if we have learned anything from the
history of morals it is that the thing to do with a moral
32
quandary is not to hide it.

It is clear, therefore, that one of Hart’s most famous arguments in reply to non‐positivism
is a genuinely moral argument, which has very little neutrality in it. A positivist definition
of law should be preferred to a non‐positivist one because it would make plain the
sacrifices and the choices that one has to make in order to impose legal obligations with
retrospective effect.

The second argument, on the other hand, is even more expressive of the moral or political
preferences of the author, for it clearly demonstrates that the allegedly descriptive
theoretical account proposed in The Concept of Law has important normative elements
built into it. Hart’s central argument in support of the idea that a legal system must have
not only “primary rules” according to which humans are required to do or abstain from
certain actions, but also “secondary rules” which are concerned with the primary rules
themselves, is grounded on a reductio ad absurdum that exposes the inadequacies of a
political organization which uses only the former type of rules. An imaginary society where
there are only primary rules of obligation would suffer from the problems of uncertainty,
for there would be no procedure for settling any doubt about the validity of a rule; of
having a static character, for there would be no means, in such society, for deliberately
adapting these rules to novel circumstances; and of inefficiency, since there would be no
procedure to keep the diffused social pressure by which rules are enforced and
33
maintained. It is because of this that we need secondary rules to provide a remedy for
these three serious problems for any legal society. Nevertheless, as Dworkin correctly
argues, such construction is far from being neutral or purely descriptive, as Hart claims it to
be. Dworkin’s words about this point are particularly illuminating:

He [Hart] develops his own account of the main
elements of law by showing how the device of a
secondary rule of recognition responds to these
32
H. L. A. Hart, Positivism and the Separation of Law and Morals, in THE PHILOSOPHY OF LAW 33 (Ronald Dworkin ed.,
1977).
33
HART, supra note 19, at 92–93.
particular defects by making possible a new set of rules
that are flexible, efficient, and certain. This, I believe,
does support my suggestion about the political basis of
positivism. 34

If Dworkin is right about this, then Hart’s advocacy of neutrality in jurisprudence is
inconsistent with his own theory, for the choice of his theoretical position is entirely
determined by political considerations. As it happens with Hobbes, Bentham and Austin,
Hart’s theoretical inquiry is far from excluding all moral and political reasons.

Last but not least, Raz’s philosophy of law appears to me as anything but neutral. First of
all, he has never attempted to avoid moral reasons in his theoretical project. Although he
holds that the law consists only of positive authoritative considerations, which implies that
one cannot appeal to moral or political arguments in order to answer questions about the
validity of a particular rule, he does not deny that there is some sort of connection
between law and morality and that the normativity of a legal system should be explained
by means of moral reasons. The Separability Thesis, for instance, is expressly denied by
Raz: “[I]t is very likely that there is some necessary connection between law and morality,
that every legal system in force has some moral merit or does some moral good even if it is
35
also the cause of a great deal of moral evil.” Perhaps one of the keys to understand Raz’s
theory that the law is a system of authoritative rules is to look at the example of
arbitration, which is one of his most famous. In that scenario, each of the parties has her
own first‐order reasons to determine her course of action, but these reasons may conflict.
When two persons have conflicting views about a particular decision, they may agree to
refer the dispute to an arbitrator. If they do that, two features stand out: (1) The
arbitrator’s decision is in itself a reason for action; and (2) The arbitrator’s decision must be
based on a set of first‐order reasons which apply to the disputants (or, in Raz’s
terminology, “dependent reasons”). In Raz’s perspective, the arbitrator has authority over
the disputants because his judgment over the reasons that apply to them replaces their
own balancing of reasons. “In agreeing to obey the decision [of the arbitrator], the
disputants agreed to follow his judgment of the balance of reasons rather than their
36
own.” This is what happens in the case of legal systems. When people join together to
form a legal community, they agree to accept an authority whose judgment replaces that
of the individual actors themselves. At this stage, we can see clearly the political root of
Raz’s philosophy of law. People accept the authority of law for prudential reasons. One of
the central ideas to explain the normativity of law is what Raz calls the “normal
justification thesis,” which can be stated in the following terms:
34
Dworkin, supra note 20, at 255.
35
RAZ, supra note 15, at 211.
36
Id. at 196.
2011]

[T]he normal way to establish that a person has
authority over another person involves showing that
the alleged subject is likely better to comply with
reasons which apply to him . . . if he accepts the
directives of the alleged authority as authoritatively
binding, and tries to follow them, rather than by trying
37
to follow the reasons which apply to him directly.

Law’s validity and bindingness can only be justified by means of prudential, political, or
moral reasons. The authority of law is justified when people are better‐off with the law’s
judgment rather than with their own. The whole argument, thus, is conducted by practical
reasons. Furthermore, one of Raz’s basic ideas is that the law claims legitimate authority,
and such legitimacy means that the authority of every legal agent must seek for a kind of
moral justification. To eliminate any doubts about the use of political and moral reasons in
Raz’s theory of legal authority, we can quote his own words about it. When talking about
his own writings, Raz has recently said: “I maintain that necessarily the law claims to have
38
legitimate authority, and that that claim is a moral claim.” We can see, thus, that Raz’s
philosophy of law also differs from Petroski’s characterization of legal theoretical inquiry as
an entirely neutral enterprise. This is enough to claim, contrary to what Petroski thinks,
that jurisprudence as a philosophical discipline neither is self‐referential nor excludes
moral and political considerations. Even in the case of those who think that the law is an
autonomous social system, theoretical inquiry over the law is hardly ever autonomous.
Jurisprudence is philosophy applied to the law, and it plays a similar role to that of
philosophy in general.

One of the most interesting questions about legal philosophy is to understand the kind of
statements we find in it. Although one can think of many different methodological
approaches to legal philosophy, which cannot be analyzed here, that which seems most
interesting to me is the one advocated by Robert Alexy. For him, “[l]egal philosophy is
argumentation about the nature of law,” and any reflection about “legal philosophy is,
39
therefore, a reflection about the nature of arguments about the nature of law.” Before
we understand the nature of such reflection, however, we need some clarification about
philosophy itself. How can we define philosophy? Again, this is one of the most difficult
tasks of philosophical discourse. Alexy’s answer is this: “Philosophy is general and
systematic reflection about what there is [ontology], what ought to be done or is good
37
Joseph RAZ, THE MORALITY OF FREEDOM 53 (1986) (emphasis removed).
38
JOSEPH RAZ, THE AUTHORITY OF LAW: ESSAYS ON LAW AND MORALITY 315 (2d ed. 2009).
39
Robert Alexy, The nature of the arguments about the nature of law, in RIGHTS, CULTURE AND THE LAW—THEMES
FROM THE LEGAL AND POLITICAL PHILOSOPHY OF JOSEPH RAZ 4 (L. H. Meyer et al. eds., 2003).
[ethics], and how knowledge about both is possible [epistemology].” 40 This definition,
according to Alexy, has three corollaries of capital importance: Firstly, the fact that
philosophy is a reflection presupposes a critical aspect that constitutes a normative
dimension for philosophical inquiry; “Philosophy as a necessarily reflective enterprise
therefore necessarily has a normative dimension.” 41 Secondly, the general and systematic
character of philosophical reflection leads to an analytical dimension (which “is defined by
the attempt to identify and to make explicit the fundamental structures of the natural and
social world in which we live and the fundamental concepts and principles by means of
which we can grasp both worlds)” 42 and, thirdly, to a synthetic dimension (which “is
defined by the attempt to unite all of this into a coherent whole”). 43

We are now in a position to define legal philosophy. Legal philosophical inquiry is a
philosophical inquiry about the nature of law. Its differentia specifica consists in its
subject. It is, thus, a general and systematic reflection about what there is, what ought to
be, and what can be known, but with specific reference to law.

Of course this concept of legal philosophy is different from that of some legal theorists
who might restrict its scope to what Alexy has called the analytical dimension.
Nevertheless, this is not how most contemporary philosophers (as opposed to legal
philosophers) understand their own discipline, and the fact that the concepts and theories
developed by legal philosophy are quite often used in practice, at least when it comes to
pivotal cases, is an indicator that the legal theoretical accounts that restrict the scope of
legal philosophy to analytical or empirical problems tend to misrepresent the function of
44
the theoretical inquiry about the law.

My main reasons for disagreement with Petroski are that if her description of legal theory
is correct, then jurisprudence lacks what I consider to be one of its crucial functions
40
Robert Alexy, The Nature of Legal Philosophy 17 RATIO JURIS 156, 157 (2004).
41
Id. at 158. The same argument is found in ALEXY, supra note 39, at 3.
42
Alexy, supra note 40, at 158.
43
Id.
44
This is precisely what Dworkin means, for instance, when he holds that in any legal dispute there is always the
possibility of a “theoretical disagreement” about the law. In hard cases, judges and lawyers may disagree not
only about whether or not a particular act falls within the scope of a master rule such as Hart’s rule of
recognition, but rather about the content of the master rule itself. The criterion on which one relies to identify
the law in a given case may depend on the theory one upholds to develop his conception of law. See, e.g.,
DWORKIN, supra note 10, at 4–5. To pick up an example, in addition to the cases quoted by Dworkin in the first
two chapters of his LAW’S EMPIRE, we can mention the decision of the House of Lords in Jackson v. Her Majesty’s
Attorney General [2005] UKHL 56. In this case, the Law Lords disagreed not only about the validity of a particular
rule, but about the rule of recognition itself. For the particulars of this case, see Michael Plaxton, The concept of
legislation: Jackson and Others v Her Majesty's Attorney‐General, 69 MOD. L. REV. 249, 249–61 (2006).
2011]
according to Alexy’s account, which is the normative or critical one. Whatever defects Neil
MacCormick’s legal theory may have, the lack of a critical or normative standpoint for
jurisprudential inquiry is not one of them. In my reading, his conception of law as an
institutional normative order includes the idea that “the proper purpose of such an order is
the realization of justice,” and one of the consequences of this conception is that it justifies
a “critical attitude” towards actual institutions of law and state. 45

C. MacCormick’s Post‐Positivism and the Nature of Legal Argumentation

I. Petroski’s Criticism on MacCormick’s Post‐Positivism

According to Petroski, Neil MacCormick is the only legal theorist to have applied to himself
the label “post‐positivist” or to have explained what this label actually means. She claims
that the use of the prefix “post” implies that “the root tradition is no longer viable in its
46
original form,” and that this is the sense in which MacCormick uses the term. After
saying that, she explains why she thinks that a legal theorist might conclude that the
tradition of legal positivism is no longer viable. She claims that there are two types of
reasons one can adduce for a post‐positivist position: conceptual and prudential reasons.
On the one hand, the former could include either (1) a “dissatisfaction with the
increasingly narrow questions addressed by positivists and their apparently increasingly
trivial conclusions” or (2) “a conclusion that one or more of the premises or methods of
legal positivism have been discredited or are irreconcilable.” The latter, on the other hand,
could include (3) “a desire to free the theorist’s work from the pejorative connotations
associated with” positivism; (4) “exhaustion with the volume of material produced under
the positivist rubric;” or (5) “a desire to mark the theorist’s work as original rather than
47
derivative.” With this classification in mind, she claims that MacCormick’s chief
motivations for his post‐positivist position can only be prudential in nature.

Petroski argues, firstly, that MacCormick “does not take the position that the premises of
legal positivism have been discredited,” 48 and attempts to show this by mentioning his
assertion that “law and morality are conceptually distinct.” 49 Secondly, she claims that
neither is MacCormick troubled by the “nit‐picking details of legal positivist controversies,
since he engages in and acknowledges the relevance of many of these quarrels.” 50 And
45
NEIL MACCORMICK, INSTITUTIONS OF LAW: AN ESSAY IN LEGAL THEORY 264 (2007).
46
47
Id. at 672‐3.
48
Id. at 674.
49
Id., at 674, note 47.
50
Id. at 674.
thirdly, she says that although MacCormick claims to be a post‐positivist because he
believes that the “law is necessarily geared to some conception of justice,” he “concedes
that it is only the ‘more austere and rigorous forms’ of legal positivism that ‘absolutely
exclude the possibility that there is any moral minimum that is necessary to the existence
of law as such.’” 51

The conclusions Petroski draws are that (1) “none of the details of MacCormick’s position
in Institutions of Law, his final summation of his theoretical commitments, distinguishes
those commitments from the core legal positivist commitments,” and (2) that the author
can only be using the “post‐positivist” label for some of the prudential purposes she
identifies, that is, as “a device for avoiding unwanted negative connotations or for
52
asserting that his intellectual position is distinguishable from others.”

In the following I am going to attempt to defend MacCormick against these criticisms. My
reasons are twofold. Firstly, my reading differs from Petroski’s when she suggests that
MacCormick does not hold that the premises of legal positivism are “discredited” or
“incorrect,” and when she states that the MacCormick’s belief that “law and morality are
conceptually distinct” implies any agreement with positivism. Some of the central theses
of positivism are that (1) there is no necessary connection between law and morality and
53
(2) it is possible that valid laws can be identified without resort to moral judgments. As a
post‐positivist, MacCormick accepts neither of these claims. This fact does not entail,
however, that law and morality should be confused or are not “conceptually distinct.”
Even an orthodox natural lawyer or an enthusiastic non‐positivist will have to concede that
law and morality are not conceptually equivalent, no matter how many necessary
connections she finds between these two normative orders. Secondly, MacCormick’s
assertion that “the law is necessarily geared to justice” seems to me to be very different
from the claim that “there is a moral minimum which is necessary for the existence of law
as such.” While the former implies that there is an argumentative connection between law
and morality, so that moral arguments are embedded in any legal discourse and
necessarily play a part in the interpretation and application of law, the latter means only
that a legal system which does not comply with some minimum moral standards cannot be
legally valid. Even if Petroski is right when she says that the former thesis can be accepted
by non‐orthodox positivists, this fact alone does not make Neil MacCormick one of them.
We can see, thus, that even though MacCormick retains some ideas derived from the
51
Id. at 675.
52
Id. at 676.
53
A contemporary positivist might reply that positivism no longer advocates the first thesis (Separability). This
kind of argument is part of a general attempt to narrow down positivism in order to escape from objections
raised by Dworkin and other post‐positivists such as MacCormick. Nonetheless, I believe this does not affect my
argument here, for it suffices to prove that MacCormick rejects the so‐called Sources Thesis in order to claim that
his theory is very different from positivism.
2011]
positivist tradition, he departs from the most distinctive theses of such tradition.
Differently from Petroski, I think that MacCormick’s theory genuinely transcends the frame
of thought which is common to legal positivists. In the following section, my job is to
provide a few examples to demonstrate this point.

II. Law, Morality and Argumentation: The Nature of MacCormick’s Post‐Positivism

Although MacCormick had not adopted for himself the label “post‐positivist” until late in
his career, his relationship with positivism has never been unproblematic. Even in his
earliest works he was not entirely satisfied with the mainstream positivist position about
the nature of law. In spite of the great influence of Herbert Hart, there were relevant
disagreements with the supposed neutrality of Hartian and Kelsenian positivisms. Perhaps
the most expressive of these disagreements concerns the issue of the justification of
decisions about the validity of a norm, considered from the internal point of view. As
MacCormick has stressed several times, legal theorists should press Hart’s insights about
54
the internal point of view further than he did in his own theoretical project. Once
theorists take into consideration all the implications of the internal point of view, the
problem of the justification of legal decisions becomes a central one, as we can see in the
following excerpt from MacCormick’s Legal Reasoning and Legal Theory:

[A] positivistic description of the system as it operates
cannot answer a particular kind of question which may
be raised internally to a legal system: the question as it
might be raised for a judge in a hard case: “Why ought
we to treat every decision in accordance with a rule
valid by our criteria of validity as being sufficiently
justified?” and that is a question which can be, and
from time to time is, raised. Nor can it answer the
question yet more frequently raised for judges: “How
ought we to justify decisions concerning the
interpretation and application of our criteria of
55
validity?”

54
To illustrate this point, we can quote the following excerpt from an interview that MacCormick gave to Manuel
Atienza on the occasion of the publication of his Institutions of Law: “The most illuminating and lasting aspect of
Hart’s writings has to do with the need to understand any conduct regulated by rules from the ‘internal point of
view.’ This is essentially to develop a clear and convincing theory of norms. But rules are just one type of norm.
The analysis of Law as a system of primary and secondary rules, although a valuable intuition, is at the end
incomplete and unsatisfactory. A fresh start is needed.” Manuel Atienza, Entrevista a Neil MacCormick 29
DOXA—CUADERNOS DE FILOSOFÍA DEL DERECHO 479, 482 (2006) (trans. unknown).
55
NEIL MACCORMICK, LEGAL REASONING AND LEGAL THEORY 63 (1978).
This passage shows that MacCormick is worried about something that is usually neglected
by contemporary positivists: the problem of the justification of legal decisions. He is
especially concerned with the justification of legal decisions because he is aware of the
subjective element that is always present, in greater or lesser extent, in the activities of
interpretation and application of law. Furthermore, he assumes the point of view of the
“norm user” when he takes up the task of constructing a theoretical explanation of the
nature of law.

Although the institutional theory of law, which is MacCormick’s legal theory in a strict
sense, presupposes some sort of detachment from the part of the theorist, such
theoretical inquiry “remain[s] nevertheless value‐oriented” in the descriptions it makes of
56
the legal institutions. According to MacCormick, “a coherent account of the nature of
law, and a coherent account of the character of any modern legal system, have to take
seriously the very general values that are inherent in the character of the legal enterprise.”
The concepts that a legal theorist uses to explain the nature of law are thus “interpretative
concepts” in the sense of Ronald Dworkin, for the theorist must adopt a hermeneutic
approach which “seek[s] to understand the practices and institutions of human beings in
terms of what makes them intelligible and worthwhile . . . to their human participants.” 57

Although the theorist has a certain degree of detachment when contrasted with the
“front‐line actors” of the legal practice, she must have, as a second‐line actor, “a relatively
high degree of engagement by contrast with purely external observers.” 58 In spite of the
fact that MacCormick claims that his jurisprudence is not aimed at the solution of
particular current practical problems, 59 the understanding which it provides for the
practical category of “law” presupposes a rational reconstruction of the legal institutions
that “yields a critical account of the governing rules in the light of the principles and values
which underpin them.” Such a critical account, MacCormick proceeds, “may indicate the
scope for future interpretations of law that will rectify anomalies in current understanding,
including current judicial practice.” 60

The understanding of legal theoretical inquiry as predominantly critical and value‐laden,
adopted by MacCormick, is coupled with a view of the practice of legal argumentation as
56
MACCORMICK, supra note 45, at 301.
57
Id. at 295.
58
Id. at 6.
59
Id. at 302.
60
Id. at 292. Like Alexy, MacCormick claims that legal doctrine has not only empirical and analytical dimensions,
but also a normative one.
2011]
also a necessarily constructive or hermeneutic process, which seems to be at odds with the
empirical essentialism that permeates most of the contemporary positivists.

For many positivists the law is some kind of real entity that is there to be “known” or
“discovered” by the theorist. Theorists very rarely engage in an interpretative or
hermeneutic activity when they attempt to “identify” the law. The law has a factual
essence which can be observed or described from the outside. The theorist looks at the
law from an external point of view, and she sees a static system of norms which can be
theorized in a sort of Platonic way, since its contents are already fixed and their meaning
does not depend on the subjectivities of the interpreter. This type of positivism, which is
widely spread in Anglo‐American jurisprudence, holds that there is a clear separation
between the “creation” and the “application” of law. One can identify quite easily the
“valid law” in any given society, although a legal official may eventually “create” new laws
when she is not satisfied with the solution provided by the legal order or, more frequently,
when she is dealing with a case not yet resolved by the set of valid laws. For this strand of
positivism, it is possible to “apply” the law without any balancing of reasons. The so‐called
“application of law” is merely the executive stage of practical reasoning. All of the law can
be found in its social sources, and the major task of the jurist is to identify these sources,
since they give her an orientation about what one is legally required to do in a particular
case.

Such clear distinction between the “creation” and the “application” of law leads to another
distinction which is typical of contemporary Anglo‐American positivism: the distinction
between “theories of law” and “theories of adjudication.” Theories of law are generally
concerned with the identification of law. They answer to the question of what the law is,
not what it ought to be. There must be a single test for differentiating legal rules from
other type of social norms, and such test normally has to do with the pedigree or the
process by means of which the rules are created (in other words, with its sources). In this
perspective, the law is self‐referential because it regulates the process through which legal
norms are created. This distinction between “theories of law” and “theories of
adjudication” presupposes some grain of positivism. In fact, it is hardly possible to find a
non‐positivist who would be willing to accept this separation.

The differentiation between a theory of law and a theory of adjudication is crucial for the
debate about contemporary positivism, since it illustrates how this kind of theory
characterizes the function of jurisprudence as a theoretical discipline. Once jurisprudence
is classified as a theory of law, as opposed to a theory of adjudication, one can notice an
increasing gap between theory and practice, which are conceived as autonomous
discursive contexts which very rarely communicate with each other. It is not the task of a
theorist to justify any practice or particular decision or to provide any guidance for the
proper interpretation and application of law. The proper task of legal theory is merely to
explain the law, rather than to develop or revise it. A good legal theorist must, therefore,
leave the object of her inquiry untouched. Legal theory and legal practice are separate
domains, which should remain separate if they are to comply with their social functions.

I have no doubt that this is how Petroski conceives the task of legal theory: Legal theorists
and judges operate within different institutional contexts. Hence, it seems puzzling to call
a practitioner “positivist” or to apply to him the concepts and positions advanced in the
jurisprudential discourse. For Petroski, when one classifies a judge as a “positivist,” this
can only mean one of these two things: (1) “that the judge speaks from an academic
domain despite his or her role as a judge, or” (2) “that the judge adopts the rhetoric of the
61
academic domain.” In both cases, however, the use of the vocabulary of legal theory by
judges or other members of the legal profession is perceived as unnatural, if not
illegitimate. For example, when an allegedly positivist judge like Justice Scalia classifies
himself as a positivist, Petroski thinks that what he is doing is merely deploying some
“interactive expertise” in order to adopt a rhetoric which would allow him to “stress on
statutory and constitutional text and precedent as the only legitimate reasons to offer in
support of a judicial decision” and to disavow any reliance on policy or moral
62
considerations. This “interactive expertise,” however, is considered to be anomalous
and interstitial, since the speaker is actually “master[ing] the language of a specialist
domain in the absence of practical competence” for doing it. 63 When we find arguments
from jurisprudence in a particular judicial opinion, this is explained not as something
natural and immanent to the activity of adjudication, as Dworkin claimed when he said
that “jurisprudence is the general part of adjudication,” 64 but rather as an intrusion into a
different domain.

This way of conceiving the spheres of jurisprudence and legal practice, however, only
becomes plausible if one presupposes a positivistic conception of legal theory. Hence,
Petroski is perhaps giving way to certain positivist prejudices when analysing a non‐
positivist theory such as that of Neil MacCormick. She assumes that the functions of
MacCormick’s theory of law and the motives that he has to undertake his jurisprudential
inquiry are the same as those of the particular type of positivism with which she impliedly
identifies herself. We should also bear in mind that MacCormick is attempting to construct
a legal theory which makes sense to the participants in the legal practice, even if it is not
primarily aimed directly at solving individual problems. When we set out for ourselves the
task of understanding a legal system, this involves both a description and a reconstruction
of the legal practice and the principles which underlie it, and this is “not simply a matter of
61
62
Id. at 686‐7.
63
Id. at 687.
64
DWORKIN, supra note 10, at 90.
2011]
detached understanding from an observer’s viewpoint. It is also a practical understanding
from a participant’s viewpoint.” 65

If one takes a different view of MacCormick’s recent theory of law, without presupposing
the positivist separation between theories of law and theories of adjudication, one is likely
to see many points where he definitively breaks down with positivism. This would be my
reading of MacCormick, which I shall consider in the following topics.

1. The Arguable Character of Law

The first point about Neil MacCormick’s institutional theory of law is that the law
comprises more than just a set of institutional facts which do not require any further
interpretation. As a normative order, “and as a practical one,” the law is “in continuous
66
need of adaptation to current practical problems.” In this particular, MacCormick
accepts Kelsen’s idea that the law presents a hierarchical structure in which the higher‐
level norms are specified or concretized in the processes of legislation and adjudication. In
European Law, for instance, the transposition process from the level of supranational law
to the municipal legal orders “is a part of the Stufenbau, that is, the step‐by‐step process
from abstract general enactment to particular decisions in concrete cases.” 67 As
MacCormick puts it with particular clarity, he sees the law as “an argumentative
discipline,” rather than an “exact science.” His first commonplace about the law is that it is
always arguable in nature. 68 Unlike the positivists who place legal certainty as the only
value secured by the Rule of Law, MacCormick believes that the proclaimed
“argumentative character of law” is something to be celebrated in democratic societies, for
it is deeply entrenched in the ideal of the Rule of Law. The recognition of the Rule of Law
as a political ideal implies the recognition of law’s domain as the “locus of
argumentation.” 69 Although the principle of the Rule of Law is oriented towards the value
of legal certainty, this value is not the only one. Rationality and justice also figure amongst
the basic values which form the basic ideal of the Rule of Law.

65
NEIL MACCORMICK, RHETORIC AND THE RULE OF LAW: A THEORY OF LEGAL REASONING 6 (2005). It is not a surprise,
therefore, that MacCormick has recently confessed that his interest in legal philosophy developed from an
attempt to reconcile philosophy and legal practice. When asked by Manuel Atienza about the roots of his legal
philosophy, MacCormick answered in the following way: “M. A.: Why have you become interested in legal
philosophy? N. M.: Because I was fascinated about philosophy, but wanted to dedicate myself to the practice of
law.” Atienza, supra note 54, at 480.
66
67
Id. at 10.
68
Id. at 14–15.
69
In this interpretation, the indeterminacy of law is not always something to be deplored. It
has to do with the ideal of the Rule of Law and with the procedural rules of argumentation
that are presupposed in the institutional structure that it provides. The ideals of
impartiality and equality between the parties in a legal dispute, as well as the fundamental
principle audiatur et altera pars, are necessarily connected to the basic idea of the Rule of
Law and to the arguable character of the legal system. In this sense, it is worth referring to
MacCormick’s words in the introductory chapter to his updated theory of legal
argumentation:

I do believe in the argumentative quality of law, and
find it admirable in an open society. We should look at
every side of every important question, not come down
at once on the side of prejudice or apparent certainty.
We must listen to every argument, and celebrate, not
deplore, the arguable quality that seems built in to
70
law.

The Rule of Law itself, MacCormick would say, implies a certain degree of indeterminacy in
the legal system. Such “indeterminacy,” according to the author, “is not merely” “a result
of the fact that states communicate their legal materials in natural (‘official’) language, and
that these are afflicted with ambiguity, vagueness and open texture,” like Hart would
say. 71 “It also results from” “the due recognition of the ‘rights of the defense’ in every
setting of criminal prosecution and civil litigation.” 72 In short, the Rule of Law implies and
in a way magnifies the arguable character of law.

If this interpretation of the political ideal of the Rule of Law is correct, then traditional legal
theorists are wrong when they present the value of legal certainty as the only substance of
the Rule of Law. Moreover, theorists are also mistaken if they claim, like Raz does in his
positivist legal theory, that the law is to be “found” in a previously determinate set of
“social sources,” by means of an entirely empirical reasoning. The validity of a law cannot
be merely a question of fact, but rather needs to be at least in a significant part a matter of
argument. In MacCormick’s view, the law is hardly ever “settled” and its rules are
necessarily defeasible, for they are inevitably subjected to a constructive interpretation
that might lead to revisions, reinterpretations, and even exceptions to their operative
conditions. Legal rules are “regarded as stating only ‘ordinarily necessary and
presumptively sufficient’ conditions for the arrangements they regulate,” since the
underlying principles that provide a general justification for the legal system interacts with
70
Id. at 16.
71
Id. at 26.
72
Id.
2011]
the more specific provisions that are found in statutes, precedents and in the secondary
legislation. 73 Such interaction may activate some background factor covered by the
justifying principle, which is likely to amount to the recognition of unstated exceptions to
the rule initially considered by the interpreter. To put it in MacCormick’s own words,
“[t]he presence of unstated elements appears to be a general feature of law.” 74

2. The Interpretative Reading of the Sources of Law

At this stage, one can see why MacCormick’s interpretative theory of law and legal
adjudication, which is in an important measure influenced by Dworkin’s conception of “law
as integrity,” is one step ahead of the positivistic formulation of the so‐called “Sources
Thesis,” which seems to be the only common point to all positivists.

It is a common assumption of positivists that the decisive criterion to determine the
validity of a legal norm, and therefore its distinctive legal character, is the pedigree or the
source of that particular legal norm. When a rule can be traced back to a source which is
legally recognized, jurists can identify it with a reasonable degree of certainty. Even the
non‐orthodox forms of positivism such as the “inclusive” theories of Coleman, Waluchow,
and the late Herbert Hart have to agree that whatever comes out of a valid source of law
has a proper legal character, provided that it satisfies the master rule established in a
75
particular legal system as a test for the validity of its norms.

MacCormick’s recognition of the immanently arguable character of law, in turn, seems to
imply a completely different doctrine of the sources of law. The materials found in the
sources of law, such as statutes, precedents, treaties, and administrative acts are not “self‐
interpreting and self‐applying.” 76 Properly considered, they are “law” only in a pre‐
interpretative sense. 77 Perhaps MacCormick’s views on the theories of precedent are a
good example to illustrate this point. MacCormick argues that the theories of precedent
based on natural law and on positivism have contributed not only to alternative
73
Id. at 241.
74
Id. at 244.
75
In this particular, an inclusive or “soft” positivist would have to concede that it is at least theoretically possible
to conceive a perfectly valid legal system that does not incorporate any moral principles to its rule of recognition.
Whatever connections one can find between law and morality will be contingent, as opposed to necessary. See
generally Jules Coleman, Negative and Positive Positivism, in RONALD DWORKIN AND CONTEMPORARY JURISPRUDENCE
(Marshall Cohen ed., 1984); WILL WALUCHOW, INCLUSIVE LEGAL POSITIVISM (1994).
76
77
Neil MacCormick, Precedent as a Source of Law, in SOURCES OF LAW AND LEGISLATION: PROCEEDINGS OF THE 17TH
WORLD CONGRESS OF PHILOSOPHY OF LAW AND SOCIAL PHILOSOPHY, BOLOGNA, JUNE 16–21 1995, ARSP—BEIHEFT 177, 183
(1998).
interpretations of the doctrine of stare decisis in the United Kingdom, but also to different
practical uses and applications of the case law. On the one hand, the so‐called
“declaratory theory of adjudication”—which presents itself as compatible with a natural
law viewpoint and dominated the British scene until early in the 18th Century—“supports
hostility to any doctrine of absolutely binding precedent, and tends towards a view of
precedents as at most defeasibly binding, on the grounds that mistakes about law are
logically possible, and precedents only declaratory or evidentiary, not strictly constitutive
of law.” 78 On the other hand, positivism, which stresses the human character of law,
“necessarily denies the premises of the declaratory theory”: 79

There is no essence of law beyond or behind what is
decided as law by some competent decision maker.
From this it follows obviously that, if precedents are
evidence of the law, they can be so only because judges
explicitly or implicitly accorded authority to make law
through their decisions. Conversely, the very
recognition of precedent as evidence of the law
amounts to recognition of the power of the courts to
80
make law.

We can see, thus, that these theories produce serious normative consequences in the way
a judicial precedent is received and applied in legal practice. It is because of such
consequences that MacCormick, in his approach to precedents, holds that we need a
theory that can overcome the dichotomy “natural law versus positivism,” for this is the
only way to transcend the limits of the declaratory theory and its positivist counterpart:

It is no longer possible . . . to rest content with a simple
contrast of natural law and positivism in the treatment
of precedent. For this, particular tribute is due to the
work of Ronald Dworkin, which has subverted the
simple dichotomy of positivism versus natural law
theory. In place of a model of law as a system of rules
derived from predetermined sources, Dworkin invites
us to re‐conceive law as an essentially “interpretive”
concept. The whole body of decisions by legislature
and judges and others that positivism has
conceptualized as making up “a legal system” Dworkin
78
Id. at 182.
79
Id. at 182.
80
Id. at 183.
2011]
tells us to acknowledge as law only in a “pre‐

interpretive” way. 81

As MacCormick explains, according to this conception the very notion of “sources of law”
needs to be revised: “For a follower of Dworkin’s view, if we treat precedent, or indeed
Parliamentary or Executive enactment, as a ‘source of law,’ the conception of ‘source’ in
play must be radically different from that assumed in the positivistic model . . . .” 82 By
following the reasoning of Dworkin, MacCormick ends up with a theory of the sources of
law according to which statutory provisions, precedents, case reports, and other legal
materials “are not really themselves law, but rather are source‐materials from which the
law‐making task of the moment must work.” The specification of the meaning of a law
depends on an interpretive process guided by the idea of “integrity,” in Dworkin’s sense, 83
or “normative coherence,” which is the same idea in MacCormick’s vocabulary. 84 In such
constructive interpretation, the interpreter makes the law operative by applying these
legal materials according to the moral principles implicit in the law, and this hermeneutic
process is always guided by the idea to achieve the best possible interpretation of these
pre‐interpretive materials. We can see, thus, that MacCormick advocates a theory of
sources of law and a theory of precedents that are incompatible with legal positivism and
with the view that the authority of a source is the only factor that is relevant to determine
the validity of the rule which can be extracted from a judicial precedent. The norms
derived from precedent will be fixed in an interpretative discourse where arguments based
on authority and arguments based on practical reason are balanced in a proper way.

3. MacCormick’s Theory of Legal Argumentation

The arguable character of law and the conception of sources of law as “pre‐interpretive”
materials rather than definitive answers to the question of the validity of a rule are deeply
consistent with MacCormick’s ambitions to provide a normative theory of legal reasoning.
If the law necessarily refers to some conception of justice and if its ultimate contents
depend on constructive interpretations in the context of disputable validity claims to
support the individual norms produced in the process of application of law, then it is
reasonable to argue that one of the tasks of legal theory should be to provide a set of
directives to help lawyers “discriminate between better and worse arguments, more and
85
less rationally persuasive ones.” In this sense, part of MacCormick’s strategy to reconcile
81
Id.
82
Id.
83
DWORKIN, supra note 10, at 95–96.
84
See generally MACCORMICK, supra note 55; MACCORMICK, supra note 65.
85
MACCORMICK, supra note 65, at vi.
the argumentative dimension of law with the value of certainty—which are both
presupposed and implied by the political idea of the Rule of Law—is to elaborate a theory
of legal argumentation that claims to be capable of acknowledging the best criteria of
rationality for legal decisions and the fundamental constraints on the process of legal
reasoning.

Although much could be said about MacCormick’s penetrating theory of legal reasoning, it
is enough for our purposes to make reference to the Kantian universalism which is
accepted in the latest version of his theoretical project. For MacCormick, any legal
decision cannot be justified unless it passes a test of universalizability. This follows from
the very idea of justification: “To justify is to show that it is right,” and “to show that it is
right is to show that, upon any objective view of the matter, the act ought to have been
done, or even had to be done, given the characteristics and the circumstances of the
86
case.” There is no justification without universalization in MacCormick’s theory of legal
reasoning: “For particular facts—or particular motives—to be justifying reasons they have
to be subsumable under a relevant principle of action universally stated, even if the
universal is acknowledged to be defeasible. This applies to practical reasoning quite
generally, and to legal reasoning as one department of practical reasoning.” 87 This is
enough to show us that MacCormick’s theory of argumentation strives towards some
universality test in the spirit of Kant’s categorical imperative, even if this test is adapted to
the form of Perelman’s ideal of the “universal audience” 88 or Habermas’ regulative idea of
the “ideal speech situation.” 89

In any case, MacCormick is very clear about the importance of the moral principle of
universalizability when he expressly recognizes that the latest version of his theory of legal
reasoning has departed from the Humean non‐cognitivism found in his earliest writings in
favor of a Kantian universalistic moral philosophy. 90 In this particular, MacCormick’s
theory of legal argumentation has to incorporate two theses put forward by Robert Alexy
in order to vindicate his claim that the principles of Kantian universalistic morality are
relevant for legal reasoning: the thesis of the “claim to correctness” and the “special case
thesis.” If the law is a special case of practical discourse and, as such, raises a claim to
correctness, then the basic rules of argumentation which apply to practical discourses are
86
Id. at 98.
87
Id. at 99.
88
See e.g., CHAIM PERELMAN and LUCIE OLBRECHTS‐TYTECA, THE NEW RHETORIC: A TREATISE ON ARGUMENTATION 31–35
(1969).
89
Jürgen Habermas, Teorías de la Verdad, in TEORÍA DE LA ACCIÓN COMUNICATIVA: COMPLEMENTOS Y ESTUDIOS PREVIOS
113–58 (1997). For a comprehensive analysis of Habermas discourse theory in English, see generally THOMAS
MCCARTHY, THE CRITICAL THEORY OF JÜRGEN HABERMAS (1978).
90
2011]
also valid for legal discourse. The same principles of discourse that one can find in Kantian
moral philosophy are also valid for legal discourse, since the latter is nothing but a case of
the former with a set of institutional constraints. 91 Without these two theses, MacCormick
would not be able to vindicate his claim, which is common to Alexy, that the principles of
moral argumentation proposed by thinkers like Kant and Habermas are applicable to legal
reasoning. Hence, it is not difficult to notice that MacCormick’s theory of legal
argumentation—at least in its latest form—presupposes a breakdown with the positivist
tradition. As I intend to demonstrate in the final sections of this paper, the theses of the
“special case thesis” and of the “claim to correctness,” in the form that MacCormick
interprets them, cannot be reconciled with any form of positivism.

4. The Thesis of the “Claim to Correctness”

One of the points in which MacCormick departs from positivism is when he claims that the
law raises an implicit pretension to justice. In fact, MacCormick holds that legal discourse
is characterized by the presence of certain implicit claims which are “necessarily bound up
with the performance of the interactive roles of the law‐making, judicial, and executive
92
institutions or agencies of the state.” Every speech act which introduces a legal norm or
exercises some sort of legal authority is connected with the illocutionary act of asserting
the correctness of the general or individual norm which is produced thereby. Any speech
act performed in the context of legal argumentation involves certain background
assumptions or implicit claims, and in particular a claim to legal and moral correctness. 93
Considered as an institutional normative order, the law strives to justice and necessarily
purports to resolve practical conflicts in a morally acceptable way: “[A] certain pretension
to justice, that is, a purported aspiration to be achieving justice (even if this be the mask of
a more partisan or sinister intent) is necessarily evinced in the very act of law‐making in
the context of a law‐state.” 94

The claim to correctness includes, although it is not satisfied with, a claim to moral
correctness, and not only a claim to correctness according to the positive laws of a given
state. Such a claim, which plays a central role in Alexy’s theory of legal argumentation and
now seems to perform an analogous function with regards to MacCormick’s latest version
of his theory of legal reasoning, is intimately connected with the idea that the law has an
91
This obviously does not mean, however, that law and morality are not distinguishable. The fact that legal
discourse operates within a set of institutional constraints implies that legal decisions are limited by the exigency
to respect the statutes, precedents, and other authoritative materials found in a legal system. It is this constraint
in practical legal reasoning that makes it a special case of practical general discourse.
92
93
Id. at 275.
94
Id. at 276.
argumentative character. The foundations of this thesis can be found in Habermas’
philosophy of language. In his famous essay “Theories of Truth,” originally published in
German in the early 1970’s, Habermas criticizes the philosophical theories which define
the truth of an assertion as the “correspondence” with a certain thing or state of affairs
that is proved to be existent in the world of physical objects. 95 In contrast to these
theories, Habermas advocates a conception of truth as rational consensus, which can be
summarized in the following excerpt:

I can predicate something to an object if, and only if, all
the individuals that could enter in a discourse with me
would predicate the same thing to the same object.
Hence, to distinguish the truth from the falsity of an
assertion I make reference to the judgment of the
others—in fact, to the judgment of all of those with
whom I could eventually enter in a discussion (among
which I counterfactually include all the hearers that I
could meet if my lifetime were coextensive with the
history of mankind). The truth‐condition of a
statement is the potential consensus of all the
96
participants in the discourse.

With this ideal discourse, Habermas seeks for the criterion of truth which he claims to be
absent in the epistemological theories that identify the truth of an assertion with a mere
correspondence, mediated by sensorial experience, to a certain state of affairs. Since we
all have different experiences, a conception of truth based solely on sensorial perception
would prove to be mistaken because it cannot guarantee the objectivity of the knowledge
it provides. Since there is no reality objectively accessible to our senses, the rationality of
the expressions used by A or B can only be assessed in terms of the discursive
redeemability of the validity claims contained in the speech acts performed by each
speaker. In other words, by understanding the discourse as an argumentative procedure,
Habermas is claiming that in the linguistic interactions between A and B “both raise claims
with their symbolic expressions, claims that can be criticized and argued for, that is,
97
grounded.” A judgment—about the truth, in the case of constatative speech acts, or
about the correctness, in regulative speech acts, where what is at stake is not the assertion
of a fact, but rather the validity of a norm 98 —can only be “objective if it is undertaken on
the basis of a transsubjective validity claim,” in such a way that the “assertions and goal‐
95
Habermas, supra note 89.
96
Id. at 171.
97
JÜRGEN HABERMAS, 1 THE THEORY OF COMMUNICATIVE ACTION: REASON AND THE RATIONALIZATION OF SOCIETY 9 (1984).
98
Habermas, supra note 89, at 130.
2011]
directed actions are the more rational the better the claim that is connected with them can
be defended against criticism.” 99 Hence, Habermas needs to adopt a concept of
communicative rationality which is based on a discourse whose formal properties are
capable of generating an unconstrained and unifying consensus. 100

Like assertions about facts, normatively regulated actions also contain signifying
expressions which are connected to criticizable validity claims. The norms to which these
actions refer also can be inter‐subjectively recognized, and the discursive redeemability of
the validity claims raised for these norms is what constitutes their rationality. 101 In the
field of ethics, Habermas adopts a cognitivist position according to which practical or
normative questions can be resolved by means of an argumentation put forward in a
practical discourse where the correctness of the norm is scrutinized. 102 This type of
discourse is a communicative process that can only be rational if it approaches the
following “ideal conditions”: (1) the speakers must be in an “ideal speech situation” which
is basically characterized by the “general symmetry of conditions,” so that each participant
in the discourse can structure his speech acts in a way that there is no coercion other that
the rational force of the better arguments; 103 (2) the argumentation, as a process of
communication, must be understood as a “a form of interaction subjected to special
rules”; and (3) the argumentation “has as its aim to produce cogent arguments that are
convincing in virtue of their intrinsic properties and with which validity claims can be
redeemed or rejected.” 104

Since I cannot move any deeper into this theory of rational discourse, I will limit myself to
stress one of the points which appear to be central to Alexy and MacCormick: the idea
that regulative speech acts always raise a claim to normative correctness. It is on the basis
of this premise that Alexy argues that in every act of creation and application of law a claim
that such act is correct is implicitly raised. In the core of this claim there would be: (1) The
assertion that the legal act is substantially and procedurally correct; (2) The claim (which
generates a guarantee) of justifiability of this assertion; and (3) The expectation of
acceptance of correctness by all addressees of the legal norm. 105 But how can Alexy justify
this thesis? He answers this question in the following way: “[A]n implicit claim can be
99
HABERMAS, supra note 97, at 9.
100
Id. at 10.
101
Id. at 15–16.
102
Id. at 19.
103
Id. at 25.
104
Id. at 25.
105
Robert Alexy, Law and Correctness, 51 CURRENT LEGAL PROBLEMS 205, 208 (1998).
made explicit by showing that its express negation is absurd.” 106 He follows, thus, a
formal‐pragmatic strategy of showing that the explicit negation of the claim to correctness
entails a contradiction between the content of the legal act (be it a statute, a judicial
decision etc.) and the content of the claim implicit in its enactment. Alexy calls this kind of
contradiction a “performative contradiction.” 107 Every participant in legal discourse who
expressly denies the claim to correctness commits a contradiction of this kind. It would be
a performative contradiction, for instance, if in the enactment of a constitution the first
article expressly asserted that “X is an unjust State.” The same would happen if a judge
pronounced a sentence stating something like “X is hereby convicted to life imprisonment,
although this is an incorrect interpretation of valid law.” This judge’s decision could be
classified as absurd because it would contain both an implicit assertion that it is a correct
decision and an explicit assertion which contradicts the implicit part.

Alexy argues, in one of the main points of his argument to justify his definition of law, that
the claim to correctness has the function of establishing a necessary connection between
law and morality, which is described as a complex connection that has at the same time a
qualifying and a conceptually necessary character: It is a necessary character because
every legal system necessarily raises a claim to correctness (and thus there can be no legal
system that does not raise the claim), but it is a qualifying character (unlike a classifying
character) because the legal systems and legal norms that do not fulfill the exigencies of
the claim to correctness may remain valid in a legal sense, despite being conceptually
108
flawed. The claim to correctness generates a weak duty to do justice according to the
law, not an absolute one.

But a positivist jurist could still ask: What is the relevance of a claim to correctness, if it has
only a qualifying character? Alexy contests this question by arguing that the claim to
correctness is not only a moral claim, but also a legal claim. And this legal claim
“corresponds with a legal obligation necessarily connected with judicial judgments to hand
109
down correct decisions.” It attributes to law an ideal character that is especially
relevant for those who analyze law from the point of view of the participants. Thus, when
application of the law results in an injustice (and therefore does not fulfill the state of
affairs required by the claim to correctness), we are facing not only a moral fault, but also a
legal fault. In Alexy’s own words, “[t]he claim to correctness transforms moral faultiness
into legal faultiness. And this is by no means trivial. It is the conversion of positivism to
106
Robert Alexy, Derecho y Moral, in LA INSTITUCIONALIZACIÓN DE LA JUSTICIA 21 (2005).
107
108
See generally ALEXY, supra note 10.
109
2011]
non‐positivism. Law’s claim to correctness is on no account identical with the claim to
moral correctness, but it includes a claim to moral correctness.” 110

The claim to correctness attributes to law an ideal character that was not present in the
positivistic accounts that defined the legal system as mere facticity or the result of the
exercise of authority. This ideal dimension of law supports a general principle of morality
which is valid as a legal norm implicit in every legal system. Along with this claim, Alexy
recognizes a norm pragmatically presupposed that contains a duty of correctly interpreting
and applying the legal norms. As a consequence of the claim to correctness, one can argue
111
that there is a methodological argumentative connection between law and morality
which implies that legal officials have a prima facie legal duty to do “justice according to
the law.” This is precisely what MacCormick means when he holds that the law is
“necessarily geared to justice,” and such claim is obviously at odds with any form of
positivism. To say the least, it attaches to law an aspirational character and introduces a
distinction between the “actual positive law” and the “ideal positive law,” 112 as well as a
discursive obligation for legal agents to attempt to reconcile the two of them in their
constructive interpretations.

5. The Special Case Thesis

Furthermore, a few words can be said about the thesis that the legal reasoning is a special
case of practical discourse. The special case thesis is relevant for the theories of legal
argumentation because it makes it easier to see the formal‐pragmatic presuppositions of
legal discourse. When we say that legal discourse is a special case of practical discourse,
we commit ourselves to the view that there are some discursive obligations for the
participants in this type of discourse. Whatever constraints are applicable to general
practical discourses are also valid for legal discourse, albeit with the additional condition
that the speakers have to comply with the positive laws with which they operate. As
MacCormick writes to express his agreement with Robert Alexy’s original formulation of
the thesis, “legal argumentation must be acknowledged to be one special case of general
practical reasoning, and must thus conform to conditions of rationality and reasonableness
113
that apply to all sorts of practical reasoning.”

110
Robert Alexy, On the Necessary Connection between Law and Morality: Bulygin’s Critique, 13 RATIO JURIS 138,
146 (2000).
111
See generally Alexy, supra note 105.
112
113
The special case thesis is thus relevant because it connects law, rationality and rhetoric. It
implies some rhetorical duties 114 for anyone who enters a legal discourse or raises a legal
claim. Hence, if Petroski is right when she holds that positivists agree that law and legal
reasoning can be characterized by the fact that they “exclude” certain reasons—namely
any kind of moral, political, or pragmatic reasons—from the set of arguments available for
legal actors, then she must be wrong when she argues that MacCormick’s post‐positivism
is not really transcending the theoretical background of mainstream contemporary
positivism.

If I am correct, MacCormick’s adherence to the special case thesis implies that moral,
political, and pragmatic reasons that apply in general practical discourse are necessarily
included in legal reasoning, which is precisely the opposite of what positivists say according
115
to Petroski’s account of their legal theories.

6. The Ethical Life of the Legal Institutions

Last but not least, MacCormick’s institutional theory of law is very different from
positivism because it refuses to accept its description of the law as an inherently
formalistic and monological system of previously given rules. In this sense, my impression
is that most of the contemporary positivists tend to be so strongly committed to their
methodological prejudices that they usually fail to address one of the most interesting
points of MacCormick’s institutional theory of law, which has to do with the ethical life of
116
the legal institutions. On the basis of a “selective” reading of MacCormick’s works, they
highlight a handful of common places in his institutional theory and end up claiming that
they find in MacCormick an ally rather than an opponent. 117 When they do it, however,
they tend to distort the theory that they are analyzing. In fact, MacCormick’s institutional
theory presupposes an element of mutual understanding among the members of the legal
community that is intrinsically built into the law. 118 The law emanates from the social
114
THEODOR VIEHWEG, TOPIK UND JURISPRUDENZ (1953).
115
If we accept this argument, we will have to deal with new problems such as, for instance, that of the potential
conflict between “positive” arguments in a strict sense and the principles of justice that can be vindicated by
means of the rules of general practical discourse. Neither MacCormick nor Alexy accept that there is always a
primacy of strictly positive reasons in such case. Nonetheless, I do not need to deal with this kind of problems
here.
116
See generally Zenon Bankowski, Bringing the Outside in the Ethical Life of Legal Institutions, in LAW AND LEGAL
ST
CULTURES IN THE 21 CENTURY (Tomasz Gizbert‐Studnicki & Jerzy Stelmach eds., 2007).
117
For some examples of positivists trying to read Neil MacCormick as belonging to their camp of theoretical
discussion see, e.g., Frederick Schauer, Positivism as Pariah, in THE AUTONOMY OF LAW: ESSAYS OF LEGAL POSITIVISM
32–56 (Robert George ed., 1996); Vittorio Villa, Neil MacCormick’s Legal Positivism in LAW AS INSTITUTIONAL
NORMATIVE ORDER 44–64 (Makysimilian Del Mar & Zenon Bankowski eds., 2009); WALUCHOW, supra note 75, at 1–4.
118
Neil MacCormick, Norms, Institutions and Institutional Facts, 17 LAW AND PHILOSOPHY 301, 305 (1998).
2011]
interaction of the members of such community, which is mediated by a discursive
procedure in which these members influence each other and are transformed by the
experiences that they gain from such interaction. This “mutual understanding” cannot be
of any unspecified kind whatsoever, since what makes a rule institutionally recognizable is
not only the conformity with a certain “constitutive rule” that determines its meaning—as
it happens, for instance, in Searle’s model of constitutive rules 119 —but rather the
“underlying principles” that constitute the “final cause” of a certain institution. 120 Similarly
to Dworkin, MacCormick believes that the meaning of the social practice called “law”
derives in part from the principles that provide the general coherence of the legal system
and that belong at the same time to morality and to positive law. 121

The difference between MacCormick’s position and that of mainstream Anglo‐American
positivism appears to be very neat when we contrast the positivist claim to treat the law as
a “self contained and autogenerative system which needs to be kept distinct from politics
in order to organise our lives” 122 with the constructivist idea, which MacCormick explicitly
accepted, that the law is open to discourses of application that may lead to its adaptation
and eventual reinterpretation on the basis of moral and ethical considerations. 123
According to the best interpretation of MacCormick’s institutional theory, the law and its
institutions create a sort of ethical space where coordination for the common good
becomes possible and the individuals living in a political community can participate in the
formation of the values and principles which apply to them. 124 The legal institutions have
a value in themselves and the members of the political community recognize their laws not
merely as a set of certain and predictable orders issued without their participation, but
rather as a normative system that expresses a form of life to which they belong and which
is constitutive to their identify and their self‐understanding. As Zenon Bankowski
convincingly argues, the positivist camp of this debate tends to alienate the individuals in a
capitalist society and to reduce them to the condition of players in a rule‐based game
where their creativity and their critical faculties are narrowed down. 125 Again in the words
119
See generally JOHN SEARLE, SPEECH ACTS—AN ESSAY IN THE PHILOSOPHY OF LANGUAGE (1970).
120
MacCormick, supra note 118, at 305.
121
Nevertheless, MacCormick criticizes Dworkin’s claim to reconcile his legal constructivism with the moral thesis
(which is difficult to reconcile with constructivism) that there is only one correct answer for each and every legal
problem. As MacCormick argues, there is an ambiguity in the heart of the Ronald Dworkin’s, legal theory. Neil
MacCormick, Dworkin as Pre‐Benthamite, 87 PHIL. REV. 585–607, (1978).
122
Bankowski, supra note 116, at 198.
123
See generally KLAUS GÜNTHER, THE SENSE OF APPROPRIATENESS—APPLICATION DISCOURSES IN MORALITY AND IN LAW
(1993); MACCORMICK, supra note 65.
124
Bankowski, supra note 116, at 202.
125
Id. at 199.
of Bankowski, who is one of the most important interlocutors that MacCormick ever found
in his lifetime, we can say that:

The implications of this social attempt to make the
fantasy of control and certainty real are that you begin
to “act without thinking about it” since the reason to
value the institution or arrangement is that it helps
construct a manageable order in one’s life. The hope is
that this might carry over to the legal system as a
whole. In that case, it could be reasonable to follow
the law for its pedigree. It is law because it has been
validly enacted. . . . We might say then that in striving
so hard to make the normative [order] predictable by
ensuring that we “do not think about it,” we start
turning the normative into the descriptive so that we
act like automata running in a pre‐programmed way. In
126
order to be free we become slaves.

None of these negative implications, however, appears in the book that Petroski is
criticizing in her paper about Neil MacCormick’s post‐positivism. MacCormick’s Institutions
of Law is indeed a critical attempt to reconcile legal theory and legal practice, law and its
moral justification, politics and the ethical life of the community in which the law is
constantly reconstructed. The core of this theoretical project, in my perception, is to
transcend the limits of legal thinking that positivists set for themselves.

D. Conclusion

In the frame of a comment, the current analysis of MacCormick’s post‐positivism does not
claim to be a fully comprehensive one, since it does not intend to cover all the far‐reaching
implications of MacCormick’s post‐positivism. It does not even attempt to explain all the
points in which MacCormick departs from positivism. To name a couple of them, much
could be said about MacCormick’s particular opinion about the proper methodology for
jurisprudence—which stands somewhere in the middle between Dworkin and Alexy’s
participant viewpoint and the point of view adopted by Hart and his fellow positivists—and
about MacCormick’s thesis that any legal system necessarily comprises some “moral
127
minimum without which purported law becomes un‐law.” Although these issues are of
fundamental importance for contemporary legal theory, I cannot address them for reasons
of time and space. My ambition here was much more modest. I do not claim more than to
offer an alternative view to Petroski’s conclusions that post‐positivism is just a new label
126
Id.
127
2011]
for old ideas and that MacCormick is still a positivist who has prudential reasons to
differentiate his theory from that of his fellow academics. When Petroski claims that
MacCormick is making “fractal distinctions” motivated by purely prudential reasons, she
seems to me to underestimate the relevance of the challenges that he poses to
mainstream positivism. If I am right in my interpretation of MacCormick’s post‐positivism,
his point is not that all the criteria of identification of law advanced by positivism should be
abandoned. Neither is it that the institutional theory that he advocates is entirely free
from the methods of positivists like Hart and his followers. It is rather that the frame of
thought of positivism is too narrow, since it cannot answer many questions which figure
amongst the key problems that are still open for legal theory. How legal decisions can be
properly justified and what weight is to be attributed to moral reasons or other non‐strictly
institutionalized reasons in legal argumentation are among these. What MacCormick is
advocating with his post‐positivism is much more than a conventional critique or a
repetition of arguments against an old but enduring idea: it is rather a new paradigm for
jurisprudential inquiry. In this new paradigm, which connects law, morality, ethics, and
politics, the current positivist legal theories cannot survive unless they are radically
128
reinterpreted—up to the point where they tend to lose their positivist credentials.

Petroski’s paper is indeed very innovative within the positivist tradition. In its final
sections, for instance, it offers some new paths for jurisprudential inquiry and explores
some of the possibilities that are open to make legal theory a bit more interesting without
departing from positivism. For those who believe that jurisprudence should remain a
purely theoretical inquiry, undertaken from the point of view of the external observer, her
account is still a sound one. If one interprets Neil MacCormick’s theory of law as also a
purely theoretical inquiry from the same point of view, Petroski’s criticisms against his
post‐positivist position might be accepted. If, however, one understands MacCormick’s
theoretical project as both connected to the legal practice and intrinsically critical, as I do,
one can see that his legal theory is connected to his theory of argumentation, and that the
theses that he needs to vindicate such theory of argumentation bring him away from
positivism.
128
Jeremy Waldron, for instance, agrees with Neil MacCormick that legal philosophy cannot be understood apart
from its connections to the rule of law and its intrinsic relations to rationality and argumentation. As he wrote in
a recent essay, “a philosophy of law is impoverished as a general theory if it pays no attention to the formalized
procedural aspects of courts and hearings or to more elementary features of natural justice like offering both
sides an opportunity to be heard.” Jeremy Waldron, The Concept and the Rule of Law 43 GA. L. REV. 1, 55–56
(2009). A legal philosophy that fails to capture the arguable aspect of law (as well as its implications for the idea
of the rule of law) is depicted as “empty and irrelevant.” Id. at 56. Hence, Waldron expressly relies on
MacCormick to hold that law is indeed an argumentative discipline and that modern positivism is guilty of the
“fallacy” of putting its emphasis exclusively on the “command‐and‐control aspect of law.” Id. Whether or not
normative positivists such as Waldron will remain positivist once they accept this new paradigm that MacCormick
is advocating, one thing is sure: Even if positivism endures, it needs to be radically modified if its upholders
manage to provide a reasonable answer to the problems that MacCormick is bringing to the front with his post‐
positivism.
Special Issue
The Spirit of Legal Positivism

By Alexander Somek *

A. Gauging the State of Legal Knowledge

Legal Positivism is dead, isn’t it? We are all legal realists now. We believe, by default, that
what really matters in law emerges from some judicial process. 1 We sense that the point
of norm‐production by adjudication is to accomplish something useful or good for either
individuals or society at large. Practice trumps theory and policy implementation overrides
respect for some scholarly edifice. What we do when we do law is to unreel formula and
rhetoric. We engage in these exercises with the aim to have judges rule in favor of our
clients. Should we be judges ourselves, we promote causes we deem to be noble and fine.
Law is a tool. Skillfully mastered legal knowledge is a prerequisite for using it well.

I concede that this is an over‐generalized and highly pointed statement. It is meant to
capture a jaded state of mind to which Duncan Kennedy has since the late 1980s referred
as “postness.” The debates are over. Controversies over great ideas have subsided. The
scene is finally dominated by those whom Oliver Wendell Holmes long suspected to be
2
mere businessmen. Law is a trade. If you are smart and know how to enact your craft,
the forthcoming financial rewards will be sizeable. You’d better adapt.

I am confident that my account of the overall intellectual situation, even though
undoubtedly impressionistic, is not entirely mistaken about what has become common
wisdom among practitioners and those pursuing an academic career. The situation is
reflected, even though in different ways, in the general attitude towards legal scholarship
found in the United States as well as in Europe. In the case of the former, what is done in
the academic realm is largely, though not exclusively, based on the understanding that
producing law review articles is a way of improving, however causally obscure, the world.
While the style of exposition ranges from more fully fledged amicus curiae briefs to larger
blueprints of human life, the underlying attitude is that of the scholar attempting to do his
or her bit to make the world a better place. The law is secondary with regard to the
*
Charles E. Floete Chair in Law, University of Iowa College of Law. I would like to thank Robert Hockett, Todd
Pettys, Robert S. Summers, Sabine Somek and Brian Z. Tamanaha for valuable comments on an earlier draft.
1
Felix Cohen, The Problems of a Functional Jurisprudence, 1 MOD. L. REV. 5, 7–8 (1937).
2
Oliver Wendell Holmes, The Path of the Law, 110 HARV. L. REV. 991–1009 (1897).

general moral obligation to work towards amelioration, which comes first. Surprisingly,
the addressee of engaged writing is—aside from a court that pays scant attention to it 3 —
some imaginary sovereign invested with plenary power to implement the program.
Without exception, however, is it taken for granted that changing the world is a piecemeal
affair, which requires neither large‐scale social transformations nor, heaven forbid,
amending the United States Constitution.

The continental European milieu is different. 4 Scholarly output is expected to be useful for
those in the trenches. Hence, one gets an overabundance of handbooks, commentaries,
textbooks and outlines. More specific contributions are supposed to signal the capacity,
on the part of the authors, to generate meticulously drafted expertise. Incidentally, the
academic world is inhabited by people who consider themselves underpaid and are
therefore eager to sell the fruits of their labor (and the time and energy of their assistants)
to whoever is willing to pay for it.

The intellectual effect of this situation is not particularly favorable to legal theory, let alone
5
legal positivism, even though I should emphasize that the overall milieu is far more
hospitable in the United States than in mainland Europe. In the case of the former, legal
analysis is approached with the desire of social reform. The basic technocratic thrust of
this vision is often mitigated by widespread moral concern. I think it is therefore fair to say
that in the United States is prevalent a tendency to produce legal scholarship on moral
high ground. By contrast, in continental Europe legal knowledge comes in monetary form.
I do not mean to suggest, of course, that whatever legal knowledge exists is subservient to
one or the other clientele; rather, the form of legal knowledge—visible in its intellectual
horizon and rhetorical demeanor—is dominated by the drive to be marketable as
expertise. This leaves an imprint on its physiognomy. Legal knowledge is largely
descriptive, technical, and deferential in relation to courts. It lacks the courage to
challenge taboos. Simplifying matters even further, it makes sense to say that while legal
scholarship is highly moralized in the United States it is, by contrast, thoroughly
intellectually commodified in Europe.

3
The chances for a law review article to be cited by the United States Supreme Court are infinitesimally slim.
4
The United Kingdom is a very special case owing to its relatively late arrival to the world of academic legal
scholarship.
5
Of course, a host of other factors contribute to the fact that legal positivism appears to be terribly outdated. A
society marked by rapid technological development and the internationalization of commerce is difficult to
reconcile with a mindset for which legal codes or commands appear to be the paradigmatic instances of law.
RONALD DWORKIN, JUSTICE IN ROBES 212 (2006).

2011]
Spirit of Legal Positivism 731
In both cases, however, legal knowledge is controlled by some external agenda. It fails to
control the reasons governing its operation. The production of legal expertise either vies
for moral praise or panders to the interest of potential buyers. Owing to the social
circumstances governing its production, it remains in an intellectually underdeveloped
state.

B. Legal Positivism’s Demise

It would be too facile to single out courts as the culprits for the unacknowledged triumph
of these two different versions of being a realist—the moral reformist, on the one hand,
and the instrumentalist legal scientist, on the other. Undeniably, however, important
tribunals, such as the European Court of Justice and the German Federal Constitutional
Court, have done their share to devalue the relevance of scholarship. Both bodies have
developed their case law in amazingly brazen, free‐style fashion. The relentless social
6
engineering by the former and the Sacra Romana Rota style of exercising authority by the
7
latter have shattered the confidence among legal scholars that canons of legal reasoning
are able to exercise intellectual constraints in addition to serving as rhetorical makers for
the dispensation of expertise.

But the problem cuts deeper. An exceedingly high regard is currently paid to emanations
of the judicial process. We do not even realize that this is the case, let alone wonder why,
because we take for granted that what matters, legally, is what courts do in fact, since that
8
is what we have come to mean by the law. We may even assume this to be legal
knowledge’s natural state. But there is no such state. Rather, the overwhelming
importance of courts reflects scholarship in a state of abdication, that is, at a point where it
has already lost faith in its critical mission.

Widespread intellectual apostasy springs, indeed, from the core of the positivistic project
whose confidence has been shattered by the realization that one of its most cherished
9
theoretical dogmas has become exceedingly difficult to sustain. The belief in the
6
See generally DIARMUID ROSSA PHELAN, REVOLT OR REVOLUTION: THE CONSTITUTIONAL BOUNDARIES OF THE EUROPEAN
COMMUNITY (1997).
7
Bernhard Schlink, Die Entthronung der Staatsrechtswissenschaft durch die Verfassungsgerichtsbarkeit, 28 DER
STAAT 161 (1989).
8
See generally Holmes, supra note 2.
9
John Gardner, Legal Positivism: 5½ Myths, 46 AM. J. JURIS. 199, 202–03, 218 (2001). Gardner would likely
protest at this point and reply that I have just succumbed to one of the myths obscuring legal positivism, namely
the belief that legal positivism, qua theoretical claim about the condition of legal validity, carries practical
implications for legal interpretation or for allocating the tasks of law application on the one hand and law‐making
on the other. Id. at 222. But see also id. at 218–19. Gardner believes that positivism is “normatively inert,” by
which he means that the position does not provide any practical guidance. Id. at 202. All that legal positivism is
said to accomplish is to provide an explanation for what it means to make a legal claim, by contrast to, for
availability of “detached” descriptions of normative meanings 10 has been seriously
undermined, in one way or another, by hermeneutics 11 or social system’s theory. 12 In a
manner that is relevant to actual legal knowledge, 13 the positivist project appears to be
viable, if at all, only when it is continued—in intellectually distorted form—as a matter of
might. The so‐called originalism rampant in American constitutional law 14 had long been
intellectually exposed for its flaws 15 before becoming increasingly influential on the
bench. 16 It needs to be taken seriously for political reasons, but not as a sound theory of
constitutional interpretation. Aside from such a positivism of might, it appears as though
the project of legal positivism can be carried on, possibly, in the post‐positivist format
created by Ralf Christensen and his former mentor Friedrich Müller. 17 In this guise it
amounts, however, to an inversion of the positivist conception of normative authority,
which is no longer deemed to originate from the norm but from the practice governing its
interpretation. The positivism of norms becomes thereby supplanted by a positivism of
legal language games.

C. Two Readings of Separability

example, claims of moral rectitude or economic efficiency. Id. at 203. Nevertheless, Gardner concedes that even
in such an understanding legal positivism can become a premise of hypothetical imperatives. Id. at 208–11. He
formulates such an imperative for someone who is supposedly under some obligation to find out what the laws of
a certain jurisdiction are. Id. In such a case, legal positivism would have this person look for the sources. Id. This
implies that what is in the sources can be described without regard to moral (or any other) merit. Id. Hence,
even a purely explanatory legal positivist, who is merely interested in uncovering the necessary features of social
practices, would have to admit that legal positivism implies further claims about the point of legal knowledge
when it plays the role of a premise in a hypothetical imperative: “If I want to know what the law is and if legal
positivism is right I had better consult sources and describe what I find therein.”
10
This belief fits even Gardner’s remarkably narrow characterisation of legal positivism according to which valid
law is valid by virtue of having a source. If it were impossible to ascertain what has sprung from the source and
therefore the case that the law would say something new in every case of interpretation we would not be dealing
with a source based system of validity. Id. at 222.
11
See generally JOSEPH ESSER, VORVERSTÄNDNIS UND METHODENWAHL IN DER RECHTSFINDUNG (1970).
12
See generally GUNTHER TEUBNER, LAW AS AN AUTOPOIETIC SYSTEM (1993).
13
See Gardner, supra note 9, at 203, 222 (presenting the project of legal positivism that abstains from carrying
normative implications with regard to whether law ought to be applied or made in singular cases and focuses
exclusively on explicating on the meaning of legal validity).
14
See generally Antonin Scalia, Originalism: The Lesser Evil, 57 U. CIN. L. REV. 849 (1989).
15
For the relevant analysis, see RONALD DWORKIN, A MATTER OF PRINCIPLE 34–57 (1985).
16
For a remarkable example, see generally District of Columbia v. Heller, 554 U.S. 570 (2008).
17
See FRIEDRICH MÜLLER & RALPH CHRISTENSEN, JURISTISCHE METHODIK (10th ed. 2010).
2011]
What is lost in both forms of continuing positivism beyond its end is its original critical
edge. The positivism of might may be a convenient tool for the pursuit of political projects.
Reversing the image of bindingness may be a good way to take comfort from simply going
on with conventional practice. But in neither form is preserved of legal positivism its
quality of being a sting in the flesh of complacent orthodoxy.

Legal positivism’s critical edge is associated closely with its most salient precept, famously
defended by Hart, namely that legal knowledge, in order to be possible, needs to
18
distinguish between law as it is and as it ought to be. This precept, whose import is not
terribly clear, can in turn be read as lending expression to what is widely referred to as the
“separability thesis.” It says that something can be legal even if it is considered to be
immoral 19 and that, therefore, what is morally repugnant can be valid law. Morality is not
a necessary condition of legal validity. 20 The separability thesis extends to other modes of
evaluating norms on their merits, for example, on grounds of either economic efficiency or
comprehensibility. Inefficient norms are just as legally valid as regulations that are too
complex to make any sense. Positivism, broadly understood, is the belief that the criteria
of legal validity are in one way or another self‐standing, 21 and only more narrowly
conceived does this mean that they contain references to social sources.

In my opinion, positivism’s critical edge resides in a certain reading of this general precept.
I should like to distinguish two such readings and suggest that it is to be found in the latter.

According to the first reading, the continuing appeal of legal positivism lies in the promise
of descriptive accuracy. If social sources matter then consulting them enlightens us as to
what the law is, regardless of whether we think that, as a practical matter, we have reason
to go forward with the way it is. Owing to the methods associated with legal positivism
norm addresses gain access to the law as it is as opposed to what they would like it to be.
18
See H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 HARV. L. REV. 593, 615–21 (1958). Gardner,
supra note 9, at 223, would likely put it differently and state in greater proximity to Austin that the source‐based
existence of law is one thing while its moral merit or demerit is another.
19
There is an alternative rendering of the separability thesis saying that necessarily legal validity cannot depend
on the moral merit of norms because the authority of law is possible only if it does not overlap with moral
authority. I guess that this would be Raz’s view, but this should not detain us here. JOSEPH RAZ, ‘AUTHORITY, LAW
AND MORALITY’ IN HIS ETHICS IN THE PUBLIC DOMAIN 226–30 (2d ed. 1995). For a critical analysis that perceives the
issue unresolved in Raz’s writings, see DWORKIN, JUSTICE IN ROBES, supra note 5, at 202.
20
I hope that the latter formulation would be endorsed by Gardner who made insightful critical comments on the
usual characterization of legal positivism as perceiving “no necessary connection between law and morality.”
Gardner, supra note 9, at 222–25.
21
I add in passing that theorists of functional differentiation would say, at this point, that autonomy of the legal
system is manifest in the code legal/illegal. See generally Niklas Luhmann, Law as a Social System, 83 NW. U. L.
REV. 136 (1989); Niklas Luhmann, Operational Closure and Structural Coupling: On the Differentiation of the Legal
System, 13 CARDOZO L. REV. 1419 (1992).
Legal positivism follows the path of sober description by mapping the law “out there”
without overlaying its object with desires or ideals harbored “in here,” i.e., within the
precincts of one’s soul. Legal positivism, thus understood, is the road to truth. The
internal link to descriptive accuracy is preserved even in those versions of legal positivism
whose point is not to guide legal practice but to account for the structure of concepts
explaining the possibility of the existence of shared legal meanings. In this understanding,
a positivistic jurisprudence explores the conditions under which law, as a social fact, can
22
exist as an object.

Descriptive accuracy, however, is only one way of understanding positivism’s most general
precept. Another way consists of engaging in constructive efforts that respond to and
actively counteract the law’s tendency towards self‐idealization and self‐obfuscation.
Arguably, this has been, in the terms of H.L.A. Hart, the “sane and healthy centre” of
Bentham’s positivism, 23 and the same can be said of Kelsen’s project as it is manifest in his
critique of “psychologism” or, more generally, in the scrutiny to which he submitted
central concepts of public law, such as statehood and sovereignty. 24 Bentham’s imperative
theory, for example, is an attempt to clarify in most sober terms that law, far from being
an embodiment of reason, is an expression of the human will and only of human will.
Bentham asserted this view in opposition to how the legal materials are presented in the
interpretations of judges, which, in his opinion, were to be mistrusted as notoriously giving
rise to mystery and confusion. 25 Legal positivism treats the self‐idealizing tendency of the
legal materials with suspicion. Hence, it cannot merely describe the law in the boastful
colors with which it claims to have authority. 26 Law is not readily available as an object of
description. Rather, law must be brought to confess to the grounds and limits that make it
possible.

In this respect, legal positivism is similar to American legal realism. 27 Whereas the latter
has no qualms about setting aside the normative appearance of legal directives on the
22
I concur, on this point, with DWORKIN, JUSTICE IN ROBES, supra note 5, at 214–15, and RONALD DWORKIN, LAW’S
EMPIRE 33–35 (1986).
23
H.L.A. Hart, Bentham and the Demystification of Law, 36 MOD. L. REV. 2, 8 (1973).
24
See generally HANS KELSEN, HAUPTPROBLEM DER STAATSRECHTSLEHRE ENTWICKELT AUS DER LEHRE VOM RECHTSSATZE (1960);
HANS KELSEN, DER SOZIOLOGISCHE UND DER JURISTISCHE STAATSBEGRIFF. KRITISCHE UNTERSUCHUNG DES VERHÄLTNISSES VON STAAT
UND RECHT (2d ed. 1928); HANS KELSEN, DAS PROBLEM DER SOUVERÄNITÄT UND DIE THEORIE DES VÖLKERRECHTS. BEITRAG ZU
EINER REINEN RECHTSLEHRE (2d ed. 1928).
25
Hart, supra note 23, at 3.
26
No legal positivist who is not completely out of her mind would say that God is the author of a constitution
whose preamble states that the constitution was adopted “in the name of God from whom all law originates.”
27
For a classical statement of the demystifying mission of realism, see generally Felix Cohen, Transcendental
Nonsense and the Functional Approach, 35 COLUM. L. REV. 808 (1935). For a useful comparison, see Brian Leiter,
2011]
ground of a comparatively keener interest in exploring how things are actually done, the
more appealing versions of legal positivism attempted to retain law’s normativity by
lending it a morally sobering construction. Rendering law normatively intelligible by
moving beyond its naive, moralistic or ideological self‐presentation has always been the
distinguishing mark of constructivism. 28 Legal positivism is a particularly skeptical member
of this intellectual family. Hence, arriving at law “as it is” as opposed to as it “ought to be”
implicates suspicion vis‐à‐vis the legal materials themselves. It may require constructing
conditions of legal validity of which the participants in legal practice may not have been
aware, such as the Grundnorm or power‐conferring norms explaining the validity of a
judicial decision, which is believed to be law by the judge and her peers on the grounds of
its merits.

If there were a point in doing so, one could argue that Kelsen’s legal theory comprises both
readings of the legal positivism project and that these remain, ultimately, irreconcilable.
But this is not what I would like to explore here. Instead, what I would like to argue is that
the spirit of legal positivism can be rescued for a different age once the first reading of its
29
mission is revealed to fall short from the perspective of the second. In other words, I
would like to demonstrate that the spirit of legal positivism resides in positivism as
constructivism. In this form, it can be preserved even for an age that has come to reject
positivism as descriptivism.

D. The Hartian Persuasion

In order to arrive at this conclusion I am going to take a detour. It is necessary, in
particular considering that I intimated at the outset that legal positivism appears to be
30
dead. But this cannot be right. I must have ignored that there exists a strong school of
legal positivism which is very much alive and kicking. I will refer to this school as the
“Hartians,” taking my cue from the name of the eminent scholar whose written work has
been accepted as the gospel by its members. 31 Membership in this circle—or, put
Legal Realism and Legal Positivism Reconsidered, 111 ETHICS 278, 301 (2001) (providing a discussion that focuses
almost exclusively on the legal positivism of H.L.A. Hart).
28
Rudolph von Jhering, 2.2 DER GEIST DES RÖMISCHEN RECHTS AUF DEN VERSCHIEDENEN STUFEN SEINER ENTWICKLUNG 385
(5th ed. 1899); see generally Stanley L. Paulson, Hans Kelsen’s Earliest Legal Theory: Critical Constructivism, 59
MOD. L. REV. 797 (1996).
29
It is, of course, possible to argue, from the perspective of the first reading, that legal positivism needs to reject
a second reading that is inconsistent with the first.
30
In fact, readers may already wonder whether I have already prepared letters of apology to be sent to members
of Balliol and University College at Oxford.
31
For a highly useful and authentic introduction into the outlook of Hartianism, see Jules L. Coleman & Brian
Leiter, Legal Positivism, in A COMPANION TO PHILOSOPHY OF LAW AND LEGAL THEORY 228–48 (Dennis Patterson ed., 2d.
ed. 2010).
differently, Hartianism as genre 32 —is mediated by five beliefs. First, through joining the
Hartians one becomes a legal positivist. Second, H.L.A. Hart was the greatest legal
philosopher of the twentieth century. 33 Third, the Hart‐Dworkin debate is of central
importance for legal theory (even though there has never been a written exchange
between the main protagonists). 34 Fourth, Hart eventually won the debate, even though
Dworkin would not learn about this defeat before the posthumous publication of Hart’s
postscript in the second edition of A CONCEPT OF LAW. Fifth, intellectual progress in legal
theory can only be made by standing on the shoulders of Hart. 35

Lest I be misunderstood, I would like to clarify that I consider Hartianism a wellspring of
erudite analytic legal philosophy. In other words, studying Hartian scholarship is an
excellent way of taxing one’s brain. This intellectual fitness aspect notwithstanding, I
argue that the Hartians present us with is at best a very lame version of legal positivism. 36
I suspect that they even pride themselves in lacking critical practical edge. 37 Of course,
while there is nothing wrong with producing a good boy/nice girl variety of legal theory, it
is doubtful whether Hartians are not thereby selling the positivist project short.

The Hartian persuasion is culturally inward‐looking. This feature may well be a resonance
of its gestation period. Hartianism, which American scholars most religiously adhere to, 38
gained momentum during the end of the twentieth century when America seemed to be
32
The allusion is, of course, to John Gardner, The Legality of Law, 17 RATIO JURIS 168, 171 (2004), in which law
itself is characterized as a genre.
33
See, e.g., Matthew H. Kramer, Introduction, in THE LEGACY OF H.L.A. HART xiii (Matthew Kramer et al. ed., 2008)
Without denying Hart his share of greatness, this is a bold claim to make, given that Hart competes in such a
contest, questionable as it is, with the likes of Hans Kelsen, Karl Llewellyn, or Carl Schmitt (I do no dare say Ronald
Dworkin).
34
See, e.g., Scott J. Shapiro, The “Hart‐Dworkin” Debate: A Short Guide for the Perplexed, in RONALD DWORKIN 22–
25 (Arthur Ripstein ed., 2007) (providing an example of the centrality of the debate); MATTHEW H. KRAMER, IN
DEFENCE OF LEGAL POSITIVISM: LAW WITHOUT TRIMMING 128–92 (1999) (same).
35
The latter is clearly reflected in Shapiro’s attempt to arrive at a legal theory by answering questions that were
only unsatisfactorily addressed in Harts work. See Scott J. Shapiro, What Is the Rule of Recognition (and Does it
Exist)?, in THE RULE OF RECOGNITION AND THE U.S. CONSTITUTION 250 (Matthew Adler & Kenneth Einar Himma eds.,
2009) (undertaking to amend the house that Herbert built by developing a theory of “plans”). This article is
supposed to answer a number of objections that have been made against Hart.
36
See DWORKIN, JUSTICE IN ROBES, supra note 5, at 188, 198 (finding that the Hartian approach is not a version of
positivism at all).
37
Gardner, supra note 32, at 174–77; see also DWORKIN, JUSTICE IN ROBES, supra note 5, at 211 (providing a related
observation regarding Hartianism).
38
Among the most distinguished members are Jules Coleman, Kenneth Einar Himma, Matthew H. Kramer, Scott J.
Shapiro, and Wilfrid J. Waluchow. I submit, however, that it is difficult to apply the rule of recognition for
Hartianism as an outsider.
2011]
setting the standards for the rest of the world. For Americans of that generation, H.L.A.
Hart was the only legal positivist whose work was somewhat congenial to their own
cultural ways. Therefore, it made much sense for them to seek guidance from him. What
is quite exceptional about Hartianism, nonetheless, is the fact that what is written about,
or in the spirit of, Hart has been quickly turned into an industry whose products are
disseminated in various forms: handbooks, monographs, and a prominent legal
periodical. 39 Even more remarkable than these outward manifestations is the
unprecedented subtlety and analytical rigor with which Hartians not only go about
expounding the work of the master but also develop amendments to the edifice. 40 For
example, the Hartians distinguished between inclusive and exclusive legal positivism, a
distinction that has quickly become part of the standard inventory of legal theory
textbooks. 41 Lest I be misunderstood, I add that the wider circle of Hartians is not
composed of legal positivist alone; even the non‐positivist among them take Hart as having
formulated the most defensible version of this position. 42

The high level of sophistication that Hartian scholarship brings to bear on the
interpretation and defence of Hart ostensibly (or maybe even “ostentatiously”?) 43 exceeds
the subtlety of the master. Since I cannot, for the purposes of this exposition, even
attempt to do justice to the various denominational instantiations of the common faith, I
should like to focus on the work of the eminent scholar whose role among the whole group
is, in a sense, most adequately described as that of the Hartian in Chief. Whom I have in
mind, of course, is Jules Coleman whose work on inclusive legal positivism ranks among the
outstanding contributions to the jurisprudence of our time. 44

39
The important periodical is Legal Theory. The monographs are virtually countless, and the flagship publication
is THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW (Jules Coleman & Scott J. Shapiro eds., 2002).
40
See generally Shapiro, What Is the Rule of Recognition, supra note 35.
41
See, e.g., BRIAN H. BIX, A DICTIONARY OF LEGAL THEORY 123 (2004) (providing the distinction). Shapiro, The “Hart‐
Dworkin” Debate, supra note 34, at 53 (same); supra note 39 (providing references to the literature).
42
For example, Brian Leiter, who is not a legal positivist, believes that, in contrast to Hart’s theory, Hans Kelsen’s
theory is out of step with developments in modern philosophy. Brian Leiter, Michael Gree and Hans Kelsen
Redux, (October 6, 2007), http://leiterlegalphilosophy.typepad.com/leiter/2007/10/michael‐green‐a.html.
43
See HAROLD BLOOM, THE ANXIETY OF INFLUENCE: A THEORY OF POETRY (2d ed. 1997) (discussing covert patricide as
stimulus of literary production).
44
See generally JULES COLEMAN, THE PRACTICE OF PRINCIPLE: IN DEFENCE OF A PRAGMATIST APPROACH TO LEGAL THEORY
(2001) (representing a monographic exposition of Coleman’s legal theory).
E. Conventionalism

When in the course of one of the heated debates of the 1920s Hans Kelsen remarked that
natural law theory merely disguises the “Gorgonian face of power” that one had better
expect to encounter underneath all more idealized appearances of law, 45 he pointed most
drastically to positivism’s sobering mission. Where participants in the legal system would
have us perceive good reasons and sound judgment it is the task of the positivist to
unearth asymmetries of power and choices that are made, potentially, to the detriment of
opponents. In this understanding, the thrust of separability is epistemological as well as
sociological. In eliminating moral criteria from the construction of legal validity, the theory
intends to bring to the fore what truly constitutes law, namely, either the legally
authorized or merely the effective exercise of power. Positivists would not deny that
moral justification is the requisite accompagnato thereto. But it needs to be turned off, as
it were, in order to see what is really going on.

46
No ambition could be farther removed from Coleman’s project. Indeed, at the heart of
his theory lies the attempt to make sense of the seemingly paradoxical claim that it is
consistent with legal positivism to see the validity of at least some legal norms depend on
moral criteria. 47 Coleman defends an extremely accommodating version of “inclusive” (or
“soft”) legal positivism, which would permit as part of the rule of recognition even a
condition saying that laws necessarily have to be defensible from a moral point of view. 48
45
HANS KELSEN, 3 VERÖFFENTLICHUNG DER VEREINIGUNG DEUTSCHER STAATSRECHTSLEHRER 54, 55 (1927).
46
Since Hartians do not share this ambition they neglect the critical import of constructions arrived at by more
advanced forms of legal positivism. Repeatedly, Kelsen is reprimanded by Shapiro for arriving at a
counterintuitive concept of the legal norm that conceives of it as primarily addressed to the law‐applying official
whose duty it is to impose a sanction. HANS KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY 26–30 (trans.
Bonnie Litschewski Paulson & Stanley L. Paulson, 1992); Scott J. Shapiro, The Bad Man and the Internal Point of
View, in THE PATH OF THE LAW AND ITS INFLUENCE: THE LEGACY OF OLIVER WENDELL HOLMES JR. 199, 204 (Steven J. Burton
ed., 2000). By insisting against Kelsen that Hart’s “puzzled man” is to be guided by legal rules Coleman ignores
the demystifying potential of Kelsen’s construction, which reveals that legal systems can operate with disregard
for how laypersons understand norms. H.L.A. HART, THE CONCEPT OF LAW 40 (2d ed. 1994). I wonder whether
Kelsen’s account is not sociologically more accurate than a legal theory that takes intuitions about guidance by
laypersons uncritically for granted. Not only is it the case that the puzzled man usually is the muzzled man; the
insistence on the “puzzled man” obtaining guidance from rules is inconsistent with Hart’s belief that law exists
only where there is unity of primary and secondary rule. The “puzzled man” resides the pre‐legal sphere
governed by primary rules.
47
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 67, 109–10. The paradox has been duly noticed by
DWORKIN, JUSTICE IN ROBES, supra note 5, at 189.
48
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 112, 126. Of course, Coleman needs to struggle at this
point with re‐establishing the authority of law vis‐à‐vis moral authority. He attempts to do so by reintroducing
the power of someone to establish for ordinary folk what, according to moral standards, is law. He thereby likens
his position again to “exclusive” legal positivism. (130, 141). Legal positivists, who, as Gardner, note 9 at 200,
reminds us, believe that all legal norms are posited and hence come into this world as a result of acts, find it quite
difficult to make sense of the idea that some laws might be void per se, regardless of whether the voidness is
2011]
It cannot come as a surprise, then, that the core of legal positivism is not believed to be
captured by separability, however understood, but rather in the puzzling claim, incidentally
attributed to all contemporary legal positivists, 49 that the criteria of “legality” 50 are a
matter of conventional social facts. 51 Conventional criteria for the set of social facts that
we call law lay the ground for the law’s existence. They are applied in a customary judicial
test of legal validity. 52 According to Coleman, this belief defines the core of legal
positivism. 53

The criteria of legal validity are believed to be components of what Hart introduced under
the name of “rule of recognition.” 54 As a convention, this rule depends on being practiced
in order to exist and to be authoritative for legal officials. 55 The existence of this rule
explains how legal systems are possible without thereby indicating why any such system is
also desirable. 56 The rule exists, socially speaking, if and only so long as it is practiced. The
rule is practiced if and so long as it is applied from an internal point of view. 57 This means
alleged to follow from immorality or illegality. See Hans Kelsen, Reine Rechtslehre (2d ed., Vienna: Deuticke,
1960) at 280.
49
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 68–69, 75.
50
By “legality” Coleman means what is to be counted as law is therefore capable of partaking of legal validity. See
also Shapiro, What Is the Rule of Recognition, supra note 35, at 240.
51
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 152, 161. It may bear emphasis that the criteria for what
might partake of legal validity are different from the criteria that account for the existence of a legal system, such
as the unity of primary and secondary rules or the existence of legal officials. Gardner, The Legality of Law, supra
note 32, at 170.
52
53
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 161.
54
55
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 77. Coleman believes the rule of recognition to be a duty‐
imposing rule, however, only on legal officials, whereby the duty itself is only incidentally related to law, for it
arises from participation in a common enterprise and is a special case of an associative obligation (77, 85, 95, 97,
159‐160). The duty imposed by the rule of recognition can therefore never be a legal obligation and is not
addressed to “ordinary folk” (139).
56
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 71, 93, 118; HART, THE CONCEPT OF LAW, supra note 46, at
257.
57
See COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 76, 86 (noting that the sloppy characterisation of a
social rule as a combination of a pattern of convergent behaviour plus internal point of view is flawed). The flaw
goes back to Hart. HART, THE CONCEPT OF LAW, supra note 46, at 255. One cannot, as a matter of judgement,
ascertain convergent behaviour without adopting a point of view in the most elementary sense of discriminating
between and among instances of a pattern. In this respect, a point of view is adopted vis‐à‐vis all others who are
also engaged in making out regularities without being actively engaged in bringing them about. The difference
between such a point of view, which is manifest in the exercise of what Kant would have called “reflexive
judgement,” and what Hart calls the internal point of view, which is relevant for social rules, lies in the fact that
that those participating in its application adopt a critical reflexive attitude towards
compliance and non‐compliance. 58 Convergent behavior would be socially beyond
recognition as rule‐following if it were not perceived as implicitly constituting itself as
instances of following the rule. 59 This explains, I add in passing, why law is constituted by
legal knowledge, a point to which I shall return below.

If the law, as a system, 60 avails of a center, then this place is occupied by the rule of
recognition, 61 for “it makes determinate which rules bear the mark of legality” and
“creates a duty for a certain class of individuals—officials—to evaluate conduct under the
set of primary rules that bear that mark.” 62 The social constitution of this rule is invariably
circular. The officials who apply it and are constituted by it in this capacity must be guided
by it 63 in order to be subject to its authority. 64 What the rule means in fact is affected by
the elements of the pattern are themselves treated as instances of rule‐following. They are distinct from mere
occurrences for which one tries to reconstruct a pattern. Hence, a point of view is applied not towards others
who are also engaged in trying to make out a pattern themselves but to whatever appears to be like an instance
of the pattern. Potential instances of the pattern are interpreted to be instances of rule‐following. The critical
reflexive attitude is directed at the claim that is attributed to them, namely, the claim to be cases of following a
rule. Social facts are thereby read as involving implicit claims whose validity is put to the test. Only thus
understood, the following statement by Coleman makes sense: “A social rule exists when convergent behaviour
is conjoined with a critical reflexive attitude towards that behaviour. The critical reflexive attitude is the internal
point of view.” COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 82.
58
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 88–89; HART, THE CONCEPT OF LAW, supra note 46, at 86.
59
The rule exists only if there is widespread acceptance of the internal point of view. COLEMAN, THE PRACTICE OF
PRINCIPLE, supra note 44, at 83, 153. I spare readers a discussion of the delicate question whether the
reconstruction of social rules from an external perspective merely suspends the application of the critical reflexive
attitude towards what it treats as instances of the rule but nonetheless extends it to those engaged in the same
hermeneutic exercise. Alas, the application of a critical reflexive attitude on the part of those engaged in an
external and theoretical description of a social practice is absent in the otherwise highly useful discussion by Scott
J. Shapiro, What Is the Internal Point of View?, 75 FORDHAM L. REV. 1157, 1160–61 (2006).
60
The belief that law is a system was most certainly held by Hart. Gardner, The Legality of Law, supra note 32, at
170.
61
Shapiro, What Is the Rule of Recognition, supra note 35, at 246.
62
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 139. The rule of recognition is only a necessary, but not a
sufficient condition for law, which also requires obedience by “ordinary folk” for its existence. Id. at 76.
63
It is not infrequently said that the rule of recognition must be “accepted” by officials. Shapiro, What Is the
Internal Point of View?, supra note 59, at 1159; Shapiro, What Is the Rule of Recognition, supra note 35, at 245.
But this broad formulation obscures an important difference. Officials must accept that their behavior has to be
intelligible and defensible as “rule‐following.” But it is not necessary that they accept the rule substantively. See
HART, THE CONCEPT OF LAW, supra note 46, at 255 (noting that rules must be accepted (merely) as guides of conduct
and criticism). See Shapiro, The Bad Man and the Internal Point of View, supra note 46, at 202 (getting it right by
saying that officials must be “committed to following” the rule). But they need not accept it as such and not, of
course, on its merits. Participants in the practice are therefore able to answer questions about how one plays by
the rules. They are not necessarily able to answer questions about whether playing by the rules is useful and
good. With this distinction, many conceptual monstrosities can be avoided, for example, the “detached attitude”
2011]
their conception of the rule’s meaning, which they imagine to be independent of their
arriving at such a conception. Their voluntary compliance and their critique of the
behavior of others must be informed by something that is independent of, and distinct
from, various individual interpretations or manifestations of the rule in single cases. 65
Authoritativeness as a rule is possible only if the rule is more than the sum total of its
applications and therefore, in a sense, something over and above them. 66 In Coleman’s
words, this means that “[i]n order for . . . behaviour to constitute a practice in the relevant
sense, it must reflect a shared grasp of the rule . . . .” 67

F. The Elusive Convention

Expositions that are intended to state what the rule really requires are arrived at from the
internal point of view. But reasonable people may come up with reasonably different
answers to this question. Arguably, the rule of recognition of European Union law says,
among other things, that fundamental rights emerge from the common constitutional
traditions of the Member States. Conceivably, it can be disputed whether this component
of the rule means that the existence of a right in the constitutional law of two Member
68
States is sufficient to constitute a common constitutional tradition.

Coleman would explain that the convention exists as long as officials agree on the content
of the rule in the abstract 69 even though they may disagree about its application. He
claims the following:

Judges may agree about what the rule is but disagree
with one another over what the rule requires. They
or the relative moral authority of the legal point of view Shapiro, What Is the Rule of Recognition, supra note 35,
at 259. Nonetheless, Patterson insists correctly against Hartians that what has to be accepted as rules by officials
in order to be in the position to arrive at judgments about rule‐following are “forms of argumentative appraisals,”
such as methods of interpretation. Dennis Patterson, Explicating The Internal Point of View, 52 SMU L. REV. 67, 73
(1999). Interestingly, Patterson, in turn, is not heeded of the fact that, just like explorations of the rule’s merit,
appeals to sound methods of interpretation mark the point at which mere conventional practice is exited. I
surmise that these matters could be further elucidated from a background that is conspicuous by its absence in
Hartianism, namely action theory. It is possible to conceive of participants in the practice as improvising actors in
the sense envisaged by Velleman. See generally J. DAVID VELLEMAN, HOW WE GET ALONG 12–14 (2009).
64
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 78, 82, 95, 134.
65
66
67
Id.
68
Case C‐144/04, Werner Mangold v. Rüdiger Helm, 2005 E.C.R. I‐9981.
69
See DWORKIN, JUSTICE IN ROBES, supra note 5, at 192–93 (discussing the convention).
could not disagree in every case or even in most cases,
since such broad and widespread disagreement would
render unintelligible their claim to be applying or
following the same rule. 70

This statement is very confusing because it involves an equivocation. The token is what
they can agree on (for example, “All are to be treated equally”), but what matters, as
Coleman’s reference to “widespread agreement” indicates, is the type (i.e., who counts as
equal). It is because of the equivocation of token and type that Coleman can claim that
“some disagreement about a rule’s requirement is not incompatible with the rule’s
71
conventionality.” Token conventionality, however, is a useless guide. It cannot
coordinate conduct, which explains why type conventionality is what Coleman is truly after
by implicitly referring to agreement in the majority of cases. Not that we should find it
disturbing that Coleman cannot quantify here. The major point is that in controversial
cases mere appeals to conventions cannot do the work. 72 The critical reflexive attitude is
directed at how rule‐following is done and not what the rule is about. When a community
has been thrown out of agreement about how things are done a convention is no longer of
any avail.

Coleman is aware of the problem and claims that the rule of recognition is a result of
“ongoing negotiations” and “a framework for bargaining how to go on” 73 or, more
generally, a shared cooperative activity, 74 which involves reason‐giving. Coleman believes
to believe that the ground of legal validity is the rule of recognition. But what he believes,
in fact, 75 is that there is a practice of exchanging arguments in which claims are made, as a
conventional matter, that x is law because it has been previously recognized as law. But
this does not mean that the connection between x and prior cases is governed by a
convention. Merely the appeal to the purported connection is conventional. Coleman
ends up reducing the conventionality of the rule of recognition to the conventionality of
appeals to its components. We may find it easier to talk to one another when, we refer, as
a matter of course, to the constitution. This does not mean, however, that interpretations
70
71
Id.
72
See DWORKIN, JUSTICE IN ROBES, supra note 5, at 190 (“Convention is built on consensus, not disagreement. . . .
When a group of people disagrees about what behaviour is required or appropriate, it seems odd to say that they
have a convention that decides the issue.”).
73
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 99–100.
74
Id. at 96–97.
75
See STANLEY L. FISH, DOING WHAT COMES NATURALLY 471–502 (1989) (formulating a similar objection against Hart).
2011]
of the constitution are thereby governed by a convention. Coleman conflates “we
conventionally appeal to x” with “x is determined by convention.” 76

G. The impasse

In order to rescue conventionality, Coleman drives a wedge between the rule itself and the
controversies over what the rule requires in single cases. 77 Because the rule of recognition
is supposed to be a conventional rule any conception of it, even though invariably involved
in its application and evolution, must be intrinsically self‐effacing. The convention is
supposed to exist even where disagreement has arisen over what it truly requires. 78 It can
only exist, therefore, when it is possible to sever it somehow from accounts of its meaning.
Paradoxically, however, the severed convention cannot exist unless a critical reflexive
attitude is considered to be manifest in its constitution.

It is a convention that fundamental rights of EU law have their source in the common
constitutional traditions of the Member States. It is left open, however, what it takes to
have such a tradition. What matters are applications that flow from the internal point of
view in one case or the other, but not how they are rendered intelligible as instances of
rule‐following from this perspective. They ought to be looked at as mere instances of
practice.

But, if applications are dissociated from how they can be accounted for as applications of a
79
rule, they can also no longer count as instances of practicing a rule. A rule exists only as
reflected in its application and from one application to the next from the perspective of the
connections that the present establishes with the past. The past can govern the present
only if the present allows itself to be governed by the past in the future. 80

Conventionalism encounters an impasse here. According to Coleman, the content of the
rule of recognition must emerge from a “pattern” 81 of behavior that one observes, as an
76
See DWORKIN, JUSTICE IN ROBES, supra note 5, at 195–96 (observing correctly that shared co‐operative activities
are not necessarily conventional).
77
See generally COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44.
78
Id. at 116; See DWORKIN, JUSTICE IN ROBES, supra note 5, at 190 (providing perceptive critical observations).
79
Here is how Coleman states the matter: “It is the shared attitude toward the pattern of behaviour that
constitutes the fact that the behaviour is governed by a rule.” COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at
79. The formulation suggests that there is a pattern which is somehow joined by the internal point of view. But
this cannot be the case. Elements of a pattern can only be relevant for the rule when it is integral to their
meaning to be followings of the rule. Supra note 57.
80
81
Id. at 91.
official, in others and oneself. The pattern says that in order to establish the existence of a
fundamental right of European Union law, appeals must be made to a common
constitutional tradition. Unfortunately, as we have seen, the mere appeal to the
convention does not give one the convention itself. Reasons of political theory must not
be decisive either, for they would introduce considerations of the merits of one or the
other substantive conception of that rule, as a result of which legal validity would no
longer be a matter of social fact. 82 Likewise, considerations regarding the right methods of
interpretation would, by virtue of what is claimed by them, transcend the purview of the
conventional.

H. The Form of the Rule of Recognition

The discussion has revealed that it is difficult to imagine how, in cases of doubt, the
internal aspect of rules (the critical reflexive attitude) can be worked into conventions with
regard to the rule’s substance. Because questions of interpretation cannot be settled with
regard to settled practice, conventionality is likely to be transcended by appeals to
methodological accuracy or moral correctness.

It cannot be ruled out, however, that the question admits of a more satisfactory solution
from a social perspective. The distinction between the substantive and the social
dimension is straightforward. What is of interest about rules, substantively considered, is
what they require. What matters about them, socially, is whether their application would
83
be supported by consent.

If the pattern itself matters, and not a statement of what the rule is over and above its
applications, then all that is needed in order to hit the convention is to trigger nodding
among one’s peers. Vis‐à‐vis “ordinary folk” (Coleman’s genteel characterization of lay
persons), who might react with bewilderment when they learn that two Member States’
constitutions are sufficient to constitute a common constitutional tradition, one stands
vindicated as long as the peers, when confronted with such a finding, nod and mutter “This
is what we do.” Is not this, if anything, the point of having a convention? “This is what we
do” and “I do as all others do” comprise the form of stating the rule’s social dimension
84
from an internal point of view. Had Coleman studied George Herbert Mead, it would
82
Id. at 161.
83
NIKLAS LUHMANN, 1 RECHTSSOZIOLOGIE 99, 105 (1972).
84
Such social perspectives are still consistent with the standpoint that Shapiro characterizes, misleadingly, as
“rule acceptance” or “internalization.” Shapiro, What Is the Internal Point of View?, supra note 59, at 1159;
Shapiro, The Bad Man and the Internal Point of View, supra note 46, at 200, 208. It is misleading, for what
matters for the internal point of view is one’s performance as a competent rule‐follower. Even—or maybe even
in particular—hypocrites can be perfect rule‐followers. I mention in passing that Hart’s distinction between the
external and the internal aspect of rules is remarkably difficult to square with his conventionalism. H ART, THE
CONCEPT OF LAW, supra note 46, at 55–56, 86. The external aspect of rules captures general regularity of conduct,
2011]
have also occurred to him that, thus understood, conventional rules are means of
exercising social control over individuals. 85 Conventionalism would arrive at the
conclusion, hence, that the right determination is that which would have been arrived at
by any other official whose intellectual demeanor exhibits all the marks of what it takes to
play the game skillfully. 86

Remarkably, the comparison usually drawn in this context with regard to (mere)
coordinating conventions 87 creates the impression that this is the immediate and natural
rendering of the internal point of view whereas more elaborate substantive conceptions of
what is takes to apply a rule correctly are mediated by interpretation and, for that reason,
more artificial. But the contrast is misleading. Mere coordinating conventions, such as the
convention to drive on the right side of the street, are not likely to be susceptible to much
interpretative elaboration, while elements of a purported rule of recognition are. Hence,
the conventionalism of nodding peers is possible only when broader interpretative or
justificatory appeals are avoided or suppressed against the backdrop of some permissive
consensus which signals to peers that peers will react with “yeah” to whatever is done by
peers simply because they are who they are. It would be wrong, therefore, to assume that
peer‐group conventionalism is logically or historically prior to the interpretative
elaborations of any justificatory appeal to the rule of recognition; rather, conventionalism
is based on denying them relevance for a practice in which peers benefit from being able
which is accessible even to an outside observer. The internal aspect is something that can be had by a rule “in
addition” to the external aspect. Id. at 55. It is manifest in people “having views” about what it takes to play by
the rules. Id. at 56. Because in the case of a convention these views have to have observance of a pattern of
conduct as their point (“Now, you do as all others do!”) the external aspect is of overwhelming relevance for
engaging the critical attitude from the internal point of view. Id. at 55. What would it take to “have views” as
regards conventions? One would have to come up with educated guesses about how peers will be judging one’s
application of the rule, for if that were not decisive it would not be a convention (“This is what we do”).
Conceivably, a lawyer with training in American constitutional law would approach an equal protection problem
by first asking which level of scrutiny has to be applied. However, if she were to do so in the context of Austrian
constitutional law she would be reproached by her Austrian peers claiming correctly that this is not the way
things are done around here (“This is not what we do”). The internal point of view states what can be observed
from the external point of view and asserts it as a matter of right. The difference between the internal and the
external aspect is not whether one anticipates reactions from a pattern or, even more intelligently, one
anticipates how peers believe all other peers to believe the pattern to be projected into the future, but whether
one does it as a member of the group or as an outsider. Only as an insider you get away with saying “This is what
we do” and do not have to worry much about doing as all others do. Then one is in a position to use as a reason
for decision what one anticipates to be the reasoning of all others. Id. at 102; CARL SCHMITT, GESETZ UND URTEIL
(1912) (reconstructing this exact conventionalism).
85
GEORGE HERBERT MEAD, THE PHILOSOPHY OF THE PRESENT 190 (A.E. Murphy ed., 1932). Had Shapiro read Mead he
would have likely explicated his “plan” theory of legal practice in the terms of a “game.” GEORG HERBERT MEAD,
MIND, SELF, AND SOCIETY FROM THE STANDPOINT OF THE SOCIAL BEHAVIOURIST 153–54 (1934).
86
SCHMITT, supra note 84, at 71.
87
Hart believed the rule of recognition to be a coordinating convention. For a critical analysis, see Andrei
Marmor, Legal Conventionalism, 4 LEGAL THEORY 509 (1998).
to claim that what they do is endorsed by others as epitome of what we do. This is a quite
artificial arrangement.

I. Going with the Crowd

The point of conventionalism is to see emerge from already existing common activity an
obligation to carry on with it. 88 The manner in which this activity is accounted for, from
the internal point of view, avoids broader justificatory appeals. It is accepted by officials as
a social rule for a variety of reasons, among which might figure the interest to earn a living,
the desire to impress one’s spouse or, according to Hart, “the mere wish to do as others
do.” 89 The rule is accepted as a standard of conduct from an internal point of view that is
indifferent to the substantive reasons of universal acceptability. This is written on the face
of statements of the “this is what we do” variety, for, as had been famously recognized by
Wittgenstein, substantively they point to “something indeterminate” (etwas
Unbestimmtes). 90 If they did not they would merge into conceptually more refined
accounts of the underpinnings of what we do. Therefore, the criteria relevant for
exercising the critical reflexive attitude need to focus on something else, which might
explain how an intellectually reticent practice can continue and arrive at a synthesis over
time.

The key to finding the answer is that what one does matters less than who one is. In other
91
words, the practice is easily possible under asymmetrical conditions. While one group,
call them the “boys”, calls the shots and gets away with saying “This is what we do” the up
and coming or the weaker members of the group, call them the “pack,” adapt, resort to
ducking and follow the maxim “I do as others do.” This pair of maxims is the essence of
conventionalism. 92 Interpretations of the rule are determined by what members of the
pack guess will trigger nodding on the part of the boys because it reflects their intuitive
88
89
HART, THE CONCEPT OF LAW, supra note 46, at 203, 257.
90
LUDWIG WITTGENSTEIN, ÜBER GEWISSHEIT 125 (1984). Coleman underscores that the internal point of view towards
a rule is different from the application of the reasons underpinning the rule in the first place. COLEMAN, THE
PRACTICE OF PRINCIPLE, supra note 44, at 89.
91
It is quite amazing that Hart and the Hartians always imagined the “group” to be egalitarian. How can one
assume that in the legal system—vis‐à‐vis courts—criticism is taken seriously or even to be legitimate? See
DWORKIN, JUSTICE IN ROBES, supra note 5, at 196 (hinting, quite correctly, at the fact that it is empirically doubtful
whether the relation among members of the United States Supreme Court can be described as a shared
cooperative activity). Contra Shapiro, What Is the Internal Point of View?, supra note 59, at 1164.
92
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 91–92.
2011]
grasp of the rule. 93 Consequently, the rule of recognition can be a conventional rule only
as long as the pack does not speak up, at any rate not too frequently, and reacts to the
boys’ piqued rejoinders in an apologetic way (“Thank you, now we have got it!”). 94 The
type of normativity that goes into creating the obligation to go on reflects the desire to
follow the crowd. 95 Going along may carry the reward of being one of the boys tomorrow
and of occupying the position, eventually, where one is free to assert, “This is what we do.”

J. From Addition to Quaddition

At this juncture, it is possible to appreciate more fully what it means that the rule of
recognition is a social rule. What matters about rule‐following is that the activity in
96
question is a specimen of social action and not some intellectual operation mastered in
the chambers of court. How one gets along with others is more important than what one
does. Being a proficient member of the group that Hart and Coleman refer to as “officials”
requires social skills, and they are that which is relevant from the internal point of view.
One does not openly criticize high courts for reasoning mistakes. One does not raise
embarrassing “theoretical” questions to which no one has an answer. One tries to go
along even when one cannot make sense of what others do. These matters reflect, at any
rate, the most elementary experiences that are made in the course of being socialized into
a member of the pack. In the eyes of students, professors are boys (even if they are
female). What students attempt to do, using their social internal point of view, is imagine
93
This is not to say that members of the pack take their guidance from boys directly: They apply the rule as they
believe it to emerge from the boy’s internal point of view. Shapiro, The Bad Man and the Internal Point of View,
supra note 46, at 206.
94
This desire may be grounded in the expected reward to become bigger than one is when one is going with the
crowd. FRIEDRICH NIETZSCHE, THE DAWN OF DAY 32–34 (trans. J.M. Kennedy, 1964).
95
It is quite remarkable that in his critique of Austin’s concept of legal obligation Hart shifts the focus from
avoiding coercion towards avoiding rebuke for not going along with the crowd. See HART, THE CONCEPT OF LAW,
supra note 46, at 84 (“Rules are conceived and spoken of as imposing obligations when the general demand for
conformity is insistent and the social pressure brought to bear upon those who deviate or threaten to deviate is
great.”). It is consistent with this understanding of obligation that the internal aspect of rules rests on the
complementary maxims of “This is what we do” and “I do as others do,” respectively. There is no reason, of
course, why one would not be “obliged” by social pressures of conformity even if they were not backed with
coercion. The critique that Hart has of Austin can be easily carried over to Hart himself. Why should the prospect
of being shunned and rebuked by, say, members of the Tea Party Movement give rise to an obligation on my part
to join the ranks of the radical right even if the large majority of Americans were to support the movement? What
matters for the application of the rule of recognition is not the acceptance of the rule but rather the social skills
necessary to pass as member of the pack. Any real acceptance of social rules as obligatory would have to be
based, by contrast, on the desire to be part of the group, for it is through identification with the group that one
can come to consider deviation from its norm as a threat to one’s identity. CHRISTINE A. KORSGAARD, THE SOURCES OF
NORMATIVITY 102 (1996).
96
See MAX WEBER, SOZIOLOGISCHE GRUNDBEGRIFFE 42 (5th ed. 1981) (explaining this activity as an action that is in
some manner heeded of the action of others).
what conceptions of rule and doctrines the professors might have in mind when talking
about cases. Thereby they arrive at guesses about the professor’s substantive internal
point of view from their social internal point of view. This is what they do because they
have chosen to do as others do. The only interesting part of this story is that they need to
make sense even of professors who, using Kripke’s famous example, mistakenly assume to
carry out additions while they in fact perform “quadditions.” 97 Professors who believe in
their own brilliance are often indeed incoherent. Students, nevertheless, need to make
sense of their conduct as instances of a rule and trail their conduct in order to perform
successfully in front of them an operation having the form “This is what we do.”

The undefeated champions of quaddition are courts. In fact, having them occupy this
position is the essence of common law. The idea of the rule of recognition as social rule
matches this situation. The court says that two Member States constitutions are sufficient
to constitute a constitutional tradition. In such a situation, it would be wrong to criticize
the court for either having misapprehended the law or for engaging in judicial legislation.
Whoever is guided by the rule of recognition is also governed by it. Once the decision has
been made, governance impacts guidance. The decision creates a new situation. The law
98
is no longer what it had been before. All efforts to come up with substantive renderings
of the rule from an internal point of view are thereby thwarted from the social internal
perspective.

This explains why the common law tradition remains at odds with the constructive legal
scholarship. A new case changes everything. This is what we do. It should not come as a
surprise that the intellectual malaise brought about by such a system comes wrapped up in
rhetoric suggesting that persistent confusion is the mere surface manifestation of how the
99
law accumulates experience and increases its wisdom.

K. Idealization and Disguise

The pair of statements “This is what we do” and “I do as others do” indeed provides access
to two different perspectives of the internal point of view whose application enters into
97
See generally SAUL A. KRIPKE, WITTGENSTEIN ON RULES AND PRIVATE LANGUAGE: AN ELEMENTARY EXPOSITION (1982)
(discussing rule‐following and introducing the example of someone who adds numbers and instead of adding one
after the other at a certain points begins adding more than one number). From the perspective of an observer
who tries to infer the rule from behavior, this rule turns out to be not the rule of addition but of something
slightly different, for example, “quaddition,” which does not use “plus” but “quus” as its operator. Id.
98
Gardner, The Legality of Law, supra note 32, at 175.
99
For an extremely useful account of the mentality of the Common Law tradition, see GERALD J. POSTEMA, BENTHAM
AND THE COMMON LAW TRADITION (1986). His account of the common law mentality has recently been
complemented by the following articles: Gerald Postema, Classical Common Law Jurisprudence (Part 1), 2 OXFORD
UNIV. L.J. 155 (2003); Gerald Postema, Classical Common Law Jurisprudence (Part 2), 3 OXFORD UNIV. L.J. 1 (2003).
2011]
the constitution of the rule of recognition as a social rule. The internal point of view is, to
repeat, the “important capacity . . . to adopt a practice or pattern of behavior as norm.” 100
The way in which the boys use the rule brims with self‐confidence: “This is what we do,”
“This is how it’s done,” “Here you have it.” The critical reflexive attitude exclaims, “Don’t
you dare.” Members of the pack, by contrast, use a more cautious form: “They are out to
lunch, but perhaps that is what they mean.” The critical question is: “What would I say if
were one of them?” Of course, the relation between these forms is dialectical. Brimming
self‐confidence may be dampened by the concern that too much audacity risks losing one’s
boyhood. One must not create an opportunity to be easily dismissed as senile or whacky.
On the other hand, one could never advance from the pack to the boys if one never tried
one’s hand at acting like one of them.

Interestingly, Hartian conventionalism hides from us these simple truths about how social
conventions work. It misses important points about the conventional existence of law, for
example, what it takes to elaborate the rule of recognition’s social nature. Such a view of
the internal aspect would indeed reveal that it is empowering to play by rules that have
been laid down by others and not to ask too many questions, in particular not those
questions that boys would be too embarrassed to answer; that it pays to be cooperative
and to show off one’s smarts, but not in an impudent way; that it is advantageous to
profess belief in the greatness of the legal enterprise; that one had better dodge
substantive constitutional issue in the course of Supreme Court confirmation hearings (the
Bork problem).

Remarkably, Coleman’s eventual rendering of the rule of recognition, qua practice, as a
101
form of shared cooperative activity does not include any discussion of the
complementarity of “This is what we do” and “I do as others do.” On the contrary, his
account of practical activity, for which he draws on Michael Bratman, merely arrives at a
more diffuse and idealizing characterization of what is essentially an asymmetrical
relationship:

[E]ach participating agent attempts to be responsive to
the intentions and actions of the other . . . . Each seeks
to guide his behavior with an eye to the behavior of
the other, knowing that the other seeks to do
102
likewise.

100
101
Id. at 96.
102
Id.
What is eclipsed here is that people go about doing things differently depending on their
position within the legal system or scholarly discourse. 103 What this view ignores is that
substantive applications of the rule of recognition are over‐determined by applications
informed by social status.

L. The Negativity of Rules

Legal knowledge, substantively considered, comes into existence in a form in which it is
encumbered and distorted by power. Coleman does not address this. Instead, he
unwittingly trails practiced legal knowledge’s tendency to produce idealized descriptions of
itself. His theory thereby falls victim to its object. Because conventionalism does not take
into account the context from which it emerges and therefore ends up producing a
distorted image of its own operation, it makes sense to treat it as a symptom of certain
intellectual situation. As a symptom, it both conceals and reveals legal knowledge in a
104
state in which theory has been exiled from practice. It even endorses this exile position.
The theoretical reflection of legal practice inadvertently and indirectly reflects the
repression of reflection within that practice.

Unwittingly, nonetheless, his analysis confesses to the prevalence of the social over the
substantive dimension of following rules. Coleman’s attempts to explain what might
account for the substantive consistency of applications of the rule of recognition
inadvertently reveal the social dimension. Coleman correctly claims the range of
convergent behavior alone cannot determine the semantic content of the rule of
105
recognition; at the same time, however, the behavior is said to supposedly “fix” the
106
rule. It would have been interesting to learn who is engaged in doing the fixing. The
rule exists only when it is practiced. It is manifest in certain paradigmatic instances of
agreement even if the participants would find it impossible to agree on the reasons
underpinning an agreement from the internal point of view. It is in this vein that Coleman
suggests that concepts, in order to function as concepts, do not require criteria but just
agreement on paradigmatic cases. 107 It’s the boys, again. 108 This is what they do even
103
I add, in passing, that had Coleman taken the non‐idealizing description into account he would have had more
difficulty claiming that this practice is pregnant with obligation.
104
See, most candidly, Gardner, Legal Positivism: 5½ Myths, supra note 9, at 203.
105
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 99. Interestingly, Coleman believes that differences
regarding the content of a rule are different from differences concerning their application. While this is true for a
difference of agreement over what a rule says on its face, which can be settled by looking it up in the relevant
rule‐book, it is completely implausible to assume that the question of whether two Member States constitutions
are sufficient to constitute a “common constitutional tradition” does not affect the content of the rule. I find it
difficult to imagine how Coleman would believe that.
106
Id. at 80–81.
107
Id. at 155.
2011]
when it is difficult to make out for the pack what the boys agree on in more general terms.
The grasp of the rule cannot be exhaustively articulated in propositional form. 109

However, spelling out the social dimension of the conventional practice by saying what
following a rule means substantively distorts the relation between meaning and practice.
If, as Coleman rightly asserts, what it takes to follow or to apply a rule correctly cannot be
articulated in propositional form it needs to be concluded that conventional practice is
constituted by remarkable negativity. 110 Conventions are socially manifest in patterns of
convergent behavior. Patterns, when seen internally, are composed of self‐reflexive claims
alleging to be faithful to the rule. 111 What is casually referred to as “patterns of behavior”
is composed of acts that carry implicit self‐interpretations. This is the point of the internal
point of view. Whatever is conceived of as candidate of rule‐following is potentially
susceptible to critique for being out of line. 112

From the perspective of a common practice, elements of a pattern come to this world as
critical of themselves. The basis of self‐critique is that which cannot be articulated in
propositional form. It is something indeterminate 113 that is given articulation when
attempts are made to explain the point of rules. The respective accounts develop what
lends authority to the practice governed by rules. Dworkin was right when he, unwittingly,
recognized under the name of “interpretative practice” 114 what Hegel has long before him
called spirit. 115 The latter stands for a reflective form of life. It integrates into itself
practices for reassuring and examining the authoritativeness of reasons that are taken to
be authoritative by their participants. Owing to its negativity, spirit is a product of itself. 116
108
A more sociologically perceptive description of this situation is given by Sunstein in his theory of “incompletely
theorised agreements.” CASS R. SUNSTEIN, LEGAL REASONING AND POLITICAL CONFLICT 46–54 (1996).
109
110
See ALEXANDER SOMEK, RECHTSSYSTEM UND REPUBLIK: ÜBER DIE POLITISCHE FUNKTION DES SYSTEMATISCHEN RECHTSDENKENS
343 (1992).
111
I am aware that Wittgensteinians would at this point say that Wittgenstein insisted that there is a following of
the rule that is not an interpretation. See, for example, JOACHIM SCHULTE, WITTGENSTEIN: EINE EINFÜHRUNG 160–61
(1989). It never occurs to them that Ludwig might have been wrong about this.
112
See generally PETER WINCH, THE IDEA OF A SOCIAL SCIENCE AND ITS RELATION TO PHILOSOPHY (1958).
113
For a profound analysis, see GERHARD GAMM, FLUCHT AUS DER KATEGORIE: DIE POSITIVIERUNG DES UNBESTIMMTEN ALS
AUSGANG DER MODERNE 140–42 (1994).
114
See DWORKIN, LAW’S EMPIRE, supra note 22, at 48; DWORKIN, JUSTICE IN ROBES, supra note 5, at 11–12.
115
TERRY PINKARD, HEGEL’S PHENOMENOLOGY: THE SOCIALITY OF REASON 64 (1994).
116
ROBERT PIPPIN, HEGEL’S PRACTICAL PHILOSOPHY: RATIONAL AGENCY AS ETHICAL LIFE 65 (2008).
It builds itself, as it were, from the indeterminacy into which practice flows out. 117
Conventional legal practice, by contrast, even though possible, 118 simply is mindless
second nature. Reflection is limited to asserting, “This is what we do” and to obeying the
maxim, “I do as others do.” The question of authoritativeness does not arise.

M. Beyond Convention

It is not for bad reasons that Coleman projects the social dimension of rule‐governed
activity (“What would the boys say?”) into the substantive dimension (“What does it
mean?”). Justifications of legal acts, as Hart conjectured correctly, tend to base
themselves on rules and not on social reputation. The substantive dimension provides
participants with access to justifications. However, it is this very same dimension that
allows participants to transcend the context of conventions. Vindications of rule‐following
are able to escape from intellectual impasse only by either exploring methods of
interpretation or by making sense of the rule itself.

Coleman goes wrong in presenting mindless social practice as though it embodied spirit.
Nowhere is this more obvious than at various points where Coleman replies to Dworkin’s
119 120
objection that a controversial rule of recognition cannot be conventional. The idea
that it is a convention ultimately begins to unravel in the course of his replies. The
ultimate characterization renders the rule of recognition as a shared co‐operative activity.
Coleman characterizes such an activity as “a framework of co‐ordination, planning, and
negotiation,” 121 which leaves wide leeway for disagreement. Coleman states explicitly:

It is not surprising that in resolving such disputes, the
parties offer conflicting conceptions of the practice in
which they jointly participate, conceptions that appeal
to different ideas of its point or function. 122

117
In anticipation of purportedly Wittgensteinian objections I hasten to add that even within a reflective practice
all justifications come to an end. That justifications come to an end (in the sense envisaged by Wittgenstein) does
not mean that they are arbitrarily cut off.
118
I guess this is what Dworkin has in mind when speaking of a “pre‐interpretative” stage of practice. DWORKIN,
LAW’S EMPIRE, supra note 22, at 65–66.
119
120
COLEMAN, THE PRACTICE OF PRINCIPLE, supra note 44, at 68, 100, 116, 153–54.
121
Id. at 157.
122
Id.
2011]
I leave it to the reader to conclude how much is left of conventionalism here. 123 I should

rather like to return to the conflict between the two readings of positivism, to which I
alluded above.

According to the first reading, positivism is commendable because it explains the
possibility of, or even gives rise to, objective descriptions of law. Coleman’s
conventionalism can be understood as an attempt to reconstruct—borrowing his own
quasi‐transcendental parlance—the conditions of possibility of law’s objectivity. The
master convention is supposed to constitute the object. It can do its work only, however,
as long as it is supported by the complementarity of “This is what we do” and “I do as
others do.”

It is, however, the case that law can overcome its conventional constitution from within.
With the conditions of its social constitution the appearance of “objectivity” also
disappears. The negativity of rule‐following unearths the intellectual resources necessary,
but not sufficient, to extricate legal knowledge from the grip of social power, which is
nonetheless indispensible for bringing law into existence. Law’s conventional origin is in its
more mindful state represented within legal practice only qua fictional foundation (as
124
Kelsen’s most skeptical rendering of the Grundnorm suggests).

This concludes my discussion of conventionalism. Coleman’s first‐version type of legal
positivism is vulnerable to criticism from the second‐version type. First, as an attempt to
reconstruct for law the conditions under which it is constituted as an object, it ignores how
law can turn itself into a more mindful form of social practice. Second, through a failure to
account accurately for the conventionality of law Coleman falls victim to idealizations that
emerge from the practice itself. The theory remains enveloped in its own object. It is
125
blinded, as it were, by law’s ideology.

N. Reclaiming Positivism’s Critical Edge

The discussion above has tried its utmost to be performatively consistent. It has submitted
the first reading of legal positivism to critical scrutiny from the perspective of the second.
The result may seem woefully counterintuitive. The true spirit of legal positivism does not
reside in the social fact dogma. Owing to the negativity of rule‐following, law has the
potential to outgrow the form in which it could be an object of description.
123
I, for one, think that Dworkin is that right in suggesting that Coleman “has wholly decamped from the
philosophical heritage he undertakes to defend.” DWORKIN, JUSTICE IN ROBES, supra note 5, at 198.
124
HANS KELSEN, GENERAL THEORY OF NORMS 256 (trans. M. Hartney, 1991).
125
That there is something like an “ideology of law,” which is part of socially relevant legal knowledge, was one of
the more astute insights of the critical legal studies movement.

What is, then, the spirit of legal positivism? It is captured in the maxim to arrive at
constructions of legal materials that de‐construct, to the greatest extent possible, the
idealizations which with these materials are encountered in society without, however,
thereby eliminating their normative meaning. For example, interpretations of an
international agreement are not infrequently defended as following from its “spirit” or
“general scheme.” 126 Legal positivists reconstruct the force of such a claim in less
exuberant terms by saying, for example, that judges simply have the legal power to make
claims of this type even if they are largely nonsensical. When the law invokes reason, legal
positivism perceives action.

Legal knowledge is an instance of social action. Law comes into this world in acts that
claim to be law. 127 Aside from being instances of social action, the meaning of those acts is
sociological in that they claim to determine authoritatively what situations mean in terms
of rights and obligations, powers and immunities or privileges and liabilities.

The validation of these draws on so‐called “sources” of law. In the German tradition, the
most salient of these sources were taken to be custom, legislation and legal doctrine. 128
Theorists of the historical school suggested that it is possible to arrange these sources in a
sequence and to reconstruct how the later source fixes problems left unresolved by the
former. 129 Indeed, from the perspective of the constitution of legal knowledge it is
possible to see that some ground of knowledge could only have been claimed to be a
source owing to the prevalence of idealizations masking specific deficiencies. For example,
one can profess belief in custom as source of law only by glossing over the fact that it
remains profoundly unclear how much usage by whom is necessary to constitute sufficient
practice and what it takes to encounter genuine opinio iuris. It is unlikely that any theory
of customary law would ever be capable of arriving at a satisfactory answer to this
question. The reason for being pessimistic is that convincing accounts would invariably
have to move custom into a direction where it would appear increasingly similar to a
process of legislation, for example, by specifying the number of confirmations, valid modes
of expressing consent and constitutional principles that it is required to respect.

Two conclusions follow from this. First, it is possible to see legislation as a more rational
rendering of what might have amounted to custom. Second, operating with customary law
126
Case 26/62, van Gend en Loos v. Nederlandse Administratie der Belastingen, 1962 E.C.R. 1.
127
KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY, supra note 46, at 9–10.
128
GEORG FRIEDRICH PUCHTA, 1 DAS GEWOHNHEITSRECHT 147 (1828).
129
See generally CARL FRIEDRICH VON SAVIGNY, ON THE VOCATION OF OUR AGE FOR LEGISLATION AND JURISPRUDENCE (trans. A.
Hayward, 1999).
2011]
as a source presupposes idealizing indeterminate factors into determinate conditions of
law‐making by claiming that there is enough use and sufficient opinio iuris.

It is not the case that legal knowledge can avoid all idealizations. Every application of rules
involves idealizations in that it perceives certain aspects of a situation as instances of this
or that rule. Unnecessary idealizations, however, cover up uncontrolled shifts of power in
the relation to adjudicating institutions or to whomever they wish to be of service.

A move from custom to legislation does not remove all unnecessary idealizations. The
belief in legislation as a source of law itself ignores that the application of one law rather
than another needs to be mediated by a way of knowing which law is to be applied in
certain cases. The appeal to legislation as a source idealizes its ability to guide as though
laws applied themselves automatically to cases. The effectiveness of legislation, however,
depends decisively on the intervention of systematic legal analysis.

In contrast to the common law tradition, which tends to succumb to gross idealizations of
the judicial process at large, the continental legal tradition is confronted with the
embarrassing realization that scholarship not only is a source of law, but compared with
others also most rational, for it establishes intelligible links between the legal system as a
whole and the analysis of singular cases. It is superior to judicial decisions in that it
develops the grounds that account for both their rationality and validity. Not surprisingly,
awareness of scholarship’s exalted position is a wellspring of idealizations of which
positivism as descriptivism is merely one notable exemplar.

Legal scholarship needs to cope most reasonably with its fate of being a source. To that
end, it has to be conducted in a spirit of self‐abdication. This requires, in particular,
130
emphasizing the extremely limited relevance of moral objectivity to legal scholarship.
Hence, instead of producing morally engaged scholarship legal analysis had better
reconstruct the true normative significance of the self‐interpretation of legal materials and
explore what it would take to accept their legitimacy. 131 It also requires rigorous attention
to choices made by judges and the real capacity to make them stick.

Unreflective legal scholarship is intellectually and socially just as treacherous as the
notorious quadditions of courts. Only legal scholarship that adopts as its task piercing
through the veil of idealizations will be in a position to rescue legal scholarship from an
encumbrance by money and power. Constructive efforts are necessary to this end, and
legal scholarship, as a source, will only attain a legitimate form by making itself subservient
130
For a good explanation, see JEREMY WALDRON, LAW AND DISAGREEMENT 176–86 (1999).
131
Because I am making grand claims here, I refer to my own example of how I believe that promise can be
delivered. Alexander Somek, The Argument from Transnational Effects I: Representing Outsiders through
Freedom of Movement, 16 EURO. L.J. 315 (2010).
to processes of collective self‐determination. Because, in this respect, the relevant
scholarship has to base itself candidly on a requisite political commitment, it can no longer
continue the positivist project. Nonetheless, it can inherit its spirit.

O. Conclusion

I mentioned above that legal knowledge is society’s mode of authoritatively reflecting on
itself in terms of rights and obligations, powers and immunities or privileges and liabilities.
Ultimately, the legal relationship reveals what we are to one another.

This is not to say that legal knowledge gives us a full and rich picture of social life.
However, insight into modern law reveals what our relationships ultimately turn out to be
when things have gone sour. Law is the ultimate point of contraction of social life. As
participants in legal relationships, we conceive of one another from a perspective of
utmost estrangement, even alienation from ourselves. Law is a social technique of
disillusionment. It is enlightenment in practice. It brings to the fore what remains of
various purportedly nice and loving dealings as soon as we want to know what we can
expect at the end of the day. When the pleasant soundings of comradery have abated we
realize that we are factors of production. When love is gone we see that matrimony is,
after all, merely a contract that involves, among other things, pledging each other the
132
reciprocal use of the genitals.

There is profound estrangement at the heart of our social existence. Law, if stated in the
terms of modest legal scholarship, brings it to the fore. Legal positivism has taught us that.
It should not be forgotten.

132
IMMANUEL KANT, THE METAPHYSICS OF MORALS 96 (trans. M. Gregor, 1991).
Special Issue
Seven Role Models of Legal Scholars

By András Jakab *

A. Introduction

In the following Article I will develop seven ideal‐typical role models for (or caricatures of)
legal scholars: the Oracle Scientist, the Prophet, the Law Reformer, the Humble Clerk, the
Wise Pragmatist, the Self‐Reflective, and the Media Star. 1 Ten features will be used to
characterize the ideal for each model: primary audience, ideal function, perverted form,
influence on the law, prestige, measure of success, time scale, use of non‐legal (moral or
social/economic) arguments, precondition of existence, typical countries, and famous
lawyers (representing themselves or at least proposing to other legal scholars to follow
that specific role model). 2

The list of role models is not meant to be exhaustive, as there are possibly other ones with
which we could sophisticate the picture endlessly. The role models themselves are also
not meant to be exclusive; it is possible that one single scholar bears features of different
role models, or writes a paper in one role and another paper in another role. Probably no
legal scholar would fully fit into any of the role models, but these as ideal types (in a
Weberian sense) still seem to have some explanatory force about how we perceive our
task. The scope of the explanation is consciously limited: The construction of these ideal
types had the purpose of explaining role models of legal scholars only in the German
speaking European countries (and to some extent, in other civil law countries such as Spain
and Italy), Hungary (as a post‐socialist country), the United Kingdom, and the United
States. The role models are thus not analytical constructions which show all logically
possible constellations of the different characterizing features; only those which seemed to
be able to explain existing legal scholarly mentalities and self‐perceptions in the named
*
Associate Professor, Pázmány Péter Catholic University Budapest, e‐mail: jakabandr@googlemail.com.
1
The Article is also a comment on Alexander Somek’s The Spirit of Legal Positivism published in this volume of the
GERMAN LAW JOURNAL. For useful remarks and critical observations I am grateful to Mátyás Bódig, Lisa Giles, Tamás
Győrfi, Gábor Hamza, Béla Pokol, and Alexander Somek; further to the participants of the research seminar held
th
at the Pázmány Péter Catholic University Budapest on the 4 of October 2010, especially to Péter Cserne and
Miklós Szabó.
2
The expressions “role” and “role model” in this Article are not used in their specific and precise social‐
psychological or anthropological senses, but only in their everyday meaning.
countries have been constructed. Though the categories of the present study might be
broadened to include further countries, this has not been tried or tested here.

The development of ideal typical role models also contests the notion of having a general
idea of the legal scholarship or legal science. Long but fruitless debates have been
conducted on the question of whether legal scholarship is scholarship or science at all.
Those who were skeptical about it annoyed or provoked the legal academia 3 and those
who confirmed it with a loose concept of scholarship or science were popular. 4 But as a
matter of fact, what legal scholars are doing under the heading of legal scholarship is just
very different with seemingly no common thread. Whether we label it with the noble
word science or scholarship is secondary; what really matters is the kind of presuppositions
behind the different ways of conducting legal scholarship (and under what circumstances
they can be accepted), and why in their genre some of the works will be considered better
or worse by their academic community. Thus, deducing good legal scholarship from an
abstract idea of the legal science is rather hopeless; what we can rather do is to
understand our own approach, what the most meaningful way of undertaking it could be,
and what pitfalls within that we should avoid. Consequently, I do not have a precise
definition of the "legal scholar" either: Everybody will be considered as such, if he or she
considers himself or herself as such.

B. Telling Judges What the Law Is: The Oracle Scientist

The Oracle Scientist’s primary audience is fellow academics, his secondary audience,
judges. He knows best what the law is, even better than judges. As a matter of fact,
judges rely on the opinion of the Oracle Scientist(s) to solve cases. It can happen either
directly by sending the file to them—see the Middle Ages German institution of
Aktenversendung which meant that courts sent the files of a case to universities for
5
decision —or by their consulting of academic literature. The reason for citing academic
literature by practitioners can also be found in positive procedural law: During the Middle
3
See generally JULIUS VON KIRCHMANN, DIE WERTHLOSIGKEIT DER JURISPRUDENZ ALS WISSENSCHAFT (1848), available at
http://fama2.us.es/fde/ocr/2006/werthlosigkeitDerJurisprudenz.pdf; ANDRÁS SAJÓ, KRITIKAI ÉRTEKEZÉS A
JOGTUDOMÁNYRÓL (1983).
4
See generally Ulfrid Neumann, Wissenschaftstheorie der Rechtswissenschaft, in EINFÜHRUNG IN DIE
RECHTSPHILOSOPHIE DER GEGENWART 385–400 (Arthur Kaufmann et al. eds., 2004) (providing an overview of the
literature).
5
See HAROLD J. BERMAN, LAW AND REVOLUTION II: THE IMPACT OF THE PROTESTANT REFORMATIONS ON THE WESTERN LEGAL
TRADITION 143 (2003) (including further references to specific provisions of contemporary procedural norms); see
also GERHARD KÖBLER, DEUTSCHE RECHTSGESCHICHTE 156 (1996). Similar practice existed in ancient Roman law, when
the praetor (an elected lay politician serving also as pre‐trial judge) and the judex (the actual judge, equally a
layman, appointed ad hoc by the praetor) asked the jurists (iuris consulti, called also men experienced in law, or
iuris prudentes) about how to solve the case. RANDALL LESAFFER, EUROPEAN LEGAL HISTORY 92–93 (2009). Gaius lists
scholarly opinion (responsa prudentium, “answers of the learned”) as a source of law. See GAIUS, INSTITUTES 1.2.
2011]
Role Models of Legal Scholars 759
Ages in certain parts of Europe, it was in some courts obligatory for advocates to cite
relevant literature (cf. "chi non ha Azzo non vada in palazzo"; "Quidquid non agnoscit
glossa nec agnoscit forum"), and while judges could be held personally liable for wrong
decisions, they avoided such liability if they followed the communis opinio (majority of
authoritative writers). To fall into the latter category, they quoted as much literature as
they could. 6 But even in modern times, in some legal orders positive law looks to the “the
common and constant opinion of learned persons” (Code of Canon Law of 1983, Canon 19:
"communi constantique doctorum sentential"), to the “approved legal doctrine” (Swiss
Civil Code, art. 1.3: "bewährte Lehre"), or to the “the teachings of the most highly
qualified publicists of the various nations” (Statute of the International Court of Justice, art.
38(1)) for the solution of cases.

Oracle Scientists who have never worked before in lower judicial offices are often invited
7
to the highest judicial positions. Her influence is thus considerable on the law, even
though she does not bear responsibility—in a legal sense—for what she is doing. 8 Her
prestige is high, her word—or at least the word of the majority of the Order of Oracle
Scientists, 9 i.e. the herrschende Lehre or the doctrine dominante—should be followed by
judges. If the legislator is afraid that its law will be “modified” by the Oracle Scientists—a
well‐founded fear in some cases—then it has to prohibit their work. 10

The Oracle Scientist is not simply a scientist expert, he is more than that. He has some
esoteric knowledge, so he can tell what the law states about a problem, even if laypersons
do not see any legal provision on the issue. He is an oracle who can see the signs from
which he reads the law. But his job is different from religious oracles in that it can be
6
William Twining et al., The Role of Academics in the Legal System, in THE OXFORD HANDBOOK OF LEGAL STUDIES 938
(Peter Cane & Mark Tushnet eds., 2003).
7
Id. at 940.
8
For the legitimacy problem of influencing the content of law, see MATTHIAS JESTAEDT, DAS MAG IN DER THEORIE
RICHTIG SEIN . . . 83–85 (2006).
9
You can become a member of the Order of Oracle Scientists if you get accepted in a long ritual by current
members. The ritual includes the defense of a doctoral dissertation, the submission of a Habilitationsschrift, and
finally the appointment as full professor. Without having fulfilled these steps of the ritual, whatever you say will
count less—independently from the content of what you say—than what the very last member of the Order says.
10
In the 6th century AD, Justinian made it obligatory to follow the opinions of some (mostly then dead) scholars
by codifying them—basically he made obligatory to follow his own code which happened to be based partly on
scholarly opinions—but forbade any new commentaries on it: “We command that our complete work, which is to
be composed by you with God’s approval, is to bear the name of Digesta or Pandects. No skilled lawyers are to
presume in the future to supply any commentaries thereon and confuse with their own verbosity the brevity of
the aforesaid work . . . .” Constitutio Deo auctore 12. With an even more radical solution, French revolutionaries
simply closed law faculties in 1793. RAOUL C. VAN CAENEGEM, JUDGES, LEGISLATORS, AND PROFESSORS 156 (1987).
learned like a science, through long years of training and practice in the Order of Oracle
Scientists.

They use a special language and logic, that of the Rechtsdogmatik. If you want to express
arguments of efficiency and justice in the language of Rechtsdogmatik, then you have to
translate them into legal arguments (“unlawful”). 11 Arguments of effectiveness, et cetera,
are not necessarily irrelevant to law, but they cannot be deployed directly (“naked”), 12 we
need the all‐knowing translator to do that for us (e.g., with the help of Generalklauseln).

But the Oracle Scientist is a very peculiar translator, because you cannot ask him to
translate just anything. He is building a legal conceptual system on his own which also
includes wishes for a better society even though he never talks about such policy issues
directly (that would be improper). But somehow, the system he builds up helps to realize
meaningful policy issues. The Oracle Scientist seems to know all our wishes and questions
about how to apply the law even before we express them. He must think in two tiers: the
13
legal language he speaks, and the real‐life (social or moral) issues he never talks about.

Her aim is to build up a conceptual system (doctrine, Rechtsdogmatik) by eliminating
contingencies and apparent gaps or contradictions. 14 The system he builds always gives
the one single legally right answer. The Oracle Scientist likes to represent himself as a
neutral, professional, and objective scientist, 15 but as a matter of fact his job is very
creative and he is building his own implied values into his system. 16 The system looks as its
very own emanation, which expands itself the Oracle Scientist being only the mouthpiece
of the autonomous Rechtsdogmatik.

11
PHILIPPE A. MASTRONARDI, JURISTISCHES DENKEN 264–76 (2001).
12
Markus Pöcker, Unaufgelöste Spannungen und blockierte Veränderungsmöglichkeiten im Selbstbild der
juristischen Dogmatik, 37 RECHTSTHEORIE 157–60 (2006).
13
See András Jakab, What Makes a Good Lawyer?, 62 ZEITSCHRIFT FÜR ÖFFENTLICHES RECHT 275–87 (2007) (discussing
two‐level thinking).
14
Contradictions have to be eliminated by way of interpretation, otherwise one cannot contribute to the solution
of future problems. If contradictions are merely highlighted, those applying the law will stare puzzled at the two
passages, then decide by tossing a coin. The Oracle Scientist’s task is to help avoiding this, thus making a
calculable functioning of the system possible. Eike von Savigny, Die Rolle der Dogmatik—
Wissenschaftstheoretisch Gesehen, in JURISTISCHE DOGMATIK UND WISSENSCHAFTSTHEORIE 104 (Ulfrid Neumann et al.
eds., 1976).
15
Oliver Lepsius, Themen einer Rechtswissenschaftstheorie, in RECHTSWISSENSCHAFTSTHEORIE 3 (Matthias Jestaedt &
Oliver Lepsius eds., 2008).
16
Vorwort, in DAS PROPRIUM DER RECHTSWISSENSCHAFT XII (Christoph Engel & Wolfgang Schön eds., 2007).
2011]
The perverted form of the Oracle Scientist is the legal scholar, who does not care about the
practical applicability and usefulness of his conceptual system. Rudolf von Jhering, one of
the greatest Oracle Scientists of all times, described the perverted and exaggerated
approach in his classic (self‐)ironic essay on the “juristic conceptual heaven” (Im
juristischen Begriffshimmel) the following way:

Concepts do not tolerate any connection to the real
world. . . . In the world of concepts, that you can see
here, there is no life in the sense you know it, it is the
empire of abstract thoughts and concepts, that follow
logically . . . from each other which for that reason shy
away from every touch of the earthly world.

Here rules only pure science, legal logic, and the
precondition of their rule and dignity is exactly . . . that
they do not have to do anything with life. . . . Life . . .
is synonymous with the death of science.

The lawyer calculates with his concepts, just like a
mathematician works with his numbers; if the end
result is logically correct, then there is nothing else to
be worried about any more.

The concepts you can see here exist, and with this
everything has been said. They are absolute truths, –
they always have been – they always will be. To ask
about their nature and justification is nothing better
than to ask ‘why is two times two four’. It is just four.
With this ‘is’ has everything been said, there is no
justification for it. It is exactly the same with concepts,
they are based as absolute truths in themselves, there
is no a justification for them. The only thing which a
thinker can do is to immerse oneself in them with full
devotion and without any restraint, in order to reveal
the amplitude of their content . . . . What he can
reveal this way is truth, and every truth has a claim to
be universally valid.

[J]ust like a natural scientist who tries to discover the
secrets of nature, also the legal researcher has no
other purpose than to unlock the beautiful secrets of
the legal world, to reveal the fine veins in the logical
organism of the law. 17

If you dare to question the Oracle Scientist by saying that he is doing Begriffsjurisprudenz
as described by Jhering, then her answer will probably be that this is her job 18 and that she
does not have the legitimacy to do more—like considering policy issues—and that she
serves legal certainty this way the best. Maybe he would also note that he is actually doing
a special type of Begriffsjurisprudenz, the so‐called Interessenjurisprudenz, which is a more
sophisticated version—and not the opposite—of the former and that with the latter he can
avoid absurdities. 19 He might also add that this type of criticism is very old, 20 and getting a
bit boring now. He would definitely point out that the critiques come and go, but the
Order of the Oracle Scientists stays. 21 If someone wants to criticize them meaningfully—so
the Oracle Scientist—then concrete mistakes should be pointed out, 22 but the job as such
should not be doubted.

He is a secret law‐maker: He pretends not to make law, as nobody empowered him to do
so, but in fact he does so, at least in a limited way. In order to conceal his law‐making, he
is telling us that he is just explaining the conceptual system of law and drawing
consequences from it. But this system is actually partly the result of his scientific work. 23
But if we ask another fellow or conspirator Oracle Scientist, then he is also going to confirm
that none of them is actually making any law.

17
RUDOLF VON JHERING, SCHERZ UND ERNST IN DER JURISPRUDENZ 252–53, 258, 274, 287, 288 (1884).
18
Lawyers necessarily work with abstract concepts. Eugen Bucher, Was ist »Begriffsjurisprudenz«?, in THEORIE UND
TECHNIK DER BEGRIFFSJURISPRUDENZ 389 (Werner Krawietz ed., 1976).
19
This is well shown by the fact that the four traditional methods of Savigny (grammatical, logical, systemic, and
historical) are not replaced by the teleological (or purposive) interpretation of which Jhering is thought to be the
inventor; rather, it is added to them as a fifth method. The opposite of Begriffsjurisprudenz is the “School of Free
Law” (Freirechtsschule). Id. at 372–73.
20
See Philipp Heck, Was Ist Die Begriffsjurisprudenz, Die Wir Bekämpfen?, 14 DEUTSCHE JURISTENZEITUNG 1456–61
(1909).
21
It does so openly and explicitly in the tradition of Begriffsjurisprudenz today. ROBERT ALEXY, THEORIE DER
GRUNDRECHTE 38 (2001). While rejecting mere logical inference, Alexy still thinks the elaboration of the conceptual
system to be the primary goal of jurisprudence, and in this aspect he explicitly sides with the tradition of
Begriffsjurisprudenz. Another remarkable contemporary advocate of this approach is Armin von Bogdandy, The
Past and Promise of Doctrinal Constructivism, 7 INT'L J. CONST. L. 364–400 (2009).
22
Only concrete logical faults have to be shown, but conceptualism as such should not be criticized. Bucher,
supra note 18, at 388; Horst‐Eberhard Henke, Wie tot ist die Begriffsjurisprudenz?, in THEORIE UND TECHNIK DER
BEGRIFFSJURISPRUDENZ 415 (Werner Krawietz ed., 1976).
23
Andreas Voßkuhle, Neue Verwaltungsrechtswissenschaft, in 1 GRUNDLAGEN DES VERWALTUNGSRECHTS § 1 n.6
(Wolfgang Hoffmann‐Riem et al. eds., 2006).
2011]
His system covers all the possible problems 24 without gaps 25 or contradictions. 26 The
structure of the system has to be built up, even if it is a chaotic uncodified system. Or
rather, it is even more important in these situations 27 as the most important feature of
legal scholarship is its systematic nature. 28

Legal scholarship does not simply describe the law, it represents the law. 29 Consequently,
a good practitioner has to know thoroughly the academic literature (nemo bonus iurista
nisi bartolista). 30 So if the practitioner does not read the Oracle Scientist, it is not a failure
of the Oracle Scientist, it is a failure of the practitioner. We do not measure the success of
the Oracle Scientist through his or her influence on the legal practice. If the legal practice
does not listen to what he or she is saying, that is their (intellectual) problem. The Oracle
Scientist is building his system for its perfectness (i.e., full, detailed and non‐contradictory)
and beauty; practicability is important (this is the official function of the system), 31 but
only secondary. Even positive law can come and go, but the conceptual system remains, it
has been built for eternity. 32 Changing social circumstances also do not really influence it.
24
Building a conceptual system instead of reproductivity is advocated also by Henke, supra note 22, at 414.
25
The ideal of gaplessness is characteristic not only of Begriffsjurisprudenz, but also of the rationalist natural‐law
tradition. GUSTAV BOEHMER, GRUNDLAGEN DER BÜRGERLICHEN RECHTSORDNUNG 2.1: DOGMENGESCHICHTLICHE GRUNDLAGEN
DES BÜRGERLICHEN RECHTES 63 (1951). For reference to CHRISTIAN WOLFF, see Begriffsjurisprudenz, in THEORIE UND
TECHNIK DER BEGRIFFSJURISPRUDENZ 432–37 (Werner Krawietz ed., 1976), especially page 436. The beginnings of
conceptual system‐building in law are traced back to scholastics—or its reflections in the works of the glossators
and commentators—by Berman. Harold J. Berman, The Origins of Western Legal Science, in THE WESTERN IDEA OF
LAW, 399–413 (1983). Special note should be taken of pages 401 and 405.
26
Even authors outside of the Begriffsjurisprudenz tradition often assume non‐contradiction in the case of a legal
system. J.W. HARRIS: LAW AND LEGAL SCIENCE, AN INQUIRY INTO THE CONCEPTS “LEGAL RULE” AND “LEGAL SYSTEM” 11, 81–83
(1979).
27
Twining et al., supra note 6, at 937.
28
“Rechtswissenschaft ist systematisch oder sie ist nicht” (“Legal scholarship is either systematic, or it is not legal
scholarship”). Lepsius, supra note 15, at 16.
29
“Rechtswissenschaft scheint das Recht nicht nur zu beschreiben, sondern auch zu vertreten.” Christoph
Möllers, Vorüberlegungen zu einer Wissenschaftstheorie des öffentlichen Rechts, in RECHTSWISSENSCHAFTSTHEORIE
167 (Matthias Jestaedt & Oliver Lepsius eds., 2008).
30
Renowned German law firms expect their applicants to hold a doctorate.
31
In lack of potential practicability, the Oracle Scientist does not build grand theories, but only middle‐level
theories—for example, the theory on proportionality—which can be used to sophisticate the conceptual system
designed to solve cases. Lepsius, supra note 15, at 26.
32
In this sense Rechtsdogmatik is conservative. Id. at 19. Similarly, see the Hungarian concept of the “invisible
Constitution” as developed by the then president of the Hungarian Constitutional Court László Sólyom in his
concurring opinion in 23 December 1990. (X.31.) AB, ABH 1990, 88, 97‐98. Sólyom famously said in different
interviews that even if there will be a new Constitution, the “invisible Constitution” consisting of the conceptual
system, remains the same—not even the constitution maker could change it. The plausibility of such opinions is
limited though, if positive law is changing so much, that there is no time to adjust the system to the new legal

Even though we do not measure the success of the Oracle Scientist by her reception or
influence amongst practitioners, you do need a legal culture where this type of behavior is
accepted. You need a demand amongst legal practitioners for the opinion of the Oracle
Scientist. If legal practitioners (especially judges) do not particularly respect legal scholars,
then the scholarly attempt to scientifically and conceptually systematize law will miserably
fail (cf. the cases of Wesley Hohfeld 33 and Christopher Columbus Langdell 34 ). There is just
no demand for the work of the Oracle Scientist in some legal cultures. 35

Oracle Scientists accept that there might be other types of legal scholars, but they consider
themselves as the most important body of legal scholars, and consider their approach as
the “heart” of legal scholarship. 36

In civil law countries where judges receive their positions more or less straight after their
law degree, 37 the chances of such an approach are typically higher than in common law
countries where the most prestigious legal job is the one of the judge. In the latter
countries the scholar either has to be either a Humble Clerk of judges, 38 or a Prophet
leading superhuman (Herculean) judges, or a Wise Pragmatist advising them to consider
non‐legal (economic, social) aspects, or a Self‐Reflective who does not care about judges
but writes only to other fellow academics. But a common law scholar using only legal
(non‐moral and non‐social) arguments telling judges about what the law is is hardly
situation. The Oracle Scientist thus struggles, if the legislator or the constituent power are too active. Möllers,
supra note 29, at 165; Vorwort, in DAS PROPRIUM DER RECHTSWISSENSCHAFT XI (Christoph Engel & Wolfgang Schön
eds., 2007).
33
Wesley Newcomb Hohfeld, Some Fundamental Legal Conceptions as Applied in Legal Reasoning, 23 YALE L.J. 16
(1913).
34
On Langdell’s vision of law, see John Chipman Gray, Langdell’s Orthodoxy, 45 U. PITT L. REV. 1 (1983). Only his
teaching method (case method) survived, but not his approach to legal scholarship. WILLIAM P. LAPIANA, LOGIC AND
EXPERIENCE: THE ORIGIN OF MODERN LEGAL EDUCATION 148–70 (1994); ROBERT STEVENS, LAW SCHOOL: LEGAL EDUCATION IN
AMERICA FROM THE 1850S TO THE 1980S 35–72 (1983).
35
As Somek points out: “The common law tradition remains at odds with the constructive legal scholarship. A
new case changes everything.”
36
For Dogmatik as the “core discipline of legal scholarship” (Kerndisziplin der Rechtswissenschaft), see Ralf
Dreier, Rechtstheorie und Rechtsgeschichte, in 2 RECHT‐STAAT—VERNUNFT: STUDIEN ZUR RECHTSTHEORIE 217 (Ralf
Dreier ed., 1991).
37
Judges are thus young, inexperienced, and receive only a modest salary. Their social standing is also modest. In
most cases, they cannot submit dissenting opinions, so they cannot make names and they remain unknown in the
legal community. Twining et al., supra note 6, at 939.
38
If judges disagree with the Prophet, then they are evil and cynical betrayers of their profession. Cf. Ronald
Dworkin, The Decision that Threatens Democracy, THE NEW REVIEW OF BOOKS, 13 May 2010. Note that this type of
criticism—as opposed to the criticism by an Oracle Scientist—is not about the judges’ intellectual capacity.
2011]
conceivable. However, in civil law countries (especially in the Germanic legal family) it is
usual practice. 39 The relationship is well expressed by Merryman: “The scholar is the
scientist, and the judge, at best, merely the engineer. The scholar provides the systematic,
scientific legal structure that the judge accepts and applies. The work of the scholar is
creative and exalted; that of the judge is, although important, on a lower plane.” 40

The reason why conceptual‐doctrinal legal thought (Rechtsdogmatik) reached its highest
level—though at the time still fragmented—in Germany in the nineteenth century is that
on one hand law was in principle quite rigid as they used the ancient and highly esteemed,
consequently unquestionable, law, that is Roman law, or its modernized version, the “ius
romanum hodiernum” (heutiges römisches Recht). On the other hand, they enshrined
conceptual elaboration into a conscious program (as an alternative to the French
41
revolutionary invaders’ codification). This combination is unique in history, and even
today’s German legal scholarship owes its conceptual sophistication to it. 42

C. Telling Judges What the Right Thing Is to Decide: The Prophet

The Prophet is primarily talking to judges; his secondary audience is legal scholars. 43 But
he does not consider his profession as an Order (or a guild) like the Oracle Scientist did. He
sees himself rather as the leader of a church, 44 whose members are not only and not even
primarily legal scholars, but judges. Legal scholarship is about judges, 45 it is also written
for judges. The Prophet knows the moral foundations of our society and he will tell us
(especially judges) what to do. He knows the right way, he has the vision, he is
superhuman. He is Hercules, and if we follow him, we can become like him. 46
39
See generally Armin von Bogdandy, Prinzipien der Rechtsfortbildung im Europäischen Rechtsraum:
Überlegungen zum Lissabon‐Urteil des BVerfG, NEUE JURISTISCHE WOCHENSCHRIFT 1 (2010).
40
John Henry Merryman, The Italian Style III: Interpretation, 18 STAN. L. R. 583, 586 (1966).
41
See FRIEDRICH CARL VON SAVIGNY, VOM BERUF UNSERER ZEIT FÜR GESETZGEBUNG UND RECHTSWISSENSCHAFT (1814).
42
Italian legal scholarship with its abstract conceptualism seems similar in many respects to the German one—
though without Savigny’s conscious and explicit program. John Henry Merryman, The Italian Style I: Doctrine, 18
STAN. L. R. 39, 45–48 (1965).
43
See RICHARD A. POSNER, THE PROBLEMATICS OF MORAL AND LEGAL THEORY IX (2002) (using the expression “moral
entrepreneurs” for these (partly legal) scholars). However, I prefer the word “Prophet” as it rather expresses the
need for followers.
44
Those who do not like this approach might consider the Prophet as a religious fanatic. For example, the
description of Dworkin as “the Taliban of the Western legal thought . . . .” Richard A. Posner, The Problematics of
Moral and Legal Theory, 111 HARV. L. REV. 1637, 1695 (1998).
45
RONALD DWORKIN, LAW’S EMPIRE 90 (1986).
46
See id. at 239 (presenting the concept of the Judge as Hercules).
Unfortunately, society is not mature enough yet and they do not see his “truth.” The
Prophet is trying to explain to them that if they do not agree with him, they are just wrong.
But it is not always successful, so it is better to talk to a more elevated section of society,
to the intellectuals and among them especially judges.

Oracle Scientists might consider the “truth” to be what the majority of their Order thinks
(herrschende Meinung). But the Prophet does not need the majority of any Order or body.
The Prophet can see the truth on his own, without any help from others. The Prophet
thinks that Oracle Scientists—and also the Wise Pragmatists—are dishonest, manipulative,
pharisaic, and conspirative legal scholars who pretend not to move along their own
political‐moral agenda and to be neutral—even though they actually do have a conscious
47
political agenda behind the facade of neutrality.

On the base of the Constitution (considered as some kind of codified natural law of the
given political community), the Prophet can find out the One Single Right Moral Vision
(OSRMV) of the given society. 48 The Constitution is actually based on the OSRMV, only
that most people do not see it. As a matter of fact, only the Prophet can see it, as he is
superhuman. Judges are the avant‐garde, they have to lead social change implementing
the OSRMV; especially judges from the highest (constitutional or supreme) court have this
task. If they do not do it, they are heretics and betrayers. 49 Democracy is of course
(morally) good, but only if the people elect politicians who want to make laws exactly
conforming to the ideas of the Prophet. But if it is not happening (as it is not happening, as
the people have not yet recognized fully the Prophet’s truth), then judges have to
implement the OSRMV (relying on the Constitution) instead of politicians. The Prophet
purports that he has been enlightened about the OSRMV from the Constitution itself, so
there is nothing arbitrary in his method. Constitutional court judges and supreme court
judges often find his ideas appealing as these help them to expand their power and at the
same time to build up a moral image.

His critics doubt, however, whether the Constitution supports only one moral vision;
50
instead, they say that several moral visions fit to the Constitution. The critics say these
things only because they do not entirely yet understand the Prophet. With time, this will
47
RONALD DWORKIN, LIFE’S DOMINION 3–29, 118–47 (1993).
48
See DWORKIN, supra note 45, at 225–28, 254–58 (basing “law as integrity” on a coherent moral vision).
49
Cf. Dworkin, supra note 38. Some Oracle Scientists—following a Weberian view of science—plainly despise the
Prophet as un‐scientific. Cf. MAX WEBER, WISSENSCHAFT ALS BERUF 25 (1996) (“weil der Prophet und der Demagoge
nicht auf den Katheder eines Hörsaals gehören.”) (“because the prophet and the demagogue do not belong to the
lectern of a lecture hall.”). Weber himself rather belongs to the Self‐Reflective category, but his quoted text
makes explicit some presuppositions shared by most of the Oracle Scientists.
50
See NIGEL E. SIMMONDS, CENTRAL ISSUES IN JURISPRUDENCE: JUSTICE, LAW AND RIGHTS 217 (2002) (including a list of
further references).
2011]
definitely change, the more advanced teachings of the Prophet will prevail and the
outdated concurrent opinions will all fail. The Prophet’s teachings are for eternity, but
these are sometimes about very concrete issues, so if a new case comes up, it is better to
ask him again, so he can teach us something new and wise.

The Prophet can work only in a country where the judiciary is prestigious enough to be
brave enough to use his OSRMV. In countries where judges have legitimacy issues, the
Prophet’s not‐strictly‐legal considerations can have only limited success, as judges do not
dare to use such arguments. If judges’ power stems from the legislator, then they are
shyer to talk about moral visions, and they rather ask the Oracle Scientist about what to
51
do. The Prophet’s teaching is also difficult to use in practice, if the respective country
does not have a written constitution, because it is even more difficult to claim the
existence of OSRMV if it has to fit a chaotic ancient precedent system. The Prophet is thus
most likely to succeed in common law countries with a written Constitution, and to a lesser
extent, in civil law countries with a strong constitutional court where prestigious judges are
sitting. 52 It also helps his work if there are strong natural law traditions in his country. 53

The Prophet wants to influence the law through judicial decisions, even though he is not a
judge himself—as he is much more than that. If judges do not follow him, he can be said
to have failed. A Prophet without followers is no one but an arrogant lunatic.

D. Advising the Legislator What the Law Should Be: The Law Reformer

The Law Reformer does not really want to talk to judges. She wants to convince the
legislator. Either because he does not trust judges, intellectually or morally, or because he
thinks it is more democratic and transparent to change the law by the legislator. The Law
Reformer does not often like judge made law at all. With the famous words of one of
them:

51
Cf. Pierre Legrand, European Legal Systems are not Converging, 42 INT'L & COMP. L.Q. 52, 74–75 (1996) (covering
the difference between common law and civil law countries). He explains that in the common law perception the
court’s power is original—it stems from the common law and not from statutes—as opposed to civil law countries
where it stems from the legislator. See also Stephen Sedley, Human Rights: a Twenty‐First Century Agenda, PUB.
L. 386 (1995) (discussing “bi‐polar sovereignty of the Crown in Parliament and the Crown in its courts”).
52
Unfortunately (for the Prophet), the constitutional court judges in civil law countries are often members of the
Order of Oracle Scientists, and they are not particularly impressed by the emotional parvenu Prophet. They
prefer to follow their centuries old traditions of (seemingly) value‐neutral scientific approach to law.
53
PATRICK S. ATIYAH & ROBERT S. SUMMERS, FORM AND SUBSTANCE IN ANGLO‐AMERICAN LAW: A COMPARATIVE STUDY OF LEGAL
REASONING, LEGAL THEORY, AND LEGAL INSTITUTIONS 230–31 (1987) (conceptualizing the Declaration of Independence
and the right to rebellion—which is very difficult to do without natural law arguments).
It is the judges . . . that make the common law. Do you
know how they make it? Just as a man makes laws for
his dog. When your dog does anything you want to
break him of, you wait until he does it, and then you
beat him for it. This is the way you make laws for your
dog: and this is the way judges make law for you and
me. 54

The Law Reformer has plans for society, how to improve it, and how to advance it. In order
to convince the legislator, he has to use policy‐arguments combined with moral
considerations.

This type of legal scholar dominated the landscape in socialist countries, where the official
doctrine of “socialist normativism” tried to turn judges into law‐applying machines, using
rather literal interpretation, and the official law‐making power only lay with the
Parliament. The Parliament followed the “scientific” and modern socialist views in order to
transform society into socialism and later into communism. For the big reform plan they
needed advisors on how to use law as an instrument of social transformation. These
advisors were legal scholars presenting de lege ferenda works offered to the legislator for
55
further use. If the legislator is legally omnipotent (i.e., there are no constitutional
constraints), 56 then we do not have to deal with intricate doctrinal questions at all and we
can concentrate on the instrumental character of law. Law is a means to change society,
and we need lawyers who can use this instrument. The most important lawyer is the
statute‐drafter or codifier: the Law Reformer. And if legal scholars want to be useful to
society (and not just theorize for no reason), then they should prepare proposals for new
laws. 57 They have to use the results of social sciences within the frames of orders
developed by politicians.

If judges happen to develop some new ideas then the scholar has to present a proposal
how to codify it, as judge‐made law is uncertain in this paradigm (as they could change
their practice and their decisions are more difficult to access than statutes): Only the
legislator makes real law. 58 A work written in this style is relevant as long as the social
54
Jeremy Bentham, Truth Versus Asshurst, in 5 THE WORKS OF JEREMY BENTHAM 233–37, 235 (W. Tait ed., 1843).
55
See András Jakab, Surviving Socialist Legal Concepts and Methods, in THE TRANSFORMATION OF THE HUNGARIAN LEGAL
ORDER 1985–2005 606–19 (András Jakab et al. eds., 2007).
56
At this point, the Westminster system and the socialist countries were very similar.
57
The real task of legal scholarship is preparing proposals for new laws; otherwise legal scholarship is just “useless
theorizing.” JÁNOS BEÉR ET AL., MAGYAR ÁLLAMJOG 16, 18 (1972).
58
The fact that judicial decisions are published or even systematized, and the fact that the legislator can change
the statute as easily as the judges their case law, do not seem to disturb the proponents of this approach.
2011]
problem to be solved is still there and/or the law for dealing with it has been made. After
this, the scholarly work can be thrown away.

The Law Reformer is successful if politicians accept his or her proposals, and if in practice
these new laws work. This approach can and sporadically does exist everywhere, but it is
surprisingly strong in some post‐socialist countries, where (despite a new constitutional
system) such old mentalities based on the unlimited central legislator and on a non‐
59
autonomous judiciary survived. It is getting weaker, but probably for several decades its
traces will remain strong in these countries.

E. Explaining to Attorneys and Law Students What Judges or Legislators Did: The
Humble Clerk

The Humble Clerk simply repeats what judges or legislators decided. His primary audience
is attorneys and law students. The Humble Clerk is never considered an authority. The
more precise he is in repeating what judges and legislators said, the better clerk he is.

The Humble Clerk can be considered as the other type of black letter scholarship besides
the Oracle Scientist. Also the Humble Clerk claims neutrality; the difference is that the
Humble Clerk is actually neutral, as he is not doing any creative work. He just repeats
(copy‐pastes), summarizes (this can be dangerously creative and imprecise) and describes
(never prescribes). Thus, he does not have to use any extra‐legal (moral or economic)
arguments either; there is nothing interdisciplinary in what he is doing.

If the Humble Clerk is trying to be creatively academic, it goes miserably wrong. He is just
not trained for such intellectual exercises. An apposite description of the problem is given
by Dicey:

Our best works, such as Smith’s Leading Cases, are at
bottom a mere accumulation of notes on detached
points of curious, rather than useful, learning. They
are deficient in all general conceptions, in all grasp of
principles, in all idea of method. . . . Turn, for
example, to a writer whose book was twenty years
ago the student’s guide to the law of contract. Mr JW
Smith opens his treatise with a chapter on the ‘Nature
and Classification of Contracts’. Of the nature of
contracts he tells his reader nothing. What he does
59
Similarly, please see the interview with Ferenc Petrik former Supreme Court judge in Hungary about the
scholarly value of codification, available at http://www.jogiforum.hu/interju/66. For a rebuttal by the present
author, see http://www.jogiforum.hu/interju/69.
tell them is that agreements consist of contracts of

record, contracts under seal and simple contracts. He
first substitutes division for definition, and then gives a
division which, for absolute uninstructiveness, may be
compared to an attempt to classify animals by dividing
them into dodos, lions and all animals which are not
dodos or lions. 60

The prestige of the Humble Clerk is low. He stayed at the university because he did not
make it to the Bar; he is a loser who does not dare to go out into real life. He is not good
enough for that, he can only become a clerk. Or he is just lazy, does not like long working
hours in law firms, and wants to enjoy the long summer holidays at universities.

In a country where the most prestigious job is the one of the judges, so if the judge is the
61
“living oracle of the law,” then the legal scholar cannot be anymore the oracle. If a
Humble Clerk writes a case note, he cannot dare to express criticism on what the judges
said. 62 Even if something is obviously wrong, he can only document what the judges said,
he cannot try to correct it or to dissent from it. 63 Dissent and correction are for judges and
legislators.

If he sees the work of the Oracle Scientists, he is partly impressed, partly critical about its
artificial system building and doubts whether it is actually the law as judges have not yet
confirmed it. 64 Humble Clerks have to be impressed how smart the judges are, and they
have to note and repeat the words of those wise men and women as precisely as
possible. 65 If judges seemingly contradict themselves or each other—a truly embarrassing
60
ALBERT VENN DICEY, CAN ENGLISH LAW BE TAUGHT AT THE UNIVERSITIES? INAUGURAL LECTURE, 21 APRIL 1883 13 (1883). As
a matter of fact, a real Oracle Scientist would consider Dicey’s works (for example, AN INTRODUCTION TO THE STUDY OF
THE CONSTITUTION (1959)) exactly as primitive and unsystematic as he does his contract lawyer colleague’s book.
61
WILLIAM BLACKSTONE, 1 COMMENTARIES ON THE LAWS OF ENGLAND 69 (1765).
62
63
See generally John Smith, An Academic Lawyer and Law Reform [Presidential Address, The Society of Public
Teachers of Law], 1 LEGAL STUD. 119 (1981) (providing anecdotes).
64
See Michelle Everson, Is it just me, or is there an Elephant in the Room?, 13 EUR. L. J. 136, 138 (2007) (criticizing
the Oracle Scientists (here, PRINCIPLES OF EUROPEAN CONSTITUTIONAL LAW (Jürgen Bast & Armin von Bogdandy eds.,
2006)) for seeing a systematic law where it actually does not exist; against the “Germanic obsession” of
systematic and elitist authoritative legal doctrine). Everson also considers the object of her criticism the
“theory”—or doctrine—built up by German scholars, most of whom, however, would probably not consider
themselves as Rechtstheoretiker, but rather as Rechtsdogmatiker.
65
See FIONA COWNIE, LEGAL ACADEMICS: CULTURE AND IDENTITIES 69 (2004) (discussing the minority complex of English
legal scholars).
2011]
situation for a clerk—then the Humble Clerk has to look again: Judges cannot have made a
mistake, somehow all the decisions have to be right, and they just must be understood the
right way. In his eyes, all judicial decisions are right, unless another judicial decision or a
statute tells you later that those were wrong.

Unfortunately, judges do not care about what the loser clerks say, they hardly quote any
scholars. 66 Judges even openly despise legal scholarship and talk about the “dangers, well
perceived by our predecessors but tending to be neglected in modern times, of placing
reliance on textbook authority for an analysis of judicial decisions.” 67 Judges in such a legal
order might “even regard it as complimentary to be told that they are suspected of having
little interest in theory or an academic approach to legal issues.” 68

If an academic wants to do something intellectually challenging in such a legal culture, then
he is doing research about law with non‐legal (sociological, political science, economic)
methods (socio‐legal studies), or he is writing about very abstract theoretical issues
(jurisprudence, constitutional theory). Thus she has to become a Self‐Reflective.

To use the expressions by Alexander Somek, the Humble Clerk claims and aims “descriptive
accuracy” about law—and what is the same for him, judicial decisions—but lacks the
“critical edge.” Whereas the Oracle Scientist also claims “descriptive accuracy” of the law,
but in the name of this descriptive accuracy he has the “critical edge” against judicial
decisions. According to Somek, we can see a change in mentality, the critical edge is
fading. I see the situation slightly differently. It is true that such mentalities change over
time, but I see the differences rather geographically. In continental countries, the Oracle
Scientist is the typical black letter legal scholar; in the U.K., the Humble Clerk; traditionally
in the U.S., it is rather what I call later the Wise Pragmatist and nowadays the Self‐
Reflective. The differences described by him are thus less a question of the history of legal
theory but rather of comparative law.

Traditionally the Humble Clerk is the typical English law school member—lacking the
scholarly ambition of the Oracle Scientist—as William Twining described the situation with
the following fictitious picture:

66
Cf. Hein Kötz, Scholarship and the Courts: A Comparative Study, in COMPARATIVE AND PRIVATE INTERNATIONAL LAW:
ESSAYS IN HONOUR OF JOHN H. MERRYMAN ON HIS SEVENTIETH BIRTHDAY 183–195 (D.S. Clarks ed., 1990) (showing that—
as opposed to thirteen secondary authority per judgment in Germany, where the academia is dominated by
Oracle Scientists—in England, dominated by Humble Clerks, you find only 0,77 in average).
67
Johnson v Agnew [1980] AC 367, 395 per Lord Wilberforce (House of Lords).
68
ATIYAH & SUMMERS, supra note 53, at 357. Until recently in England, many judges have never been to law school,
but after a non‐legal university degree went directly to the Bar and qualified there, so academics counted hardly
anything for them. Id. at 348.
[A] case exhibiting a sample of recent publications by
the Faculty, four slim monographs, about a dozen fat
books addressed to the student market (three of
which are past their third editions) and a number of
offprints with obscure titles, which some might think
are self‐addressed. No room could be found here for
rather more lucrative publications, such as nutshells
(or other student aids), contributions to loose‐leaf
practitioners’ services, and occasional journalism.
First impressions suggest that this is primarily a
teaching institution, which is quite a vocationally‐
oriented, but which is trying to build up his research
69
profile.

In England, as opposed to the Continent or the U.S., even for major scholarly tasks needed
by the government, judges seem rather to be the right candidates. 70 If there is a need for
an important report about the legal system, then rather judges seem to have the necessary
intellectual and moral qualification for it. 71 Legal scholars can quote them later and
disseminate the judge’s profound ideas.

In countries, where instead of the judge (and the judge made law) the emphasis is on (the
literal meaning of) codified law, the Humble Clerk notes and repeats what the legislator
said. If the legislator changes the law, all the former works of the Clerk can be thrown
away. 72 In this perception, it is possible to run a legal system without any kind of legal
scholarship. 73 Legal scholarship is thus not an intellectually valuable work, even though it
might be useful to do the secretarial part of running the legal system—as the legal material
is just big to overview it. 74
69
WILLAM TWINING, BLACKSTONE’S TOWER: THE ENGLISH LAW SCHOOL 69 (1994).
70
ATIYAH & SUMMERS, supra note 53, at 386–87.
71
See Glidewell Report (The Review of the Crown Prosecution Service, 1998), Woolf Report (Report on Access to
Civil Justice, 1996), Auld Report (Review of the Criminal Courts of England and Wales, 2001). Even if academics
write sometimes such official reports, it is not obvious for the English—as opposed to the continental or the US
perception—that academics should be the primary authors in such cases.
72
See KIRCHMANN, supra note 3, at 17 (“[D]rei berichtigende Worte des Gesetzgebers und ganze Bibliotheken
werden zu Makulatur.”) (“Three correcting words of the legislator and whole libraries become waste paper.”).
That is the reason why Oracle Scientists often oppose law reforms: it would ruin their oeuvre. Cf. id. at 11. For a
powerful response by Karl Larenz, see Karl Larenz, Uber die Unentbehrlichkeit der Jurisprudenz als Wissenschaft,
JURISTISCHEN GESELLSCHAFT BERLIN (2006), availale at http://www.juristische‐gesellschaft.de/schriften/26.pdf.
73
KIRCHMANN, supra note 3, at 7.
74
An even more despised category of legal scholars are clerks of party politicians in a dictatorship—so not even of
judges or of legislators. During the toughest times of socialism, legal scholars did not have the standing of Law
2011]

F. Advising Judges About Socially Best Decisions: The Wise Pragmatist

The Wise Pragmatist is writing for judges; he is advising them about socially best decisions.
In doing so, law in a strict sense is only one factor for him besides common sense,
sociology or economy; he is using thus policy arguments very often.

In general, he takes the normativity of law less seriously. To formulate it in a cynical and
provocative way: “The prophecies of what courts will do in fact, and nothing more
75
pretentious, are what I mean by the law,” or “[g]eneral propositions do not decide
76
concrete cases.” His doubt in the normativity of law makes him characteristically
different from the Oracle Scientist, the Humble Clerk, and the Prophet. He also lacks the
intellectuality of the Oracle Scientist and the moralism of the Prophet. He is rather
instrumentalist and utilitarian.

As opposed to the Law Reformer, he does not distrust judges. It is rather the opposite: a
precondition of the existence of the Wise Pragmatist‐type legal scholar is exactly a
prestigious and trusted judiciary which dares to use not strictly legal arguments—in a
society where judges are rather considered to be state bureaucrats of the “mouthpiece of
the law” type this approach is unlikely to succeed. Also as opposed to the Law Reformer,
the difference between legal and non‐legal arguments is less sharp for him. The Wise
Pragmatist does not take seriously the difference between what the law is and what it
77
should be, whereas the Law Reformer wants to see a sharp line. The Law Reformer
protects his own competence for changing the law—or at least advising the legislator
about it—in this way, whereas the Wise Pragmatist is helping to expand judges’
possibilities by blurring the frontier. In this sense, the Wise Pragmatist is inherently anti‐
positivist—or at least, as Holmes, an anti‐formalist positivist—and the Law Reformer is
positivist and formalist.

Another precondition for the existence of Wise Pragmatists is the inefficiency of
legislation. In the U.K. there are no such problems (cf. parliamentary sovereignty meaning
in practice the sovereignty of the House of Commons in which normally only one party is in
Reformers, they were just clerks of communist politicians repeating party resolutions, not proposing law reform.
It was (existentially) risky to propose anything new, as communist leaders might have disagreed, so the legal
scholar could lose his job or even more. See MICHAEL STOLLEIS, SOZIALISTISCHE GESETZLICHKEIT: STAATS—UND
VERWALTUNGSRECHTSWISSENSCHAFT IN DER DDR (2009) (providing additional details and stories).
75
Oliver Wendell Holmes, Law in Science and Science in Law, 12 HARV. L. REV. 443 (1899).
76
Lochner v. New York, 198 U.S. 45, 76 (1905) (Holmes, J., dissenting).
77
See GERALD J. POSTEMA, BENTHAM AND THE COMMON LAW TRADITION 147–336 (1986) (pointing out that Bentham’s
legal positivism can be explained by his purpose to reform English law).
government, and this party’s internal discipline is strong), and this type of legal scholar
could not emerge. 78 In the U.S., however, where the separation of powers in the law‐
making power (possessed by three politically distinct and powerful organs together:
Senate, House of Representatives, and President), the lack of party discipline in the
Congress (and consequently the numerous unexpected amendments proposed by
members of the Congress), 79 the rather low quality of legislative drafting, 80 the
constitutional limits (as interpreted by the Supreme Court), and the higher speed of
changing social circumstances 81 all contribute to the eminent need for judicial law‐
making. 82 Judges seem to be able to make law in the U.S. more speedily 83 and more
cheaply 84 than the legislator. It also means that judges in the U.S. have to look at the
future (solutions for social challenges) rather than at the past (case law). 85 Consequently,
judges adhere less to the doctrine of stare decisis, and deviate from past decisions more
often. 86 But they are grateful for advice as to in which direction they should do so. This
advice is very often not legal(istic) at all, but uses economic or social arguments (e.g., in
the form of a Brandeis Brief). 87

Judges in the U.S. are also often rather politicians (state judges are sometimes directly
elected by the local population for a limited term, federal judges are appointed mostly
exactly because of party membership and loyalty, so their audience is much less the legal
78
See ATIYAH & SUMMERS, supra note 53, at 141, 149 (explaining that Legal change in the UK is rather introduced by
legislation, which is prepared by highly educated civil servants and not by legal scholars).
79
Id. at 314.
80
Id. at 37, 334.
81
Id. at 134.
82
See id. at 270 (“[T]here can be little doubt that one of the principal reasons that American courts make so much
law is that [by comparison with British Parliament] American legislatures make so little.”).
83
Maurice Rosenberg, Anything Legislatures Can Do, Courts Can Do Better?, 62 A.B.A. J. 587 (1976).
84
RICHARD NEELY, HOW COURTS GOVERN AMERICA 30, 71 (1981).
85
See ROSCOE POUND, AN INTRODUCTION TO THE PHILOSOPHY OF LAW 47 (1922) (discussing the concept of Judges as
“social engineers”).
86
ATIYAH & SUMMERS, supra note 53, at 118–27. In the UK, higher courts bind themselves (with some exceptions);
in the U.S. they can overrule their own past decisions. In the U.S., even a lower court can overrule a past decision
of a higher court if it expects the higher court to overrule it (“anticipatory overruling”). The deviation from the
stare decisis is partly caused by the size of the U.S.; there are just too many judicial decisions to have a strict
stare decisis. Once the stare decisis is loosened up, there is also more room to follow academic opinions. Id. at
32, 128–30, 148.
87
Louis Brandeis, judge of the Supreme Court between 1916 and 1939, argued as an attorney in the case Muller v.
Oregon, 208 U.S. 412 (1908), by delivering a detailed sociological presentation of the social effects long working
hours for women.
2011]
profession than the electing people or the appointing politicians), so non‐legal (political,
social, policy, moral) arguments are much more acceptable for them than in other
countries. 88

As the law is difficult to change by legislation, it is better not to have “descriptive accuracy”
about law, because it would mean that we might discover the need for (cumbersome)
legislation. Rather, problems can be solved in a practical manner if we consider law as
aiming for meaningful social purposes and interpret it in the courts liberally. To find out
these purposes, academic literature can be of great help. The Oracle Scientist could also
give such advice, but in the U.S. nobody really believes that the law is in fact as systematic
as all Oracles Scientists profess. Law seems rather to be a practical device to manage
society. The technical language of advice by the Oracle Scientist does not give any
substantive reasons either why the final decision should be accepted (by the people or by
89
the politicians). And also importantly, except for some European immigrants there are
just no Oracle Scientists in the U.S., and even if they write about a question in their
legalistic style, an American judge just does not get them as their scholarly language and
style are out of touch with her needs and interests. If somebody tries to be an Oracle
Scientist in such a legal culture (like Langdell did), he will be dismissed as a “formalist” and
he will fail. 90

Because of the non‐professional factors in the selection of judges, sometimes the
necessary intellectual capacity is also missing to be able to understand the Oracle
Scientists: “It is amazing how many judges— especially, but not exclusively, state judges—
lack the basic intelligence to understand a moderately complex legal argument. Some are
just plain stupid; others lack the necessary legal education; still others are lazy and
91
impatient.”

For those judges who lack the doctrinal‐technical legal knowledge, it is just easier to decide
the case on its substantive merits. 92 To counter the mentioned unfavorable intellectual
88
ATIYAH & SUMMERS, supra note 53, at 342, 344, 350–51, 379. This is very different from both civil law countries
and England. For the latter, see Lord Scarman in McLoughlin v. O’Brian [1983] 1 AC at 340 about the necessary
abstinence of English judges from policy.
89
In a non‐elitist culture, the reasons have to be understandable and acceptable also for the general public. But
see ATIYAH & SUMMERS, supra note 53, at 38, 225, 232 (contrasting with the U.K.).
90
I am afraid that Professor Somek fights within the same environment as Professor Langdell did more than
hundred years ago, and his ambition that the critical edge of legal scholarship should be used in a “constructive”
(rather than clerk‐wise) manner is similarly doomed to fail. Professor Somek wants to proclaim the power of
Oracle Scientists in a country where there is hardly anybody with such an ambition and where the general
(practitioner, esp. judicial) mentality simply does not allow it.
91
ALAN DERSHOWITZ, THE BEST DEFENCE 111 (1982).
92
ATIYAH & SUMMERS, supra note 53, at 358.
phenomena, sometimes prestigious Wise Pragmatist professors are appointed to the
highest judicial positions without any previous practical experience (but with well
identifiable political sympathies), where they often bring in their sociological or economic
approaches. 93

As American legislation is sporadic, and the law making power is in a much larger extent in
the hands of judges than in other countries, and as the stare decisis is less adhered to, the
danger of legal uncertainty is more imminent. The legislator cannot encounter this legal
uncertainty because of its inefficiency, the judges cannot do it because of the (case by
case) nature of judicial law making either. So, legal scholars (Wise Pragmatists,
cooperating with judges and practitioners) have to do it, in the form of Restatements of
94
the Law. The Restatement is more creative than the Humble Clerk’s work, but it does
not have the scientific ambition of the Oracle Scientist’s writings either. 95 It just wants to
help judges, like the Wise Pragmatist always does. But it also gives the opportunity for
(some) law reforms: The case law is so chaotic, contradictory and huge that you can use
your own policy ideas to choose the “right” ones.

Even if the legislator did not agree with the way judicial case law is evolving (under the
influence of the Wise Pragmatists), it cannot counter it so easily because of its mentioned
efficiency problems and because of the constitutional constraints. 96 The constitutional
constraints themselves are in the hands of judges—“The Constitution is what the judges
say it is” 97 —and the Constitution is practically unamendable.

The balance between the legislator and the judge is very different in the U.K. and the U.S.
Consequently—and notably contradicting to some rough continental European
prejudices—the role and prestige of legal scholars is also very different (prestigious Wise
Pragmatist vs. nobody Humble Clerk):

93
Id. at 340, 345.
94
Similarly, see the work of the American Law Institute, available at http://www.ali.org/. In their self‐description,
“[t]he American Law Institute is the leading independent organization in the United States producing scholarly
work to clarify, modernize, and otherwise improve the law.” The Restatements of Law do not have any formal
validity, but are nevertheless very influential with a strong persuasive force.
95
In international law, a well‐known example of the Wise Pragmatist approach is the New Haven School. MICHAEL
REISMAN, THE VIEW FROM THE NEW HAVEN SCHOOL OF INTERNATIONAL LAW: INTERNATIONAL LAW IN CONTEMPORARY
PERSPECTIVE (1992). This approach never had any success in the Oracle Scientist dominated German international
legal scholarship.
96
ATIYAH & SUMMERS, supra note 53, at 269.
97
BERNARD SCHWARTZ: CONSTITUTIONAL LAW, VII (1972) (citing Charles Evans Hughes, Speech at Elmira, May 3, 1907).
2011]
Perhaps nothing illustrates the gulf which still

separates English and American law schools more than
they way in which England’s premier law journal, The
Law Quarterly Review, has been taken over by the
practising profession, being edited from ‘chambers’ in
London by a Circuit judge. We doubt if anything short
of a constitutional amendment could shift the control
of the Harvard Law Review to a Wall Street firm of
attorneys! 98

Another factor helping the influence of the Wise Pragmatist is that American federal
judges have law clerks. They are fresh law school graduates who were often student
editors of law reviews and are keen to use that knowledge or their interdisciplinary
education. 99 Their interdisciplinary orientation is due partly to the fact that law is a second
degree in the U.S., so they have to have a first degree in something non‐legal, and partly to
the fact that American legal education in general tends to be more interdisciplinary (maybe
also because of the above described need for interdisciplinarity) than other countries’ legal
education.

G. Theorizing About Law and Legal Scholarship for Other Legal Scholars: The Self‐
Reflective

The Self‐Reflective writes for fellow academics; she does not even try to be useful for legal
100
practitioners. Most legal theorists—at least who do not want to make a direct impact
on how the legal order is run—belong to this category of legal scholars. 101 But also
representatives of “law and” movements or comparative lawyers would mostly count to
this group, unless they give practical advice either for the legislator or for the judges, what
would make them belong to one of the formerly mentioned role models.

98
ATIYAH & SUMMERS, supra note 53, at 388 n.5.
99
Ronald A. Cass & Jack M. Beerman, Throwing Stones at the Mudbank: The Impact of Scholarship on
Administrative Law, 45 ADMIN. L. REV. 1, 8–9 (1993). Law clerks remain only one year at the court, ensuring that
new intake always brings fresh academic air to the court.
100
Also, academics belonging to other groups sometimes themselves write Self‐Reflective works (e.g., the quoted
writings of Jhering), but they do not consider it the primary task of legal scholars. For them, it is rather a
methodological pre or meta‐work. And on the other hand, Self‐Reflectives might make short excursions into
other role perceptions.
101
In the present categorization, basically all legal theorists are Self‐Reflectives unless they have explicitly
proposed another role model for legal scholars (e.g. Bentham).
As opposed to the former models in which there is a hierarchy in prestige between legal
scholars and practitioners (the Oracle Scientist, the Wise Pragmatist, the Prophet and the
Law Reformer are higher than the practitioner, the Humble Clerk is lower), here there is no
hierarchy. The Self‐Reflective is simply playing in a different field. She is not impressed by
practitioners, and she does not want to impress practitioners. She wants to stay full time
in the ivory tower, she does not really care what is going on outside in the real world. Her
audience are either other Self‐Reflectives, or the Oracle Scientists, the Wise Pragmatists,
the Prophets, or Law Reformers. She is successful if she can influence what her audience
thinks or writes, and can be measured by the number of quotation she gets or by the fact
that she publishes in renowned and popular scholarly forums (e.g. in peer reviewed law
reviews). Maybe she does not even want to influence these other academics, but she just
102
wants to provoke them.

As she does not consider herself as part of the legal (practical) discourse, but rather as
leading a meta‐discourse about the legal discourse, she is not limited to legal arguments.
She can use moral, economic or philosophical (philosophy of science, etc.) arguments. 103
As her ideas are bound to the changing law, but are very general, she aims to write for
eternity, and her chances to be readable long term are actually better, than the similar
ambitions of the Oracle Scientist or of the Prophet.

You can find such legal scholars (they do not necessarily consider themselves as legal
theorists) in most law schools, but the likelihood of their existence is higher, if there is a
regular bureaucratized scholarship assessment. In the latter case, legal academics will
evaluate the work of legal academics and they obviously appreciate more works which
have been addressed explicitly to them. Thus, academics will want to impress only fellow
academics—who decide about their tenure—and do not care at all about practicability.

The phenomenon is well‐known in the U.S., where they talk about the growing gap
104
between legal academia, including legal education, and the legal profession. The legal
scholarship is rarely quoted by courts and not respected by legal academics, and the
respected works are absolutely useless for practitioners’ use. 105 Statistically, there is a
102
This provocation might consist of destroying or at least pointing out the self‐deception of Oracle Scientists
about their neutrality and objectivity. COWNIE, supra note 65, at 51–53 (discussing Critical Legal Studies from this
perspective).
103
Interdisciplinary and multidisciplinary approaches are also used. GUNNAR FOLKE SCHUPPERT, STAATSWISSENSCHAFT
(2003); GUNNAR FOLKE SCHUPPERT, VERWALTUNGSWISSENSCHAFT (2000).
104
Harry T. Edwards, The Growing Disjunction between Legal Education and the Legal Profession, 91 MICH. L. REV.
34, 34–78 (1992); Twining et al., supra note 6, at 931. Another reason for the gap in the U.S. is that law schools
are rather on the political left, whereas judges are rather conservative. Twining et al., supra note 6, at 935.
105
Deborah Jones Merritt & Melanie Putnam, Judges and Scholars: Do Courts and Scholarly Journals Cite the
Same Law Review Articles?, 71 CHI.‐KENT L. REV. 871, 871–99 (1996). The answer to the question of their article
title according to the authors is a clear no.
2011]
clear decline in judicial quotes of scholarly works, 106 which can also be explained by the
popular topics chosen by newer generations of legal academics.

In the U.K., where for a long time such mechanisms did not exist, this gap was unknown, 107
but the RAE (Research Assessment Exercise) 108 is pushing legal scholarship in this
direction. 109 The unfavorable side effect of such changes is that less talented people will
work as Oracle Scientists, as Humble Clerks, or as Wise Pragmatists, thus that type of black
letter legal research is fading. 110

Another problem is that the self‐perpetuating, self‐referential, and closed discourse of the
Self‐Reflectives will begin to live its own life. You are successful in the discourse if you are
successful in the discourse. This tautological way of measuring success results in try‐hard
originality, sexy titles (which catch the eyes), provocative quotes, and counter‐intuitive
theses. Such papers will probably never be read by anyone. The latter seems to be a
failure for the Self‐Reflective (and to a certain extent it is), but in a bureaucratized
scholarship assessment system it can become secondary: once the output is there (in a
good law review), its content or influence (which is more difficult to measure) countless.
So the main point in publications will be to look like something very original, paradigm‐
111
shifting and brilliant in order to be accepted for publication. In the U.S., where such
behavior is flourishing, “[a] law professor with no formal qualifications in philosophy might
undertake a project of remarkable philosophical ambition and publish it in a journal edited
by individuals who not only are equally unlikely to have any philosophical qualifications but
who have yet to become qualified in law.” 112 As a matter of fact, you just have to look like
a deep thinker, you have to sell your product once, you don’t give any guarantees, and
then you can go to the next law review. 113 The best ones can aim for a successful
106
Louis J. Sirico Jr, The Citing of Law Reviews by the Supreme Court: 1971–99, 75 IND. L.J. 1009 (2000).
107
108
Since 1986, approximately every five years, the quality of research at UK higher education institutions will be
evaluated on behalf of the higher education funding councils. The core of the evaluation is that full‐time
members of the higher education institutions submit a limited number of writings which will be read by other
legal academics and evaluated accordingly. For more information see http://www.rae.ac.uk/. The name of the
system is changing, but its relevant content (as to the topic of this paper) remains the same, see
http://www.hefce.ac.uk/Research/ref/.
109
COWNIE, supra note 65, at 136; Twining et al., supra note 6, at 925.
110
111
Daniel A. Farber, The Case Against Brilliance, 70 MINN. L. REV. 917 (1986).
112
Neil Duxbury, A Century of Legal Studies, in THE OXFORD HANDBOOK OF LEGAL STUDIES 957 (Peter Cane & Mark
Tushnet eds., 2003).
113
As opposed to Alexander Somek, I do not think that the general belief would be that the “production of law
review articles is in some manner, however causally obscure, a way of improving the world.” A considerable
provocation in the debate (and not the solution of any problem), but the discourse is just
so big, that your voice seems hopelessly low in the loud crowd. It looks like you work for
eternity, but you just work for your next publication, because this is expected from you at
the university. From the ambitioned eternal truth, so will be an unreadable amount of
mediocre interdisciplinary (“law and”) scholarship – with only a few quality exceptions. 114

If you also measure influence by the number of cites, then the picture looks a bit different.
The cite‐counting method has been heavily criticized as it does not give a real picture of
the quality of legal scholarship one produces. 115 Sometimes you get cited several times,
because everybody wants to show how wrong you were. Sometimes you get cites only
because the other scholar wants to fill up footnotes. Often treatises are quoted, which just
summarize the law, but which do not offer any new ideas. And if an article becomes
widely accepted, then the distortion can work in both ways: either you will be quoted by
everybody, 116 or you will not be quoted because your opinion is so widely accepted that it
is not any more connected to you personally. 117
Even a kind of partly ironic self‐help literature has evolved to analyze the nature of
citations and to give advice on how to win cites. The advice is the following:

Maxim One: (Make sure that you have already)
Attend(ed) Harvard, Yale, or the University of Chicago
Law Schools. . . . Maxim Two: Publish all of your
articles in the Harvard Law Review, the Yale Law
Journal, or the University of Chicago Law Review. . . .
Maxim Three: Take a job as an assistant professor at
the Harvard, Yale, or University of Chicago Law
118
School.

If the three maxims are followed, then your articles are much more likely to get cites than
otherwise. As a matter of fact, you do not get positions at good universities because your
portion of U.S. law reviews contain basically either Self‐Reflective papers, which are a Selbstzweck with no—not
even a pretend—usefulness for the world outside academia. The rest of the papers fit Professor Somek’s
description, and he is right in the sense that the majority of papers falls into his category.
114
For a similar view, see Edwards, supra note 104, at 36; Kenneth Lason, Scholarship Amok: Excesses in the
Pursuit of Truth and Tenure, 103 HARV. L. REV. 926 (1990).
115
Cass & Beerman, supra note 99, at 2–3.
116
J.M. Balkin & Sanford Levinson, How to Win Cites and Influence People, 71 CHI.‐KENT L. REV. 843, 845 (1996) (“By
gaining an increasing presence in that [discourse] space, the canonical work may create an increasingly hospitable
environment for its own reproduction in the minds of future academics.”).
117
Cass & Beerman, supra note 99, at 3.
118
Balkin & Levinson, supra note 116, at 843–69.
2011]
articles are often quoted, but rather, the other way around: You get quoted because you
have a position at a prestigious university. Students or young scholars aiming for tenure
will quote you, because they cannot judge the actual quality of articles, so they go for the
proxies (i.e., who published it and in which journal). 119 After they get a tenure, their
productivity will normally decline so the real market for cites is not the one of tenured
professors (for whom very often students do the research anyway), but rather students
and assistant professors.

You should also forget about what is important for the real world; you should rather ask
yourself what can be quoted: “If we have one basic piece of advice about topic selection,
it’s this: Never confuse what’s important in the world outside law schools with what is
120
important in law reviews.”

Even if self‐cites do not count in most systems, it is still useful to advertise your own
writings wherever it is possible, so you should not shy away from that. Thus for a
conscious and successful impression management you need some shamelessness:

(1) Cite yourself, early and often. (2) Get your friends
to cite you whenever you can. . . . We realize that it
takes a bit of chutzpah to shamelessly self‐cite. But
after a while, you’ll get over it. Believe us, many other
people in the legal academy already have. . . . Friends
are usually more than happy to cite you, especially if
you offer to cite them in return. Sometimes, however,
they need a bit of cajoling doing the right thing. Make
your friends feel guilty if they don’t cite you in all of
their articles. Tell them how hurt you are that they
are neglecting you and your ideas. If all else fails,
accuse them of insensitivity, plagiarism, or worse.
Sure it may strain the friendship, but aren’t the extra
121
cites worth it?

And finally, the actual content of the article counts even less, than we like to admit it. The
article rather needs new keywords, or it has to be able to symbolize a movement, an
approach or a political stance, i.e. it has to become an “icon.” If you manage to catch this
role, than your article will become canonical and get cited again and again—often without
119
Id. at 860.
120
Id. at 855.
121
Id. at 856–57, 859.
being read. 122 So, when you choose topic or keywords, you should bear this feature of
legal scholarly discourse in mind.

H. Telling the Public What the Law Is: The Media Star

In every country we know different legal scholars turning up on TV or being interviewed
regularly in newspapers. Some of them do have an otherwise respected scholarly
activity—Alan Dershowitz as a Wise Pragmatist; 123 others just do not—the latter ones
should remain unnamed out of courtesy.

The Media Star has to be understandable, even at the cost of being precise, in cases he
even has to be funny. He has to understand what the people actually do not understand.
He has an important function in bringing law closer to the public, thus making it more
transparent or even democratic in an epistemological sense. It helps to run the legal
system either by making addressees aware of the law or by making them believe that the
law makes sense and that it is not a secret and intrinsic knowledge. He is not running the
legal system from inside as a Media Star, but is only helping outsiders understand what is
happening inside.

Unfortunately, their colleagues do not really respect him—or at least not for his being a
media star, but for his other scholarly activities—maybe because his audience are not even
lawyers and because he consequently has to be imprecise in order to remain
understandable to lay persons. It is also a dangerous function in the sense that the prompt
opinions required by the media do not leave enough time to think about them thoroughly,
so even the best scholars make mistakes more easily and damage thus their own scholarly
reputation.

The Media Star might even express his or the general public’s wishes to change the law,
and it might even have an effect, if politicians think that leaving unchanged the law would
outrage the public. As opposed to the Law Reformer he does not give his advice in person
to politicians about such changes, he rather proclaims in the media what the people want.
He can be considered as successful if he can influence public debates and if his media
coverage is high. His views are aimed at the moment; nobody expects them to be useful
long term.

If the Media Star proposes to change the law, then she is usually using non‐legal
arguments. If, however, she simply explains what the law is, then she remains within the
strict limits of legal reasoning. The Media Star can exist in countries where the press is free
122
Id. at 861–62.
123
For a self‐promoting account of his media activities, see www.alandershowitz.com.
2011]
and where the public is interested, but also in dictatorships if the regime decides that
population needs some legal education through TV, etc.

The Media Star has nowhere near a full‐time job; there is no country where he would be
dominant amongst legal academics. Even if somebody is fulfilling this function as a legal
scholar, he or she fulfils one of the other role models as mentioned above. If not, then she
or he is usually not even considered as a legal scholar, only as a media figure. By
definition, there are only a few Media Star legal scholars in every country. He is a Media
Star exactly because the media coverage possibilities are limited and he managed to catch
a considerable part of it.

I. Which One Is the Ideal Legal Scholar?

Asking which of the above role models is “the” ideal is similar to asking which football
player is the ideal one: the goalkeeper, the defender, the midfielder or the forward. They
do different jobs, but they could all contribute to a “better common result.” In a legal
system, the role of legal scholars could be perceived as similarly manifold, but—just like in
a football team—you have to be careful about role proportions. A team consisting of ten
goalkeepers is unlikely to win. And a legal order full of Prophets is unlikely to function
properly. So the question should be: What is the ideal proportion of role models?

But as we have seen, it is not even about ‘universal’ proportions, as much depends on the
institutional context and on the potential audience of the respective country where the
legal scholar is working. Some role models are just not needed at all in certain legal
124
cultures, while other role models are needed a bit less or a bit more. The “real” task of
legal scholars is thus only partly an issue of legal theory; it is in an even bigger part rather a
problem of comparative law. In continental European countries the Oracle Scientist seems
to be the typical and mostly needed legal scholar; in the U.K., the Humble Servant; in the
U.S., the Wise Pragmatist; and in moderate dictatorships or post‐dictatorial regimes the
Law Reformer; though in very small numbers, we can discover almost every (but not every)
role model in the named countries.

The Self‐Reflective seems to be the most international role of all—as he does not try to
help to run a legal system which is different from his original legal culture—so this is the
role one can take up the most easily when changing country as a legal scholar. Self‐
Reflectives can simply give new insights into general questions of law or they can coach
125
other legal scholars. If she is in her original legal culture, then the second seems like a
124
Globalization means that certain legal institutions or rules are borrowed from each other, but basic mentalities
(as described in this Article) will continue to differ in different legal cultures. Cf. Legrand, supra note 51, at 52–81.
125
Cf. Peter Lerche, Maunz/Dürig, Grundgesetz, in RECHTSWISSENSCHAFT UND RECHTSLITERATUR IM 20. JAHRHUNDERT,
1026 (Dietmar Willoweit ed., 2007) (describing “theory based practicality” (theoriegestützte Praxisgewandtheit)
more often needed task. But nobody can expect a truly European scholar in the U.S. to
train Wise Pragmatists, not because he could not, but because it would contradict his
whole ars poetica. And on the other hand, we do not invite American scholars to Europe
to coach us about methodological problems of Rechtsdogmatik. We do it because we
want to exercise our brains by seeing something which is very different from what we
believe in, and besides satisfying the pure curiosity about American law, we also hope to
understand our very own differing role perception as a legal scholar a little bit better.
as the ideal type of commentary, a typical genre of the Oracle Scientists); Mátyás Bódig, Legal Theory and Legal
Doctrinal Scholarship, 23 CAN. J.L. & JURIS. 483, 494 (2010) (indicating doctrinal legal scholars are the main
audience of legal theorists).

Annex

Audience Ideal Perversion Aimed Prestige in Measure of Time scale Non‐Legal Precondition of Typical
Function Influence the General Success (Moral or Existence Countries and
on the Legal Policy) Lawyers
Law Community Arguments
and Amongst
Other
Scholars
Oracle academics, furthering forcing artificial high high perfectness aimed for very limited, doctrinally well German
Scientist judges legal conceptual of the eternity, disguised, trained judiciary (Austrian,
certainty by absurdities on system but in rather Swiss, Italian,
building a real life, losing practice through and Spanish)
conceptual practical medium vague legal law
system relevance term concepts professors,
Langdell,
Hohfeld
Prophet judges, community hijacking law by high rather high influence aimed for substantial prestigious Dworkin,
academics building in a own preferences on judges eternity, (moral judiciary which human rights
moral sense behind the mask but in arguments) dares to use litigator
of moral dignity practice also not strictly professors
medium legal
term considerations;
constitutional
limits on the
legislature
Law politicians promoting extreme high medium influence medium substantial central Bentham,
Reformer (legislators) social instrumentalism, on (until the legislator which (post)socialist
development serving politicians social is able and countries
or reform dictatorships (legislator) problem willing to make
and real life has been new laws
success of solved)
the
proposed
law
Humble attorneys, documenting mindless near to low how rather zero growing Kirchmann,
Clerk law what the law repetition of zero precisely short, until amount of law, English black
students is (and not what judges or s/he notes the next and the letter
what it ought legislators said and repeats relevant following need scholarship
to be) the law judicial to document it
(typically decision
judicial
decision)

Audience Ideal Function Perversion Aimed Prestige in the Measure of Time scale Non‐Legal Precondition of Typical
Influence General Legal Success (Moral or Existence Countries
on the Community and Policy) and Lawyers
Law Amongst Other Arguments
Scholars
Wise judges advising armchair high high influence medium substantial prestigious Posner,
Pragmatist judges about sociologist or a on judges (policy judiciary which traditional US
socially best Prophet arguments) dares to use legal
decisions disguised as also not strictly scholarship
social legal (legal
scientist; cynic considerations; realists), law
about the efficiency & economics
normativity of problems with
law the legislature
Self‐ fellow understanding self‐ zero, or rather high (if publications aimed for substantial no specific Kelsen (in his
Reflective academics better general perpetuating collateral understandable) in eternity, preconditions, theoretical
questions of community influence esteemed but in but writings),
law; coaching without any (e.g., peer practice bureaucratized Somek,
other types of practical use reviewed) mostly scholarship Jakab, major
legal scholars for the legal forums, rather assessment part of
system influence medium induces it modern legal
on term scholarship at
academic top US
discourse universities
(quotations)
Media general explaining the pure media medium low media very short it depends technical Alan
Star public people what figure without coverage, development; Dershowitz;
the law is, so relevant influence either free nowhere
helping them scholarly work on public press and dominant,
to accept and and without political interested only as a
follow it real debates public, or a part‐time job;
knowledge dictatorial parallel work
regime deciding in any of the
for general above
legal education categories
through TV

Special Issue
Kelsen, Sander, and the Gegenstandsproblem of Legal Science

By Christoph Kletzer ∗

A. Introduction

One of the main problems which has emerged in recent years in the debates of legal
positivism has been a rather defensive twist in its self‐understanding (i.e., its self‐
affirmation in terms of what it is not). Now, whereas such a negative approach does not in
itself pose a problem, in the case of legal positivism it has led to a series of rearguard
battles against claims stemming, on the one hand, in more general philosophic
developments, and, on the other, in doctrinal legal scholarship. The result has been that
some have wondered if between those conceptual and institutional demarcations,
between the philosophic and the juridical departments, there actually remains anything for
positivism to stand for. Is there anything it actually claims? Accordingly, a perceived
dilemma of positivism has emerged along the following lines: insofar as legal positivism
makes a sound claim, this claim is very weak and, by itself, not very interesting; insofar,
however, as positivism tries to make a strong and interesting claim, this claim can be
shown to be fundamentally misguided.

However, the described problems do not follow from any genuinely positivistic
commitments themselves, but are the result of a historical rooting of the Anglo‐American
strand of legal positivism in a not thoroughly reflected empiricism and in the lack of a
proper engagement with the “problem of being an object of cognition in general,” or what
in the Continental tradition has been called the Gegenstandsproblem of legal science.

This paper tries to introduce this problem by means of outlining the heated debate
between Hans Kelsen and his student Fritz Sander on this issue, and it submits that it is this
early engagement of the Pure Theory with fundamental philosophical questions of object‐
relation which has ensured the non‐naive, non‐empiricist character of the Pure Theory.
The early engagement with the Gegenstandsproblem has, so to speak, immunized the Pure
Theory against the disease of philosophic shallowness which has left contemporary, Anglo‐
1
American positivism defenseless against even the most perfunctory attack.
∗
Lecturer in Legal Philosophy, King’s College London, e‐mail: christoph.kletzer@kcl.ac.uk.
1
Surprisingly, it has been precisely this philosophic sophistication of the Pure Theory which has led many
theorists to avoid it and has driven them to the mixture of quasi‐philosophy and quasi‐sociology, which currently
makes up the tradition of Hartian positivism. Nevertheless, the same philosophic sophistication of the Pure
Theory has, in recent years, also sparked a renewed interest in Kelsen and even something of a Kelsen
renaissance in the Anglo‐American world of jurisprudence and legal theory. For evidence, see only the lively

B. The Gegenstandsproblem

The Gegenstandsproblem deserves a technical term because the better part of
contemporary jurisprudence takes, as its object, the positive law sans phrase, and does not
very often consider there to be such a thing as a Gegenstandsproblem, but rather, takes a
straightforwardly empiricist approach. 2

This approach is especially surprising as, its antitype, Hartian positivism, apart from being a
theory of the positive law, is so obviously also a theory of the division of labor between
itself and the positive law: Hartian positivism teaches that it lies within the concept of the
positive law that the positive law determines its own content. Implicit in this claim is the
claim that it is for legal philosophy to determine the concept of the law. This is seen as an
important advance from natural law theories which claimed competence over both
concept and content of the law. The advance, it is believed, lies in the self‐restraint of
philosophy, in the effort to let the law be.

Now, this division of labor is a peculiar intellectual maneuver. At the same time, it
expresses supreme intellectual self‐restraint, and a hidden but vast arrogation of
competence. Relying on the self‐determinative powers of the law, it withholds judgment
about the content of the law but it can do so only in holding on to a much more sweeping
competence, namely, the competence to determine competences, or what can be called
the Kompetenz‐Kompetenz. So there are three competences involved: the competence to
determine the content, the competence to determine the concept, and the competence to
determine the competences.

The problem of a large part of contemporary jurisprudence consists in not dealing with the
problem of this third competence and in not seeing a problem with the implicit arrogation
of it. It is believed that positivism can determine its own limits from within, so to speak,
that is without already overstepping those limits when setting them. However, the limits
of jurisprudence can only be defined by claiming competence to set those limits; and
whereas the limited competence of positivist jurisprudence is in constant view of
3
jurisprudence, the limiting competence has remained something of a blind spot.
debate ignited by a blog‐post from Michael Steven Green, in which scholars like Brian Leiter, John Gardner,
Anthony D’Amato (if their web‐signature can be trusted and the posts were really from them) and many others
have engaged. To follow the debate, see http://leiterlegalphilosophy.typepad.com/leiter/2007/10/michael‐
green‐a.html or http://lsolum.typepad.com/legaltheory/2007/10/should‐we‐study.html. See also Michael Steven
Green, Hans Kelsen and the Logic of Legal Systems, 54 ALA. L. REV. 2 (2003).
2
See, e.g., Green, supra note 1.
3
The dilemma of naive positivism’s failure to be positivist enough is of course isomorphic to the dilemma of naive
liberalism’s failure to be liberal enough to accept illiberal substantive moral theories. Liberalism and positivism
have a conjoined fate. Both are attempts of philosophy to “let the world be” (in both the extreme quietist and
2011]
Gegenstandsproblem 787

To put it more casually, there is an odd intrusiveness about the seeming positivist self‐
restraint, in a way similar to a stranger telling you, “I am not going to order you to give me
money.” As if he could!

But how does Kelsen avoid this trouble? One way to think about this issue is to claim that
Kelsen guarded himself against the traps of empiricism by sticking to a form of Kantian or
Neo‐Kantian “conceptualism.” A classic formulation of this approach has been provided by
4
Cotterrell.

Cotterrell takes empiricism to be the attempt to represent the law as something the truth
of which can be ascertained by observation. A theory, according to this empiricism, is a
“direct representation of empirical reality, with its concepts derived from observation of
and generalisation about that reality and so corresponding with it and testable for truth
against it.” 5

Conceptualism, in contrast, starts from the following insight:

Empirical reality—the world of objects and experiences
“out there”—does not, in fact, present us with
evidence which we can merely package together or
generalise about to arrive at scientific truth. Concepts
need to be formed in advance—a priori—in order to
organise empirical evidence. The previously
established concepts not only determine what is
empirically relevant but also reflect a view of why it is
relevant. Thus, theory aiming at scientific explanation
of any object of knowledge cannot take its concepts
from observed experience but must deliberately
construct concepts as a means of interpreting
experience, of imposing order on it. A theory is not an
attempted representation of observable reality but an
intellectual construction—a logically worked out
model—which can be used to organise the study of
what can be observed in experience . . . Whereas an
the metaphysic sense of the phrase) and to demarcate a limit of philosophy from within. The reflexivity of this
approach (i.e., the degree of its non‐naivety) will be the sole criterion of the success or failure of both positivism
and liberalism.
4
See ROGER B.M. COTTERRELL, THE POLITICS OF JURISPRUDENCE: A CRITICAL INTRODUCTION TO LEGAL PHILOSOPHY (1992);
Green, supra note 1.
5
COTTERRELL, supra note 4, at 85.
empiricist view would say that a theory’s truth can be

tested in the light of experience, a conceptualist would
claim that it is its usefulness, not truth which is the
issue. 6

Now, Cotterrell is clearly right in describing Hart as an empiricist; and he does a masterful
job in illustrating how, despite Hart’s early attack on empiricism, 7 Hart nevertheless
“institutes a new kind of empiricism in place of the old.” 8

This new kind of empiricism is “linguistic,” and, as is well known, it rests on some
Wittgensteinian and J.L. Austinian insights. According to Cotterrell, Hart refuses the old
empiricism and its convictions that legal concepts do have a uniform, stable meaning:
“Linguistic philosophy, in this form, focused not on the meaning of words in some
definitional manner, but on clarifying the way in which words are used in various linguistic
9
contexts.”

Instead of asking what legal terms designate, the task now was to find out how those
terms are used in a legal system:

And this is—although Hart never calls it such—a new
kind of empiricism because what is required is a “close
examination of the way in which statements e.g., of
legal rights or of the duties of a limited company relate
to the world in conjunction with legal rules.” It is, thus,
necessary to examine the actual conditions under
which such statements are regarded as true. The
observable reality which legal statements represent is
not a range of identifiable entities which are referred to
by words such as “corporation.” It is the reality of
linguistic practices of people living within a legal system
and orienting their conduct and expectations in relation
10
to it.

6
Id. at 86.
7
The viewpoint of Hart’s inaugural lecture of 1953 suggests an attack on empiricism because concepts are no
longer to be seen as representing anything in a one‐to‐one fashion. The meaning of a legal concept, according to
this view, cannot be defined as if the concept represented some invariant state of affairs.
8
COTTERRELL, supra note 4, at 91.
9
Id. at 90.
10
Id. at 91.
2011]
According to his focus on the performative nature of language, Hart is not any more
concerned with terms and concepts as representations of social reality, but with those
words themselves as social reality. We can do things with words. Words themselves can
already be deeds.

So much for Hart’s empiricism, which Cotterrell describes precisely. However, I think
Cotterrell errs in describing Kelsen as a conceptualist along Neo‐Kantian lines. Cotterrell
describes this Neo‐Kantianism as follows:

We only begin to understand the empirical reality by
imposing concepts on it which enable us to organise as
meaningful what we observe. Concepts do not reflect
experience; they organise it and make it intelligible.
Every science, every knowledge field, must, therefore,
create its own conceptual apparatus. Because of this
necessity, each science, or form of systematic
knowledge, is unique and distinct from all others.
Consequently, legal science must have its own unique
framework of concepts which cannot be shared or
integrated with those of other sciences. It follows,
therefore, that Kelsen wholly rejects what he calls a
11
syncretism of methods.

Unfortunately, the case is much more complicated than this.

As I hope to be able to show, at least from his engagement with Sander onward, Kelsen
cannot any more be described as a “conceptualist” in Cotterrell’s sense. Moreover, even
before the engagement with Sander, Kelsen was less of a clear‐cut conceptualist, and more
a Humean empiricist masquerading as a Kantian.

C. Sander’s Intervention

Early in his career, Kelsen stumbled into an intellectual struggle with Fritz Sander, one of
his most promising students and campaigners, in which Kelsen let some of his other
students, namely Emanuel Winternitz and Felix Kaufmann, do a lot of the writing and
arguing. This struggle ultimately led to an ugly personal and legal battle which had to be
resolved in front of an academic arbitration committee. Kelsen felt that Sander had
accused him of plagiarism and had claimed things which, if true, would be incompatible
with his reputation as a scholar and academic teacher. Sander, in turn, felt that Kelsen had
been scheming and arrogantly avoiding the real intellectual issues by continuing to treat
11
Id. at 107.
him as a disciple, and failing to notice that he had actually already incorporated into his
work more of Sander’s insights than he was willing to concede or himself had even noticed.
Much of the conflict is, of course, more of a psychological or psychoanalytical interest, as
similar phenomena can be found in the circles and schools of other “great intellectual
fathers.”

However, underneath the unpleasantness of the anecdotal and historiographical Kelsen‐
Sander controversy, there lies an exciting intellectual struggle about the possibility of
rationally relating to the positive law to be uncovered.

What somewhat complicates a contemporary reception of the Kelsen‐Sander debate is the
fact that the debate was conducted nearly entirely in Neo‐Kantian terms, terms that now
sound cumbersome and antiquated to us. Still, on reflection, those terms are actually
rather helpful insofar as they help us detach the debate form contemporary empiricist
commonplaces.

The Neo‐Kantian background that informed the debate can be expressed as set of a more
or less diffuse transcendental convictions: (α) in knowing an object we in a way construct
it; (β) the human sciences have an analogous structure to the natural sciences; (γ) legal
science is a human science.

Even though I do think that most of the authentic Kantian insights that are involved in
these Neo‐Kantian convictions are, indeed, correct, we do not have to accept Kantian or
Neo‐Kantian epistemology here in order to follow the debate. All we need to accept is that
knowledge is not a purely passive phenomenon (i.e., that there is at least a minimal sense
in which in knowing an object we are not only determined by the object but also
determine it).

As already touched upon in the discussion of Cotterrell, whatever we know to be true, it is
at least the respect in which we know it to be true that depends on our making and on our
approach to the object of knowledge. This is what Rickert claimed in his famous
statement, that “the same empirical reality . . . becomes nature when we view it in respect
to its universal characteristics; it becomes history when we view it as particular and
12
individual.”

I do not take this to be a particularly controversial claim, especially since it does not
include the more radical idealist claim that the content of our knowledge is constructed by
our act of cognition.

12
HEINRICH RICKERT, DIE GRENZE DER NATURWISSENSCHAFTLICHEN BEGRIFFSBILDUNG: EINE LOGISCHE EINLEITUNG IN DIE
HISTORISCHE WISSENSCHAFT 234 (1902).
2011]
Now, as it concerns our knowledge of nature, we have to distinguish three levels: (1)
nature; (2) natural science; and (3) critical philosophy. Critical philosophy teaches that
nature is not simply given to natural sciences as a unified and comprehensively determined
whole, but that the unity of nature, and thus the “natureness” of nature, 13 has to be
constructed. This construction is effected in the natural sciences. In determining natural
laws, the sciences establish a concrete universe, or a continuous relation, and thus unity
within nature.

The question is, of course, whether the law, legal science, 14 and legal philosophy stand in a
similar relation.

Level 2 Philosophy Legal Philosophy
critical
Level 1 Natural Science Legal Science
constitutive
Ground Level Nature Law
Table 1: Classical Model

Kelsen started off his academic career with the vague presentiment that they do, even
though he seemed reluctant to follow through the analogy. Maybe this was the case
because Sander, in a kind of fit of anticipatory obedience to his master, outdid Kelsen and
published an article in which he pushed the analogy between natural and legal science to
an extreme. 15
13
IMMANUEL KANT, CRITIQUE OF PURE REASON 466, A 418, B 446 (Paul Guyer & Allen W. Wood eds., 1998)(defining this
material nature (natura materialiter spectata), as “the sum of appearances is so far as they stand, in virtue of an
inner principle of causality, in thoroughgoing interconnection.”).
14
By “legal science” I refer to doctrinal legal scholarship, or legal research (i.e., what we do in law schools and get
money for from research bodies and the government). The Germans call it “Rechtswissenschaft.” The term “legal
science” is generally avoided by Anglo‐American lawyers and some think this has to do with the narrower
meaning of the English “science” in comparison with the German “Wissenschaft.” But I tend to think the problem
has its roots in the differences of Continental and Anglo‐American legal education stemming in a difference of
reception of Roman Law.
15
Sander tried to defend the Pure Theory against attacks by the legal sociologist Bernhard Stark, who in a
polemical article insisted that legal science has to be conducted as a science of facts (i.e., as sociology). Doctrinal
legal studies, conversely, have to be “nihilised,” just as astrology and alchemy have been previously. See
Bernhard Stark, Die jungösterreichische Schule der Rechtswissenschaft und die naturwissenschaftliche Methode, in
DIE ROLLE DES NEUKANTIANISMUS IN DER REINEN RECHTSLEHRE 422 (Stanley Paulson ed., 1988); see also Fritz Sander,
Rechtswissenschaft und Materialismus, 47 JURISTISCHE BLÄTTER (1918).

Kelsen reacted coldly to Sander’s initiative. He took a backseat and let the others do the
adventurous work. In the meantime, however, Sander had made a momentous turn in his
assessment of the relation of legal and natural sciences. His new position begins with the
newly found insight that “legal science has at all times been under the misapprehension
that the positive law stands in an analogous relation to it as nature does to natural
science.” 16

He wonders whether or not a “transcendental philosophy was conceivable that would set
as its task a critique of the law itself and not of legal science . . . Notably, a transcendental
critique and not an empirical critique, which would be natural law.” 17 And, finally, he
comes to an affirmative assessment:

Not legal science but the law corresponds to the
transcendental, constitutive sphere of “cognition:” the
synthetic judgements of the law constitute the analogy
to the synthetic judgements of the mathematical
natural sciences. The relation of legal science to the
law, conversely, corresponds not to the constitutive
relation of natural science to “nature” but to the
reflective relation of transcendental logic (the theory of
experience) to mathematical natural sciences. Legal
science is possible only in reflective relation to the fact
of law, it is possible only as a theory of legal experience.

Thus, as soon as its metaphysical (natural law) content is
destroyed and its political content is discharged, legal
science is essentially philosophy of law . . . . With this
determination of the merely ancillary, reflective method
of legal science, with the establishment that the
sovereign, constitutively determining and thus creating
syntheses of the law take place exclusively within the
sphere of the objective law, in the legal process an
unshiftable line is drawn between natural law and the
18
positive law.
16
FRITZ SANDER, RECHTSDOGMATIK ODER THEORIE DER RECHTSERFAHRUNG? KRITISCHE STUDIE ZUR RECHTSLEHRE HANS KELSENS 93
(1921).
17
Fritz Sander, Die transzendentale Methode der Rechtsphilosophie und der Begriff der Rechtserfahrung, 1
ZEITSCHRIFT FÜR ÖFFENTLICHES RECHT 468, 476 (1919–20).
18
SANDER, supra note 16, at 94–95.
2011]
The relation which Sander proposes thus shifts the right hand stack of relations one level
up, and introduces “legal facts” as the new ground level.

Legal Philosophy
?
Level 2 Philosophy Legal Science

critical critical ?
Level 1 Natural Science Law
constitutive constitutive
Ground Level Nature Legal Facts

Table 2: Sander’s Model

Sander’s revolution thus involves three new and momentous theses, the relevance of
which cannot possibly be overstated: (1) legal science essentially is philosophy of law; (2)
the positive law is cognitive; and (3) legal facts are not given, but have to be constituted.

In what follows we will focus on the second thesis, the cognitivity thesis, as it is clearly the
most puzzling of the three.

What brought Sander to such a radical claim is the insight that the positive law is neither a
purely factual occurrence nor merely a complex of norms, but rather, that it is the relation
of both. The legal process, or the legally regulated application of law, is this relation of fact
and norm. The legal process produces judgments which synthesize given facts by
employing spontaneously produced concepts into legally relevant facts. From here, it is
only a small step to claim that it is the legal process itself, and not legal science, which is
analogous to the process of cognition in the Kantian architectonic.

This turn of localizing the relevant act of constitution, not in legal science but in the law
itself, must at first, of course, seem highly counterintuitive. In fact, Sander, throughout his
academic career, has not succeeded in underlaying what must seem like little more than a
learned intuition with a philosophically satisfactory theory. However, what implicitly
underlies this learned intuition is an essential deepening of the understanding of the
Kantian achievement. Sander’s revolution is based on an insight that was too often
forgotten in Kelsenian circles, namely, the insight that despite some wavering and a certain
slackness in language. Kant’s most profound contribution to philosophy was not the
delivery of a psychology of knowledge, nor the definition of the substratum of cognition or
the substance of knowledge; but the layout of his critical philosophy as an inquiry into the
claims regarding the validity of knowledge, into the normative dimension of knowledge. 19
After all, the theme of the transcendental deduction is not the psychological problem of
the empirical conditions of the possibility of any synthesis, but the problem of the
transcendental conditions of possibility—in short, the question as to the possible validity
of synthesis.

The fact that “valid synthesis” is then determined as “experience” in the individual subject
appears as a mere terminological fact. In the law, too, such a synthesis takes place, a
synthesis, by the way, which in contrast to the syntheses of legal science, can bear
comparison to the synthesis of natural science: by means of its legal norms, the law
autonomously and spontaneously determines the entire realm of legally relevant facts. It
is only within the continuous connection of the legal process
(“Rechtsverfahrenszusammenhang”) that the extralegal fact is integrated with the legal
norm to a judgment which determines both. Facts without legal norms are irrelevant or,
as Kant would have said, “blind.” And legal norms without facts to which they can be
applied are ineffective or “empty.” In any case, the relevant synthesis is affected in the law
itself and not in legal science, since it is the law which determines which fact has which
legal relevance.

Insofar, Sander has surely corrected the skewed analogy of natural and legal science:
whereas nature as an ordered whole is inconceivable without an experience that orders
the given material, the positive law as an ordered whole is very well conceivable without
the existence of legal science; whereupon again the ordered whole of legally relevant facts
would be inconceivable without the positive law.

So Sander does seem to have followed by all means a plausible intuition. However, he
lacked the intellectual means to philosophically capitalize on it. For, even though his
insights may have led him to a locus classicus of post‐Kantian philosophy, so to speak, into
the middle of the turn of Kantianism via Fichte to Hegel, he nevertheless remained dazed
in the dark of his own insights.

He does not realize that according to his own intellectual progress, the problem of the law
runs parallel to the problem of the “you” in idealist philosophy insofar as, in both cases,
the constitutive gaze of the subject meets an “object” which in its self‐constitution resists
the constitution by the subject. As soon as I subject a “you” to the machinery of
categorical determination, I have turned the “you” into an object, which by definition is
subject to the comprehensive principle of causality and thus also subject to the
mathematical natural sciences. From such an object, the personal “you” cannot be
19
See ROBERT B. BRANDOM, MAKING IT EXPLICIT: REASONING, REPRESENTING, AND DISCURSIVE COMMITMENT 9 (1994).
2011]
derived. Only in suspending my own constitutive force and, as was first demonstrated by
Fichte against Kant, in recognizing the self‐constitution of the other can I ever be able to
meet the other in terms of cognition.

D. The Criterion of Law

There remain, of course, many problems and open questions with Sander’s account. Felix
Kaufmann, for instance, asked where precisely the “criterion of law” in Sander’s theory
should be found: “For, through the ‘immanent analysis of the methods of creation’ of the
law we will never reach a knowledge about what the law is; it is rather this knowledge of
20
the law which allows us to identify certain methods of creation as legal.”

Kaufmann seems to have a valid point here. Do we not first need a concept of law in order
to speak about the law? Does not Sander’s entire enterprise rest on an implicitly
presupposed concept of law?

Well, that might indeed be the case. However, Sander can grant it without running into
trouble. The question is, after all, whether the concept of law finds its determination in
philosophy, in legal science, in pre‐scientific everyday language, or recursively in the law
itself.

What Kaufmann does not sufficiently take into account is that a concept of law, or a
“criterion of law,” is inscribed into the law itself: the legal process decides in a constitutive
and not mere declaratory way on the legal quality of putatively legal entities.

Kaufmann would very likely have claimed that such an intrinsic concept of law may very
well exist, but that what he was demanding was more than that: a comprehensive concept
of law. In claiming the legal nature of this intrinsic concept of law, Kaufmann might say
that we are not denying, but rather presupposing, the existence of this higher,
comprehensive concept of law.

Let us briefly follow through this distinction between such an intrinsic and a
comprehensive concept of law and see if the latter actually can do any work: a
comprehensive concept of law, if it is to be truly a content concept of law and not merely
an empty logical limiting concept, has to have a determinate content. Now, if it has such a
content, then the positive law can always make reference to it, and it can explicitly make it
legally relevant, thus turning this comprehensive concept into an intrinsic concept of law.
There is nothing in the comprehensive concept that can prevent it from becoming an
intrinsic concept. Whatever conceptual determinations we come up with for the
20
Felix Kaufmann, Theorie der Rechtserfahrung oder reine Rechtslehre? Eine Entgegnung, 3 ZEITSCHRIFT FÜR
ÖFFENTLICHES RECHT 244 (1922).
comprehensive concepts, the law itself will have to decide whether they are legally
relevant (i.e., whether the concept is a true concept). The law is competent to decide on
any concept of law, and thus turns the comprehensive into an intrinsic concept.

So we have to introduce a further distinction, namely, the distinction between a
provisionally comprehensive and an ultimately comprehensive concept. The provisionally
comprehensive concept of the law is not truly comprehensible because it will become an
intrinsic concept of the law, if it is a concept of the law at all. The ultimately
comprehensive concept of the law, however, cannot have any content; it is only a limiting
concept, a pure marker that highlights the impossibility of coming up with a
comprehensive concept of law, which could not be turned into an intrinsic concept.

The ultimately comprehensive concept of the law is, of course, what Kelsen will call the
21
basic norm. It is the signet of the impossibility of legal philosophy to come up with an
ultimately comprehensive concept of the law, and this impossibility takes the form of a
legal norm. The basic norm is thus what stands between jurisprudence and the law; it is
the legal incarnation of the independence of law from jurisprudence.

The driving idea behind Kaufmann’s criticism was the vague conviction that we need
something “outside” the law to determine the law. This outside, however, is either legally
relevant, and thus awaiting confirmation in and being dependent on the legal process; or it
is legally irrelevant, and thus incapable of doing the work it claims to be doing. The
question that remains is whether this “outside,” which drives Kaufmann’s criticism,
conceptually makes sense at all.

It is one of the fundamental claims of positivism, properly understood, that it does not.
For the law, there is no outside of the law, and for the outside of the law, there is no law.
It is in this sense, I think, that Sander and Kelsen have to be understood as being positivists.
Much more than standing in a tradition of positivism, which muses about the possible
relations of law and morality (a tradition that has, understandably, only taken off after the
Second World War), Kelsen stands in the other, older tradition of legal positivism. This
tradition of positivism goes back at least to Savigny: “In relation to this quality of the law,
according to which it already has actual existence as given in every situation, we call it
22
positive law.”

21
A pure theory, in its proper development, must thus be able to develop all of its pronouncements from an
analysis of the logic of the basic norm alone. The rule of recognition, conversely, cannot feature as an ultimately
comprehensive concept of law, because it actually already presupposes a range of specific positivist content (e.g.,
the difference between officials and citizens).
22
FRIEDRICH CARL VON SAVIGNY, 1 SYSTEM DES HEUTIGEN RÖMISCHEN RECHTS 14 (1840).
2011]
Accordingly, the decisive insight of positivism is not that the law is conceptually separable
from morality, but that the law has already been there. The positivity of the law is a
function of its temporality: the law was there before we were there. Because of this,
every “outside of the law” has to remain an empty abstraction.

There is, at least for us, no “outside” of the law; there is no ahistorical vantage point from
which the question of the criterion of the law could intelligibly arise. We can, of course,
always consistently deny the existence of the law in toto, and nobody could prove to us the
contrary. However, we have then not reached the outside of the law, but only an absolute
outside without any law. The core claim of positivism thus is, if there is to be positive law
at all, then there cannot be an outside of the law from which attempts at an analytical
determination of the concept of law could make sense. Rather—and this is Kelsen’s
solution—the concept of law has to be a norm. It can only be derived by reflecting on the
totality of the law and the impossibility to come up with a criterion of law. This
23
impossibility is the basic norm.

E. Kelsen’s Criticism of Sander’s Position

But how did Kelsen himself react to Sander’s turn? Kelsen’s reaction to Sander’s radical
proposition went through three stages: confusion, rejection, and incorporation.

At first, Kelsen was confused by Sander’s work. In his reply to Sander, Kelsen oddly attacks
24
both Sander’s early position and his later reversal. This is especially surprising as Kelsen
must have known that, in his attack of Sander’s early position, he is charging an open door.
Can one criticize someone who has explicitly revised—nay, reversed—his original position
for this original position in a piece of work which aims at refuting the later position?

In terms of substance, however, Kelsen at this point of the development of his theoretical
position tried to reject Sander’s analogisation of law with natural science on the basis of
two claims: (1) that Sander’s thesis is counterintuitive, and (2) that legal science does
indeed constitute the law.

I. Sander’s Thesis is Counterintuitive

Kelsen claims that Sander falls prey to an equivocation of the terms “process,” “judgment,”
and “law,” which, as homograms, connect the sphere of law with the sphere of
constitutive cognition, but have no relation in terms of content.

23
See infra text accompanying note 31.
24
See HANS KELSEN, RECHTSWISSENSCHAFT UND RECHT: ERLEDIGUNG EINES VERSUCHES ZUR ÜBERWINDUNG DER RECHTSDOGMATIK
3 (1922).
But Kelsen misses the point here. What he takes to be “two a priori quite heterogeneous
things” are, given a closer inspection, less heterogeneous than common sense takes them
to be: what determines the sphere of cognition as such is, as already mentioned, not some
kind of cognitive substratum, but only a specific form of connection or synthesis.

Or, to put it differently, the law has to be seen to have a cognitive function because the
truth about the law is produced in the legal process. Any pronouncement legal science or
25
legal philosophy makes about the law can be falsified by the law itself.

The problem lies in the uncomfortable alternative that legal science faces: it can relate to
the law either as a mere fact or as a legal fact. If it relates to the law as a mere fact, then it
does not relate to the law as law; if, however, it relates to the law as law, it has to relate to
the law as a legal fact. However, whether something is or is not a legal fact depends
entirely on the law.

It is surprising that Kelsen, who throughout his career defended the nature of things
against common sense, here himself relies on common sense against the nature of things.

II. Legal Science Constitutes Law

Secondly, Kelsen clarifies his position vis‐à‐vis Sander as follows:

The synthetic judgements of natural science are, even
though they are “produced” by natural science,
nevertheless just as determined by the “material”
which is to be unified in them (this is why they are
judgements “about” nature), as the synthetic
judgements of legal science, the Rechtssätze, are. In
the latter the material given to legal science (the
statutes, ordinances, court judgements, administrative
acts, etc.) is transformed into Rechtssätze, just as the
material of sensation is transformed in the synthetic
judgements of the natural sciences; and these synthetic
judgements of legal science, the Rechtssätze, are just as
25
The importance of Lon Fuller’s still too readily neglected contribution to jurisprudence should become apparent
here. Fuller correctly saw that an anti‐positivist project could only hope to be successful as a skeptical
reconstruction (i.e., if it started from the insight that the content of any legal philosophy has to be limited to that
which the law cannot consistently claim to be, to claims about the law which cannot be falsified by the law itself).
This focus on formal consistency, however, is troubled by the fact that the law, so to speak, creates itself anew at
every moment and does not even rely on formal consistency. Very often it retrospectively declares
inconsistencies to be consistent and vice versa, faults can heal and be rendered retrospectively legal and valid
pronouncements can be rescinded.
2011]
determined by the material given to them as the

judgements of natural science are. In the Rechtssätze,
which are judgements, and as such are a function of
cognition, of a scientific consciousness, only this
material may be processed which is given and assigned
to them, just as in the judgements of natural science
only this material may be processed which is given by
sensation. This, and nothing else, is what is meant by
demanding that legal science should not be a legal
source. However, legal science still may, well, it must,
“produce” the Rechtssätze as judgements. And thus
there can only be judgements of legal science and there
cannot be judgements of the law; just as it would be
absurd to declare the synthetic judgements, in which
nature is made an object of cognition, in which it is
made the nature of natural science, to be judgements
26
of nature itself.

Disregarding the fact that Kelsen himself here adopts a position which tries to find a strict
analogy between natural and legal science, a position which he heavily criticized in Sander
in the same article, this argument errs in many respects. Firstly, he confuses Kant’s
transcendental philosophy with a Humean sense‐date theory. He has turned Kant into a
“picture theorist” of cognition. He seems to be reading Kant in the following way: science
has to “produce” judgments which correspond to the “true relations” present in the world,
or to the objective order of the world. This, of course, is not Kant but Hume, since, for
Kant, the “true relations” are not given but produced themselves. Or, to put it more
carefully, if we want to make sense of our knowledge of the world, we cannot conceive of
the world as in itself ordered; rather, we have to go through the Copernican revolution of
understanding and accept the “discursive” nature of knowledge (i.e., the fact that we can
only know nature by applying concepts). It is, after all, the point of Kantian philosophy—a
point which Kelsen at this stage of his development does not yet seem to comprehend—
that the judgments of natural science are not determined by the given material, but that,
conversely, the capacity of judgment determines the material to nature. Here, it shows
most clearly that Kelsen at this stage was neither Kantian nor Neo‐Kantian, but rather a
Humean empiricist who only embellished his empiricist positions with Kantian vocabulary.

Secondly, Kelsen seems to overlook his own demur against Sander’s earlier position,
namely, that the law differs in one fundamental respect from nature as an object: whereas
no connection can be given in pure receptive sensation, such a connection can very well be
given in the law. Whereas the logicity, which is an indispensable element in the cognition
26
KELSEN, supra note 25, at 181–82.
of nature, can never be assumed to be simply given in nature—such an assumption would
inevitably turn our project into uncritical metaphysics—and thus has to be thought of as
being spontaneously produced by understanding, the law can very well be conceived as
being in itself ordered. After all, the entire transcendental effort is no vain play with
concepts, but has turned out as the only consistent option to conceptualize and represent
to man human knowledge of nature (i.e., a kind of knowledge that has to start from the
assumption that we simply do not have any direct access to the “how” and the “if” of the
natural order). All the transcendental turns and wrenches become necessary only because
we want to know how we can know something about an object, about which we should
not be able to know anything: such knowledge is possible only if it is we who produce the
object of cognition.

In relation to the positive law, however, the problem vanishes. In contrast to nature the
law is in itself ordered, and can find its unity only in itself.

Kelsen obfuscates this difference between nature and law in claiming that legal science has
to picture, reproduce, or map the order of the law. If this is the case, legal science is not
constitutive of order and thus is not science in the Kantian sense of the term. If, on the
other hand, it would be claimed that legal science is constitutive of the order of the law,
then it might be a science in the Kantian sense; however, it would cease to be positive legal
science and become a theory of natural law. Legal science thus faces the dilemma of
either not being science in the Kantian sense or being natural law.

The debate between Kelsen and Sander has thus come to the following unspectacular
conclusion as it relates to legal science: legal science faces the alternative of either not
being a science in the Kantian sense, or of being a theory of natural law. Surprisingly,
however, this is a result with which both Sander and Kelsen can live. Both agree that
natural law has to be avoided and both agree that legal science is not a science in the
Kantian sense. The difference lies in the accentuation. Whereas Sander would formulate
the result as, “Since legal science is not a science in the Kantian sense, it is not science at
all,” Kelsen would express it as, “Legal science may not be a science in the strict Kantian
sense of the word—this, however, does not matter, as it does not make that claim.”

F. Kelsen’s Partial Adoption of Sander’s Position

However, the important outcome of the Kelsen‐Sander controversy is not a
reinterpretation of legal science, but of law. The law, as it turned out, is “constitutive.” To
put it differently, that, of which Kant claims that it happens in knowledge, actually happens
in the law. This result is relevant even for weaker than Kantian epistemologies (i.e., even if
one does not agree with Kant about the constitutive nature of knowledge, one can still
agree with Sander that that what Kant claimed happens in knowledge actually happens in
the law).

2011]
What is more, despite his attempts to make Sander sound ridiculous, it can be argued that
Kelsen later on actually incorporated many ideas which have a distinctive Sanderian ring to
them. To what extent Kelsen has been directly influenced by Sander, and whether he
would have acknowledged any kind of indebtedness, must, of course, remain unknown to
us. However, what is beyond doubt is that Kelsen’s development of his ideas on the
dynamic nature of the legal order and his theory of international law have clearly benefited
from his skirmish with Sander, as he reached positions more close to Sander’s than to his
own previously held views.

What is even more interesting for us, however, is Kelsen’s matured position on the
“cognitive” nature of the law.

Kelsen agrees with Sander that it is the law itself which is constitutive of legal facts.

But how, according to Kelsen, does the law constitute these legal facts? It does so by
27
schematizing interpretation (i.e., by allowing it to understand something as something).
It is only by means of this function as a scheme of interpretation, that the law can
constitute legal facts.

That some law schematizes interpretation does not seem to be a very controversial claim.
Take the following example: suppose that you see a room of people, with some standing
up while others remain seated. From those facts alone, Kelsen claims, nothing objective
can be concluded. It is only the norms of the constitution that allow us to interpret the
assembly, say, as a Parliament, and the goings‐on as the enactment of a statute:

External circumstances are always a part of nature, for
they are events perceptible to the senses, taking place
in time and space; and, as a part of nature, they are
governed by causal laws. As elements of the system of
nature, these events as such are not objects of
specifically legal cognition, and thus are not legal in
character at all. What makes such an event a legal (or
illegal) act is not its facticity, not its being natural, that
is, governed by causal laws and included in the system
27
Together with (1) the doctrine of authorisation; (2) the doctrine of the hierarchical structure of the legal order;
(3) the doctrine of the dual character of the legal act; (4) the doctrine of the complete legal norm; (5) the doctrine
of the Fehlerkalkül; (6) the doctrine of the basic norm, and (7) the doctrine of the scheme of interpretation make
up the seven doctrines of the Pure Theory of Law. Lippold has eight (quite similar) “elements” of the Pure Theory
of Law. See RAINER LIPPOLD, RECHT UND ORDNUNG: STATIK UND DYNAMIK DER RECHTSORDNUNG 526 (2000); Christoph
Kletzer, Das Goldene Zeitalter Der Sicherheit: Hersch Lauterpacht Und Der Modernismus, in HANS KELSEN UND DAS
VÖLKERRECHT: ERGEBNISSE EINES INTERNATIONALEN SYMPOSIUMS IN WIEN, 1–2 APRIL 2004 (Klaus Zeleny, Robert Walter &
Clemens Jabloner eds., 2005).
of nature. Rather, what makes such an event a legal

act is its meaning, the objective sense that attaches to
the act. The specific legal sense of the event in
question, its own peculiarly legal meaning, comes by
way of a norm whose content refers to the event and
confers legal meaning on it; the act can be interpreted,
then, according to this norm. The norm functions as a
scheme of interpretation . . . . That a material fact is
not murder but the carrying‐out of a death penalty is a
quality, imperceptible to the senses, that first emerges
by way of an act of intellect, namely, confrontation
with the criminal code and with criminal procedure.
The aforementioned exchange of letters [between
merchants] means that a contract has been concluded,
and it has this meaning solely because these
circumstances fall under certain provisions of the civil
28
code.

The law thus allows a certain interpretation of the world; it allows treating certain facts as
having a certain meaning.

c b
Table 3: Scheme of Interpretation

Norm a allows us to treat the fact b as having the meaning c. 29

But what is the relation of this legal scheme of interpretation to other schemes of
interpretation?

Well, even if the people in the aforementioned assembly were to refer to themselves as
“Parliament” and to their action as “enactment of statute,” we would still need a legal rule
28
HANS KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY: A TRANSLATION OF THE FIRST EDITION OF THE REINE
RECHTSLEHRE OR PURE THEORY OF LAW 9 (Stanley L. Paulson ed., 1992).
29
Note the similarity of Kelsen’s notion of the scheme of interpretation with Searle’s concept of constitutive rules
and institutional facts which work according to the same scheme: “X counts as Y in C.” The advantage of Kelsen’s
account, apart from it being first in history, is that Kelsen does not need to rely on the psychological notion of
“acceptance,” as Searle does, and which makes his conception either circular or metaphysical. See JOHN R. SEARLE,
THE CONSTRUCTION OF SOCIAL REALITY (1995).
2011]
to confirm this self‐interpretation and exclude the possibility that the body calling itself
“Parliament” here is actually only the assembly of an unsuccessful revolutionary party or
the “Parliament” in a big theatre production. And even if a sociologist were to tell us that,
according to the criteria of his sociological theory, the assembly “really is” a Parliament,
there is still the possibility that the legal criteria and the sociological criteria differ. It is the
law itself which stands in competition with sociology. Kelsen’s point is that the law itself
has a sociology, just as it has a morality. It is only because the law itself is both sociological
and moral in nature that the study of the law has to steer clear from moral and sociological
considerations. If it did not, it would replace its object’s sociology and morality with its
own and it would thus miss its object. The purity of the pure theory is thus purely an
epistemological demand.

Now, because this interpretation of the going‐ons in accordance with the constitution can
falsify the immanent and even the “scientific” interpretation of what is happening, Kelsen
calls it the “objective” interpretation and he calls the immanent and scientific
interpretation “subjective.”

We might at first be surprised by this use of the subjective/objective dichotomy. Is not the
immanent and scientific interpretation just as objective as the legal interpretation? What
is it about an interpretation in accordance with the constitution that should make it
objective? Well, the difference between an interpretation in accordance with sociological
and one in accordance with legal rules is that the legal rules are “effective.” They are
schemata which are actually used. Scientific rules might be true, however, whether or not
someone actually uses them to identify certain objects; they are not part of their validity as
rules. For legal rules, in contrast, it is one of the conditions of their validity that they are
part of a system which is by and large effective. Interpretation in accordance with legal
rule is thus “objective” in the sense that the rules which function as schemes for this
interpretation are themselves “out there.”

With this notion of “objectivity,” Kelsen has, of course, already overstepped the confines of
Kantian transcendental philosophy towards the Hegelian notion of objective spirit: with
“objective,” Kelsen does not refer to that which is constructed by a subject in accordance
to certain categories and schemes which somehow exist in the subject; but, rather, he
refers to a view of the world in accordance to rules and schemata which themselves are
objectively “out there,” which are effective in the world.

What might trouble some readers immediately is the following: does not the constitution
in order to supply the goings‐ons with what Kelsen calls the objective meaning itself have
to be valid? And does not the validity of the constitution presuppose another norm which
can function as a scheme of interpretation in relation of the act of the creation of the
constitution? And does not all of this lead us into classical Agrippa’s trilemma of being
either faced with an infinite regress, a circular argument or a dogmatic denial of the need
of derivation?

The worry is premature. We should not jump ahead but focus on the problem at hand:
the problem we have before us is not the absolute validity of a single legal rule or the
legality of the entire legal system, but only the relation of the goings‐ons at hand to a rule
which can feature as a scheme of interpretation. What we are interested in at this
moment is only the relation of a valid rule to actual facts and not the relation of valid rules
to other such rules. In order to examine this relation, there is no problem in simply
presupposing the validity of the rule for our purposes, and to postpone an intellectual
30
resolution of this presupposition to a later doctrine.

So the issue at hand is the question as to the objective interpretation of a factual
occurrence. In order to arrive at such an interpretation, we need a valid rule. If we
presuppose its validity, the constitutional provision does indeed function as a scheme
which allows an objective interpretation in our context.

Now, the interesting thing about Kelsen’s treatment of the scheme of interpretation is that
he views it not as a secondary function of the law, but as a central feature of the positive
law. According to Kelsen, the scheme of interpretation is both a universal and an
exhaustive doctrine: all law functions as a scheme of interpretation, and all that law can
immediately do is to function as a scheme of interpretation.

The law thus provides for a specific interpretation of facts, mostly of political facts. The
law here does not tell us what to do, but it is first and foremost an instrument of
reinterpretation of social facts and thus also an instrument of social self‐reflection. It helps
us understand the social world. The law itself, of course, does not bring about this
interpretation; rather, it “offers” or “allows” us to interpret the world according to the
scheme.

30
This doctrine is, of course, the doctrine of the basic norm, the most misunderstood doctrine of Kelsen’s work.
The latter is not simply an odd addition made at the top of an existing legal system, but the explicit philosophic
formalisation of the presuppositions of validity already made at every stage of legal cognition. It is not that we
could first start with, say, a local statute and then work our way up via the federal stature and the constitution to
the basic norm. Viewed in such a way, residing at the top of the legal system, the basic norm must of course
appear as arbitrary and absurd. Rather, Kelsen’s claim is that without presupposing validity, we cannot even start
with the local statute, we could not start anywhere. There would be no question to answer. The basic norm is
simply the legal formulation of this presupposition of validity. So it is not the case that without the basic norm
there would be a legal system which would somehow lack validity, but there would be no legal material in the
first place! Just as it is said that life is understood backwards, but lived forwards, the law is understood backwards
but formalised forwards. It is the complex notion of a “presupposition” (Voraussetzung) which does most of the
work here and which is not studied enough in jurisprudence. This is particularly unfortunate as there is an
intimate relation between the “positivity” of the law and the “presuppositivity” of its validity.
2011]
G. Why Purity?

Now, it is only because of this “cognitive” function of the positive law that a theory of law
has to be “pure.” The Gegenstandsproblem is, accordingly, the key to the purity of the
Pure Theory. The purity of the Pure Theory thus does neither stem in some kind of
intellectual austerity 31 nor in Kelsen’s political philosophy, 32 but entirely in epistemological
considerations. Kelsen’s positivism is neither analytic nor political; it is philosophical.

In order to further substantiate this claim, let us thus have a brief look at law’s relation to
both sociology and morality.

I. Law and Sociology

The debate about a sociological jurisprudence is based on the classical model of the
relation of law, legal science, and legal facts. The possibility of, and demand for, a
sociological jurisprudence rests on the idea that both doctrinal legal science and sociology
relate to the positive law as their respective material or object.

Level 1 Sociology Legal Science
Ground Level Law
Table 4: Classical Model of Sociological Jurisprudence

Given this setup, it must seem arbitrary to demand legal science to remain free from
sociological considerations.

Sander’s insight, however, changes the picture fundamentally, and Kelsen’s life long
struggle against sociological jurisprudence shows how much he took up Sander’s ideas.
According to Sander, it is the law itself and not legal science which stands in competition
with sociology about the determination of legal facts as legal facts. It is only the law, in
operating as a scheme of interpretation—to use Kelsenian terms again—which allows us to
interpret certain facts as being legal facts.

31
See Joseph Raz, The Purity of the Pure Theory, in NORMATIVITY AND NORMS: CRITICAL PERSPECTIVES ON KELSENIAN
THEMES 238 (Bonnie Litschewski Paulson & Stanley L. Paulson eds., 1998) (proposing this, but leaving out all
questions relating to the possible reasons for this purity).
32
See, e.g., LARS VINX, HANS KELSEN’S PURE THEORY OF LAW: LEGALITY AND LEGITIMACY (2007).
Level 1 Sociology Law

constitutive constitutive
Ground Level Legal Fact
Table 5: Sander’s Model

In this setup, any legal sociology has to be pointless. Facts by themselves are just facts.
There is nothing in their quality as facts that could mark them out as legal. It is only the
law that can mark out certain facts as legal facts. In order to be legal sociology, in order to
relate to the objective sphere “legal facts,” sociology would have to accept the criteria
given by law. This, however, would make it unscientific, because science has to produce
and test its own criteria of object‐relation. “Legal sociology” and “sociological
jurisprudence” are thus oxymora: if they are legal, they are not scientific, and if they are
scientific, they are not legal.

Accordingly, Kelsen throughout his career was an avid critic of the possibility of legal
sociology. His rejection of legal sociology stems not in any dissatisfaction with the
discipline as such, but in principled epistemological considerations: sociology has to
determine legal facts, but the law too determines those legal facts. Now, in case there is
disagreement between legal and sociological determinations, sociology faces the dilemma
of either becoming unscientific or missing its object: it can either give up its own criteria of
determining legal facts and submit to the law’s determinations, at the cost of becoming
unscientific; or it can stick to its own determinations, but then it is unclear in what sense it
33
still is legal sociology.

II. Law and Morality

The dynamic understanding of the law, which treats the law itself as being the realm of
constitutive relation of facts and norms, also has important repercussions on our
conceptualization of the relation of law and morality. Some of these might already have
become obvious in the previous debate: if the law is a constitutive relation of fact and
norm, this not only makes difficult a separate factual (i.e., sociological) approach but also a
separate normative (i.e., moral) approach.

For Kelsen, the problem with respect to the relation of law and morality thus is not that
morality is superfluous in legal matters; or that it is not strictly speaking part of the
33
This is the reason most Kelsen scholars try to steer clear of Luhmann, whose ideas, despite their closeness to
Kelsenian themes, nevertheless remain sociological.
2011]
concept of the law, that we could conceivably talk about immoral law; but rather, the
problem is that law and morality both want to do the same thing.

Kelsen stressed that the positive law is “ethical‐political speculation,” 34 that it is an
ideology, 35 and that the law itself is a relative morality. 36 The law shares its content with
ethical institutions like positive morality, churches, the family, and so on. 37 What sets it
apart from those institutions is not the content but the form: “The question about the
relationship between law and morals is not a question about the content of the law, but
one about its form.” 38

This is of crucial importance: it is not the content, which distinguishes the law from
morality, but only its form. But what is it about the form that differentiates the law from
morality? The difference in form is that law is a system of schemes of interpretation which
is structured procedurally, whereas morality lacks this procedural quality. Procedurally
here means that it is for the law to determine according to its own schemes what is to
count as what and, most importantly, what is to count as legally relevant, as a “legal fact.”
The difference in form is thus the constitutive nature of the positive law itself, which, as
39
Sander has shown, the law shares with the natural sciences.

The legal moralist is deluded by the fact that the law shares the content with his enterprise
and falsely concludes from this that the law can learn something from morality. He misses
the crucial point that the law is of different form.

“Since morality always evaluates actuality, a moral evaluation of the positive law directly
only relates to the acts which create norms, and it relates to the norms themselves only
34
HANS KELSEN, DER SOZIOLOGISCHE UND DER JURISTISCHE STAATSBEGRIFF: KRITISCHE UNTERSUCHUNG DES VERHÄLTNISSES VON
STAAT UND RECHT 46 (1928).
35
See HANS KELSEN, THE PURE THEORY OF LAW 104–05 (2d ed. 2000).
36
See id. at 65.
37
This is, of course, a very Hegelian claim: “It is philosophical insight which recognises that Church and state are
not opposed to each other as far as their content is concerned, which is truth and rationality, but merely differ in
form . . . . In contrast with the faith and authority of the Church in relation to ethics, rights, laws, and institutions,
and with its subjective conviction, the state possesses knowledge. Within its principle, the content in no longer
essentially confined to the form of feeling and faith, but belongs to determinate thought.” G.W.F. HEGEL,
ELEMENTS OF THE PHILOSOPHY OF RIGHT 299, §270 Addition (H.B. Nisbet trans., 1991).
38
KELSEN, supra note 36, at 65.
39
Again, consider Hegel: “Science . . . has the same element of form as the state,” and “The state knows what it
wills, and knows it in its universality as something thought.” HEGEL, supra note 38, at 300, §270 Remark, 290,
§270 Addition.
indirectly.” 40 This means the following: whenever we think we are morally evaluating “the
law,” we are actually primarily evaluating a specific act of creation of law. When we say,
“this statute is morally reprehensible,” what we are primarily saying is something like,
“Parliament should not have enacted this statute.” Now, the problem with this is that the
act in question can only be law if another law evaluates it as being lawful.

Kelsen’s argument is thus that the moralist misunderstands the law and his own possible
relation to it: the moralist may think that he relates to the law as the object of his moral
assessment, whereas in reality he competes with the law in his assessment of legally
relevant facts. He interprets them as “immoral,” the law interprets them as “legal.”

The problem thus is not simply that law and morality are conceptually unrelated, but that
they are related in a way which reveals the strict impossibility of a moral treatment of “the
law.” As “the law” itself is not a thing, but a relation of norms and facts since it itself is an
“assessment” of facts, it is no suitable object of moral assessment.

So, again, Kelsen’s argument that an engagement with the law has to remain free from
moral considerations follows an entirely different path than Hart’s. Whereas Hart’s
empirical approach leads him to claim that morality is not part of the concept of law and
thus has to be kept separately, Kelsen’s engagement with the Gegenstandsproblem, or
with the possibility of our relation to the law, leads him to the claim that because law is
fundamentally moral in nature.

This should also help explain both the reasons Kelsen has had for the following claim and
the difficulty Hart has had with it: “From the point of view of positive law as a system of
41
valid norms, morality does not exist as such.”

Confronting the positive law with moral claims (“opinion”) is like confronting natural
science with private perceptions of natural phenomena. Thus, if I, in all sincerity, tell a
physicist that, last week, I have observed a second sun on the evening sky, and on his
probing simply insist that I am really certain to have observed it, he will react with infinite
indifference.

The point here is not that it would be uninteresting for physics if there indeed were a
second sun or the appearance of a second sun on the evening sky. The problem is not the
40
HANS KELSEN, REINE RECHTSLEHRE 69 (2d ed. 1967). The quote is taken from a footnote in the German original
which has been omitted in the English translation.
41
HANS KELSEN, GENERAL THEORY OF LAW AND STATE 374 (1999); see also H.L.A. Hart, Kelsen Visited, in NORMATIVITY AND
NORMS: CRITICAL PERSPECTIVES ON KELSENIAN THEMES 83 (Bonnie Litschewski Paulson & Stanley L. Paulson eds., 1998);
cf. HEGEL, supra note 38, at 294, 301, §270 Remark (“All that need be mentioned here is that the attitude of the
state towards opinion—in so far as it is merely opinion, a subjective content which therefore has no true inner
force and power, however grandiose its claims—is one of infinite indifference.”).
2011]
content of my claim, but its form. The problem is that, in my pronouncement, I claim that I
observed something of relevance to physics and at the same deny the applicability of the
criteria physics has established of something being relevant to it (i.e., confirmability, etc.).
In my claim, I exhibit that I do not subject myself to the logic of physics—of universality—
but ground the strength of my claim in its privacy and, particularity, something to which
physics has to remain indifferent.

Physics cannot deal with such claims, and this is not because they would not be interesting;
rather, it is because, for physics, the content of such claims carries within itself a logic of
confirmation or refutation (a logic along the lines of, “if what I observed really was a sun,
then it will have to be there tomorrow again so let us look tomorrow if there is still the
appearance of another sun, let us start an inquiry if others have had the same apparition,
etc.”), a logic which the claim in its form refuses to follow. The elaboration of this and
actualization of this claim of course already is physics.

So in my claiming that I have observed a second sun and at the same time only relying on
past perception and not subjecting the claim to the methods of confirmation provided by
physics, I exhibit that I have actually misunderstood the meaning of my own perception.

It is in this sense that claims of physics and the claims of private observation do not touch
each other and thus cannot contradict one another, but this is not because claims of purely
private observation were alien to physics, but because they are, so to speak, “incomplete
physics”; they are the “trigger” of physics, they can initiate a process of scientific inquiry
but—just as the incomplete cannot refute the complete—they cannot in themselves refute
valid claims of physics.

In a similar sense, morality is “incomplete law.” The positive law has to react to
expressions of moral opinion with “infinite indifference,” and it has to do so not because
the positive law was somewhat amoral, or in a wrong sense “pure,” or because it was not
interested in morality; but because the positive law is already the actual elaboration of
what is only subjectively claimed in private moral utterances. Morality may be the trigger,
the occasion of a legal process, but it cannot by itself contradict the law. If I approach the
law with, say, the claim that property is intrinsically immoral, and if I insist that it is really
the case that private property is intrinsically immoral; then the law has to remain
indifferent and it has to remain so, not because the content of my claim was somewhat
irrelevant to it, but because the form of my claim in its subjectivity is insufficient. The
content of this claim carries within itself a logic of realization and decision (along the lines
of, “if there is no property, then we have to decide differently about the allocation of the
use and transfer of resources; maybe a party‐apparatus should decide on those allocations
according to the following process, and so on.”). The point is that the positive law is the
actuality of this logic. Private morality is thus incomplete positive law, or yet to be
completed positive law. The positive law and the claims of private morality do not touch
each other and thus cannot get into conflict, not because moral claims were alien to
morality, but because they are incomplete law.

H. Conclusion

In this paper, I have tried to show that an analysis of the Gegenstandsproblem of legal
science is not an awkward pet passion of some dead Continental legal philosophers, but
that it forms the prolegomenon of any philosophic treatment of the positive law. Without
awareness of the very specific problems that the law as an object in general poses, we
cannot hope to make any progress in the deepening of our understanding of the positive
law, and are doomed to run the circles sketched out by the naivety of our approach.

The reflective approach proposed is of course idealistic (in the sense of German
“Idealism”), at least in style. It tries to solve problems of ontology by transforming them
into problems of epistemology: It tries to answer questions like “What is the law?” by
translating them into “What can the law be for us?”; and questions like “What is the
concept of law?” into “What concept of law can we possibly have?”. I hope to have made
at least a preliminary case for couching this move in the language of absolute idealism
rather than in the language of transcendental idealism.

I do think that all prominent problems and enigmas in contemporary debate of the nature
of the law would benefit significantly from infusing the debate with a heightened
awareness of the Gegenstandsproblem; and it is only due to the constraints that come with
a journal article that further discussion of how an analysis of the Gegenstandsproblem can
help us deepen our understanding of the problems pertaining to, say, the normativity and
the authority of law has to be omitted.
Special Issue
Comment on Kletzer—Positive Law and the “Cognitivity Thesis”

By Luís Duarte d’Almeida ∗

A.

Christoph Kletzer’s paper is by and large an attempt (a failed one, I will contend) to argue
for what he calls the “cognitivity thesis” 1 —the succinctly rendered claim that

(CT) The positive law is cognitive.

The idea behind this opaque (and admittedly “puzzling” 2 ) proposition is traced back to the
work of Fritz Sander, and ascribed to Hans Kelsen as part of his “matured” views. 3 It is also
clearly endorsed by Kletzer himself. In this brief commentary, I shall concentrate directly
on (CT), and leave aside, for the most part, issues of Sanderian or Kelsenian hermeneutics. 4
I will maintain that the arguments adduced by Kletzer in support of (CT) are flawed; the
“cognitivity thesis” is every bit as wrong as its “highly counterintuitive” 5 make‐up suggests
it to be.

B.

First, however, some scene setting is in order. Let us begin by noting that (CT) is one of a
group of three “new and momentous” 6 theses of Sander’s which—amounting in Kletzer’s
judgment to nothing short of a “revolution” 7 —are said to overlie the “insight” that “legal
∗
University College, Oxford; and LanCog Group and Law Faculty, University of Lisbon. Email:
luis.duarte‐dalmeida@univ.ox.ac.uk; luisduartealmeida@gmail.com
1
See Christoph Kletzer, Kelsen, Sander, and the Gegenstandsproblem of Legal Science, 12 GERMAN L. J. 794 (2011).
2
See id.
3
See id. at 803.
4
The suggestion will be nevertheless be made in passing that Kletzer’s case for holding Kelsen hostage to the
“cognitivity thesis” is unconvincing.
5
Kletzer, supra note 1, at 794.
6
Id.
7
Id.
science has at all times been under the misapprehension that the positive law stands in an
analogous relation to it as nature does to natural science.” 8

The proposition here rejected—that “legal science” 9 is to positive law as natural science to
nature—is one which Sander is said to have endorsed in an earlier piece. 10 We should
nonetheless keep in mind that both defense and dismissal of such an analogy share a
philosophical background: Neo‐Kantian epistemology. 11 The relevant assumptions, in
Kletzer’s summary, “can be expressed as a set or more or less diffuse transcendental
convictions: (α) in knowing an object we in a way construct it; (β) the human sciences have
an analogous structure to the natural sciences; (γ) legal science is a human science.” 12

We are reminded, in particular, that under this philosophical mindset “if we want to make
sense of our knowledge of the world we cannot conceive of the world as itself ordered.” 13
“Nature” is thus “not simply given to natural sciences as a unified and comprehensively
determined whole.” 14 Rather, “the unity of nature and thus the ‘natureness’ of nature has
to be constructed” 15 —a “construction” “effected in the natural sciences.” 16 “Cognition in
the Kantian architectonic” 17 is, in the stock idiom, “constitutive.” 18

“In order to follow the debate,” nevertheless, we do not need “to accept Kantian or
Neo‐Kantian epistemology;” “all we need to accept is that knowledge is not a purely
passive phenomenon (i.e. that there is at least a minimal sense in which in knowing an
object we are not only determined by the object but also determine it).” 19
8
Id. at 793.
9
See Kletzer, supra note 1, at 792, n. 14 (defining “legal science” as “doctrinal legal scholarship, or legal research
(i.e. what we do in law schools and get money for from research bodies and the government). The Germans call it
‘Rechtswissenschaft.’”.
10
See id. at 792.
11
See id. at 791.
12
Id.
13
Id. at 801.
14
Id. at 791.
15
Id.
16
Id.
17
Id. at 794.
18
Cf. id. at 802.
19
Id. at 791.
2011] Comment on Kletzer 813
These summary remarks, however generally cast, throw sufficient light on the question
under scrutiny. Recall the “three levels” which “as it concerns our knowledge of nature”
are to be differentiated: nature, natural science, and critical philosophy. 20 Sander’s—and
Kletzer’s—claims do not question, but indeed assume, that a similar structure may be
discerned in the legal domain. The first question, then, is “whether the law, legal science
and legal philosophy stand in a similar relation” 21 to, respectively, nature, natural science,
and critical philosophy.

The answer to this question seems at first glance to be affirmative. Here, for example, is
Kelsen’s take on the issue:

[A]ccording to Kant’s epistemology, the science of law
as cognition of law, like any cognition, has constitutive
character—it “creates” its object insofar as it
comprehends the object as a meaningful whole. Just as
the chaos of sensual perceptions becomes a cosmos,
that is, “nature” as a unified system, through the
cognition of natural science, so the multitude of
general and individual legal norms, created by the legal
organs, becomes a unitary system, a legal “order,”
through the science of law. But this “creation” has a
22
purely epistemological character.

Kletzer, however, takes the contrary view. He sides with Sander in denying that natural
science finds its analogue in “legal science.” Let me call this their negative claim. It is
supplemented with another, positive claim, pointing out what, according to Sander and
Kletzer, the precise legal counterpart of natural science is. This positive claim, though, is
less than accurately stated. Compare the following passages:

Not legal science but the law corresponds to the
transcendental, constitutive sphere of “cognition:” the
synthetic judgments of the law constitute the analogy
to the synthetic judgments of the mathematical natural
23
sciences.
20
Id.
21
Id. at 792.
22
HANS KELSEN, THE PURE THEORY OF LAW 72 (Max Knight trans., 2d ed.) (1970) [hereinafter KELSEN, PURE THEORY].
23
Kletzer, supra note 1, at 793, quoting FRITZ SANDER, RECHTSDOGMATIK ODER THEORIE DER RECHTSERFAHRUNG? KRITISCHE
STUDIE ZUR RECHTSLEHRE HANS KELSENS 93 (1921).
[I]t is the legal process itself, and not legal science,
which is analogous to the process of cognition in the
Kantian architectonic. 24

These two excerpts (the first of which is by Sander) are clearly seen by Kletzer as different
ways of rendering the very same points. The negative claim just mentioned is clearly set
forth in both passages. Moreover, both purport to impart the corresponding positive
claim, identifying the relevant legal correlate of “natural science.” What exactly is this
correlate, then? The reader is seemingly expected to take “the law” and “the legal
process” as equivalent expressions in this context. In that case, what do they mean?
Things are further confused by Kletzer’s use of yet another phrase for the very same
purpose—“the positive law,” as employed in the formulation of (CT).

The surest way to dissipate these confusions may be simply to circumvent Kletzer’s
terminological prodigality and directly to address his argument for the “cognitivity thesis.”
This is an argument from the character of “legal process.” What brought Sander to put
forward the “radical” cognitivity thesis, Kletzer says, was “the insight that the positive law
is neither a purely factual occurrence nor merely a complex of norms, but rather, that it is
the relation of both. The legal process, or the legally regulated application of law, is this
25
relation of fact and norm.”

Let us attempt a reconstruction of the underlying reasoning. What is the “legal process”?
Kletzer speaks here of the “legally regulated application of law.” By this, I gather, he
means to refer not solely to judicial and administrative application of general positive law,
but, more broadly, to any act of law‐creation which can simultaneously be described as an
act of law‐application. Kletzer, then, it appears, subscribes to Kelsen’s well‐known
dismissal of the traditional, absolute contrast between law‐creation and law‐application.
26
“Borderline cases” apart, says Kelsen, “it is not quite correct to distinguish between
law‐creating and law‐applying acts,” for “every legal act is at the same time the application
of a higher norm and the creation of a lower norm”: 27
24
25
Id.
26
See HANS KELSEN, GENERAL THEORY OF LAW AND STATE 133 (1945) [hereinafter KELSEN, GENERAL THEORY]; or KELSEN,
PURE THEORY, supra note 22, at 234. These “borderline cases” are, on the one hand, the “acts which are only law
application, not law creation,” i.e. “the acts . . . by which the coercive acts, authorized by the legal norms, are
executed,” and, on the other hand, the “act of positive law creation, which is not the application of a positive
legal norm: the enactment of the historically first constitution.” See Kelsen, supra note 22, at 236.
27
KELSEN, PURE THEORY, supra note 22, at 234; see also HANS KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY: A
TRANSLATION OF THE FIRST EDITION OF THE REINE RECHTSLEHRE OR PURE THEORY OF LAW 70 (Bonnie Litschewski Paulson &
Stanley L. Paulson trans., 1992) [hereinafter KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY]; KELSEN, GENERAL
THEORY, supra note 26, at 133–4.
Each law‐creating act must be a law‐applying act—it

must be the application of a legal norm that preceded
the act in order to be considered as an act of the legal
community. Therefore, law creation must be
understood as law application, even if the higher norm
determines only the personal element, the individual,
who has to render the law‐creating function. 28

It is in this very same sense that Kletzer characterizes the legal process as “the relation of
fact and norm.” The image of the “continuous connection of the legal process” is similarly
familiar: Kelsen stresses, e.g., that

[c]ontemplated from the point of view of the dynamics
of the law, the creation of individual norms by the
courts represents a transitional stage of the process
that begins with the establishment of the constitution,
continues via legislation or custom to the judicial
decision, and leads to the execution of the sanction.
This process, in which the law keeps renewing itself, as
it were, proceeds from the general (abstract) to the
individual (concrete); it is a process of increasing
29
individualization and concretization.

The law is therefore said to “regulate its own creation.” 30 The implication is that no fact
counts as a fact of law‐creation unless so determined by a “higher” norm:

[A] norm whose creation is not determined by a higher
norm cannot be considered as a norm created within
the legal order and therefore cannot be a part of it;
and an individual cannot be considered as an organ of
the legal community (his function cannot be attributed
to the community), unless he has been determined by
a norm of the legal order, that is: unless he has been
authorized by such a higher norm to perform his
31
function.

28
KELSEN, PURE THEORY, supra note 22, at 235.
29
Id. at 237; see also KELSEN, GENERAL THEORY, supra note 26, at 135.
30
KELSEN, PURE THEORY, supra note 22, at 71, 221; see also KELSEN, GENERAL THEORY, supra note 26, at 39, 124ff.
31
Kletzer concurs: “by means of its legal norms,” he says, the law “determines the entire
realm of legally relevant facts,” 32 and thus “the legal process decides in a constitutive and
not mere declaratory way on the legal quality of putatively legal entities.” 33 Some of
Kletzer’s chosen terminology is however of quite a different stripe. Consider these
extracts:

The legal process produces judgments which synthesize
given facts by employing spontaneously produced
concepts into legally relevant facts. 34

In the law . . . a synthesis takes place . . . : by means of
its legal norms, the law autonomously and
spontaneously determines the entire realm of legally
relevant facts. 35

This talk of “syntheses” and “judgments,” I hope you will agree, is highly conspicuous in the
context of Kletzer’s discussion. It is also unwarranted. There may possibly be some
colloquial, non‐laden sense of the word in which one might perhaps speak of a “synthesis”
of norms and facts when characterizing the legal process (although to employ “judgment”
for these purposes is to risk overstretching both word‐meaning and the reader’s
indulgence). But there is certainly no room to use these terms here with the specific,
“cognitive,” sense which they have in the context of Kantian or Neo‐Kantian epistemology,
to whose terminological stockpile they notoriously belong. Yet Kletzer trades on the
equivocation, and goes as far as to paraphrase Kant’s famous dictum:

It is only within the continuous connection of the legal
process (“Rechtsverfahrenszusammenhang”) that the
extralegal fact is integrated with the legal norm to a
judgment which determines both. Facts without
norms are irrelevant or, as Kant would have said,
“blind.” And legal norms without facts to which they
can be applied are ineffective or “empty.” In any case,
the relevant synthesis is affected in the law itself . . .
since it is the law which determines which fact has
36
which legal relevance.
32
33
Id. at 796.
34
35
36
Id.
Stipulation is free, to be sure. Merely calling the product of the legal process a “judgment,”
however, does not make it so. Nor does it do any argumentative work for the conclusion
that “the positive law is cognitive.” How, then, can the abovementioned “insight” that
“the legal process” is a “relation of facts and norms” be supposed to lead us toward
accepting (CT)? Kletzer appears to believe that the inference is straightforward; he asserts
that

the law has to be seen to have a cognitive function
because the truth about the law is produced in the
legal process. Any pronouncement legal science or
legal philosophy makes about the law can be falsified
37
by the law itself.

There are several reasons why this argument is wrong. For one, it sounds self‐defeating,
given that Kletzer’s “legal process” in fact presupposes the possibility of “legal cognition” in
the very sense which he purports to reject. Why? The point, which is in fact fairly trivial,
can be expressed in Kelsenian terms. Think of Kelsen’s well‐known contrast between “the
function of legal cognition and the entirely different function of legal authority.” 38 The
former’s task is “to know the law—as it were from the outside—and to describe it”; the
latter’s, “to create the law so that afterward it may be known and described by the science
of law.” 39 But, of course, Kelsen adds,

[i]t is true that the law‐applying organs also have to
know—as it were from the inside—the law they are
applying. The legislator who applies the constitution
ought to know the constitution, and the judge who
applies the law ought to know the law. 40

So if (in Klezter’s phrase) “the truth about the law is produced in the legal process,” then
knowledge or “cognition” of the applicable law is a precondition of every step in the
process, a precondition of the performance of every single norm‐producing act. Does this
mean that the legal process is “cognitive”? Certainly not. Not in Kletzer’s sense. The legal
37
Id. at 799.
38
39
Id.
40
Id.
process does not, pace Kletzer, produce judgments. 41 It produces norms. And “[l]egal
norms are not judgments, that is, they are not statements about an object of cognition.” 42

This point is connected with a second flaw in Kletzer’s argument. He says, as we saw, that
“any pronouncement legal science or legal philosophy makes about the law can be falsified
by the law itself.” 43 Taken literally, this is plainly false—at least in what concerns “legal
science.” 44 Kletzer tangles here, I think, legal “dynamics” and legal “statics.” An analysis of
the workings of the legal process is, as he stresses elsewhere, a study in legal dynamics. 45
“Knowledge” of the law, in turn—“legal science”— is relative to the law taken (as Kelsen
would put it) “as a system of valid norms.” 46 More precisely, the material given to legal
science is “the multitude of general and individual legal norms, created by the legal
organs”—the law “in its state of rest.” 47 Let me elaborate.

In the standard account, “legal science” is said to “describe” the law in truth‐apt
statements (“statements of law,” “legal statements,” “Rechtssätze”). A well‐noted feature
of such statements is particularly relevant for our discussion. This feature is often referred
to as their relativity to “legal systems.” None the less, as this latter term carries with it
some distracting theoretical connotations, I prefer to adopt a less frequent usage and to
speak instead of legal “materials” or legal “source‐materials.” The relativity of legal
statements (of tokens, not types, of statements) to sets of legal source materials may be
generally characterized, I suggest, along the following lines. First, any token statement of
law can be assessed as true or false only if contrasted with a set of legal source‐materials.
Any token statement of law, that is, may be true when contrasted with a given set of
materials, and false if contrasted with a different set. Second, the set of materials with
which any token statement of law is to be contrasted is the set of materials relative to
which it is put forward by a speaker (a jurist, a scholar—a “scientist”). The definition of the
relevant set, which has to be made discernible to the intended hearer or audience, is
entirely up to the speaker. Now, the set of source‐materials to which a legal statement is
relative may be defined either intentionally or extensionally. It may be extensionally
41
Cf. Kletzer, supra note 1, at 794.
42
43
44
“Legal science”, not “legal philosophy,” is the subject of the present discussion, and the claims of legal science
are very different from those of legal philosophy. Not surprisingly, Kletzer’s footnote to this sentence concerns
only the latter; but what one needs here instead is an argument applicable to the former. (I am taking the term
“legal science,” of course, in Kletzer’s explicitly chosen sense: cf supra note 9).
45
46
47
Id., at 70, 72; cf. also KELSEN, GENERAL THEORY, supra note 26, at 39.
defined, for example, as the set comprised by articles x, y, z. It may be intentionally
defined, for example, as the set of source materials recognized as valid in society s at time
t. Although the hearer needs to be capable of discerning what the relevant set in each
case is, it does not have to be the case that such a set is expressly mentioned or identified
in the statement itself. In fact, in what concerns what are arguably the most practically
important of our statements of law—statements relative to the whole set of source
materials which at the moment the statement is made are valid in the society to which the
hearer belongs—the relevant set is contextually evident and thus seldom made explicit.
The common operators (e.g. “it is the law that . . .,” “according to the law . . .,” “legally . .
48
.,” etc. ) are accordingly interpreted to yield statements of what the law is, at the time the
statement is being made, in the society to which the hearer and/or the speaker belong at
that same moment. The corresponding set of materials may thus be called a “static,” 49 or
“momentary,” 50 set or system.

Perfunctory as they surely are, these remarks should more than suffice for us to see that
Kletzer’s Kirchmannian‐flavoured protest—that “any pronouncement legal science . . .
makes about the law can be falsified by the law itself”—is wide of the mark. The set of
source‐materials which at any given moment is valid in a society may, and typically will,
vary. By the enactment and repeal of legislation, to take the simplest of examples, the set
of valid legal materials is through time expanded and contracted. The diachronic
subsistence of what we call a “legal order” involves the chronological succession of static
sets of source‐materials—each set a product of the “legal process.” And, of course, the
possibility that no equivalence obtains between the sets of materials valid at any two
moments t and t+1 does not entail that a true token statement of law made at t relative to
the former set becomes at t+1 a false statement. Rather, if true, it remains true. For
whether a token of the same statement‐type is true or false relative to any other set of
materials (including the set of materials valid at t+1 in the same legal “order”) is
inconsequential for the assessment of whether a token statement relative to the former
set is true.

Nor does the fact, which Kletzer also brings up, that the law “very often” “retrospectively
declares inconsistencies to be consistent and vice versa” and that “faults can heal and be
51
rendered retrospectively legal and valid pronouncements can be rescinded,” lead to
48
See, e.g., Joseph Raz, Legal Reasons, Sources, and Gaps, in THE AUTHORITY OF LAW 63 (2d ed. 2009).
49
KELSEN, PURE THEORY, supra note 22, at 70, 279 (distinguishing a “static” theory from “dynamic” one: a “static”
theory “attempts to comprehend the law without consideration of its creation, only as a created order”); see also
KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY, supra note 27, at 91; KELSEN, GENERAL THEORY, supra note 26,
at 39, 42, 122.
50
JOSEPH RAZ, THE CONCEPT OF A LEGAL SYSTEM 34 (2d ed. 1980).
51
Kletzer, supra note 1, at 800, note 25.
different conclusions. The set of materials which is at any moment t deemed valid at t may
differ from the set of materials which is at t+1 deemed valid at t; and, again, a statement
which is true relative to the former may well be false (or true) relative to the latter, or vice
versa.

Kletzer has failed to make a convincing case for the “cognitivity thesis,” which was his
positive claim. His contention that “[t]he law, it turned out, is ‘constitutive.’ Put
differently: that, of which Kant claims that it happens in knowledge, actually happens in
52
the law” comes out as simply arbitrary. What about his negative claim, the denial that a
relevant analogy does hold between “legal science” and the natural sciences? It, too, is
ill‐founded. Or so I will now suggest.

C.

Kletzer’s reiterated argument against analogizing legal and natural sciences is made
manifest in passages such as these:

[W]hereas nature as an ordered whole is inconceivable
without an experience that orders the given material,
the positive law as an ordered whole is very well
conceivable without the existence of legal science;
whereupon again the ordered whole of legally relevant
facts would be inconceivable without the positive
law. 53

[T]he law differs in one fundamental respect from
nature as an object: whereas no connection can be
given in pure receptive sensation, such a connection
can very well be given in the law. . . the law can very
well be conceived as being in itself ordered. 54

In contrast to nature the law is in itself ordered, and
can find its unity only in itself. 55

Is it the case, then, that, as we read in the first of these excerpts, “the ordered whole of
legally relevant facts would be inconceivable without the positive law”? There is in this
52
Id. at 803.
53
Id. at 795.
54
Id. at 801‐2.
55
Id.
sentence, on the one hand, the whiff of tautology: it is part of Kletzer’s own notion of a
“legally relevant fact” that no such facts are “conceivable” “without positive law.” Facts
are “legally relevant” only if such an interpretation is assigned to them by the law itself:
“Facts by themselves are just facts. There is nothing in their quality as facts that could
mark them out as legal. It is only the law that can mark out certain facts as legal facts.” 56

This is in essence the household doctrine that norms operate as “schemes of
interpretation.” 57 On the other hand, however, Kletzer errs in suggesting that the “whole”
of legally relevant facts is ipso facto—i.e. independently of legal science—“ordered” in the
suitable sense. This suggestion is clearly confounded by the possibility that any given set
of legal source‐materials, whilst validly produced by the adequate “legally relevant facts,”
contains or expresses conflicting norms. As yet again Kelsen would have it,

it is undeniable that legal organs may create conflicting
norms—that they perform acts whose subjective
meaning is an “ought” and which may be in conflict
with each other if their subjective meaning is
interpreted as their objective meaning. 58

The legal process may of course in some sense be said to be an “ordered” process for the
creation of legal source material. The point, though, is that the legal material itself cannot
plausibly be deemed “ordered” in the sense that matters here—the sense in which, in the
jargon of (Neo‐)Kantian epistemology, nature is indeed said to be “ordered.” On the
assumption, then, that “the cognition of law, like any cognition, seeks to understand its
59
subject as a meaningful whole and to describe it in noncontradictory statements,” the
analogy which Kletzer strives to lay off does appear to hold quite neatly. Recall this
exemplary passage of Kelsen’s, which for expository convenience I now quote again:

[A]ccording to Kant’s epistemology, the science of law
as cognition of law, like any cognition, has constitutive
character—it “creates” its object insofar as it
comprehends the object as a meaningful whole. Just as
the chaos of sensual perceptions becomes a cosmos,
that is, “nature” as a unified system, through the
cognition of natural science, so the multitude of
56
Id. at 809.
57
See KELSEN, INTRODUCTION TO THE PROBLEMS OF LEGAL THEORY, supra note 27, at 10; KELSEN, PURE THEORY, supra note
22, at 3ff.; and also KELSEN, GENERAL THEORY, supra note 26, at 41.
58
59
Id. at 206.
general and individual legal norms, created by the legal
organs, becomes a unitary system, a legal “order,”
through the science of law. But this “creation” has a
purely epistemological character. 60

One may therefore plausibly refer to the “judgments” of legal science, differently from
those supposed “judgments” which Kletzer claims to devise in the positive law itself, in a
sense which bears comparison to the “judgments” of the natural sciences. The parallel is
in fact far from superficial. In Kelsen, it is famously rendered as an analogy between, on
the one hand, the principle according to which the natural sciences describe their object
(the “Ordnunsprinzip” of causality, which connects two elements as cause and effect) and,
on the other hand, the principle according to which normative social sciences in general
describe their object (the principle of imputation, “Zurechnung,” which unites two
elements according to an essentially normative copula). The contrast is here between two
“methodological forms”: 61 the form of the scientific judgment, “if a is, then b is (or will
be),” and that of the normative judgment, “If a is, then b ought to be.” 62

Somewhat surprisingly, Kletzer takes issue with Kelsen’s claim that

[t]he synthetic judgments of natural science are, even
though they are “produced” by natural science,
nevertheless just as determined by the “material”
which is to be unified in them . . . as the synthetic
judgments of legal science, the Rechtssätze, are. In the
latter the material given to legal science (the statutes,
ordinances, court judgments, administrative acts, etc.)
is transformed into Rechtssätze, just as the material of
sensation is transformed in the synthetic judgments of
the natural sciences; and these synthetic judgments of
legal science, the Rechtssätze, are just as determined
60
Id. at 72.
61
On the intellectual genealogy of this notion, relating the autonomy of “normative sciences” to the respective
“methodological forms,” and Windelband’s and Rickert’s influences on the formation of Kelsen’s thought, see
Stanley L. Paulson, J. W. Harris’s Kelsen, in PROPERTIES OF LAW, esp. 14ff (2006).
62
See KELSEN, PURE THEORY, supra note 22, at 89–90; and KELSEN, GENERAL THEORY, supra note 26, at 45–6, 163–4.
The “form” of the legal judgment—the legal proposition, or Rechtssazt—is in Kelsen but a species of the general
form of the normative judgment or proposition—the Sollsatz. It corresponds to a partial interpretation of the
latter, where “b” is interpreted as an “act of coercion” determined by a given legal order. See KELSEN, PURE THEORY,
supra note 22, e.g. at 58, 89; and HANS KELSEN, GENERAL THEORY OF NORMS 272 (first endnote) (Michael Hartney
trans., 1991).
by the material given to them as the judgments of
natural sciences are. 63
Kletzer offers this as proof that Kelsen simply does not “comprehend” “the point of
Kantian philosophy.” 64 For Kelsen, he says, “seems to be reading Kant in the following
way: science has to ‘produce’ judgments which correspond to the ‘true relations’ present
in the world, or to the objective order of the world.” 65 There is nothing in the quoted
Kelsenian passage, however, that even remotely mandates such a charge. The problem,
apparently, lies in Kelsen’s reference to the judgments of natural science being
“determined” by the “material given to them,” when “after all,” says Kletzer, it is “the
point of Kantian philosophy . . . that the judgments of natural sciences are not determined
by the given material but that, conversely, the capacity of judgment determines the
material to nature.” 66 But this is uncharitable and slightly offhand. Surely Kletzer does not
mean to intimate that the judgments of natural sciences do not at all depend on what,
content‐wise, the given material of sensations is? Any given legal‐science judgment will
similarly depend on the contents of the legal material to which it is relative. To different
sets of legal material there correspond different sets of true statements of law. No sort of
commitment to “Humean sense‐data” or “picture‐theories” of cognition is evinced by the
notion that the “judgment” is in this sense “determined” by the material; nor is this notion
inconsistent with the acknowledgment that legal science is, to put it in Kletzer’s words,
67
“constitutive of the order of the law.” I am doing no more than stressing the obvious.
But should further evidence be required for the unreasonableness of this objection of
Kletzer’s, we need only note how elsewhere in the paper he himself finds it perfectly fitting
to assert that “knowledge is not a purely passive phenomenon (i.e. there is a minimal
sense in which in knowing an object we are not only determined by the object but also
determine it).” 68

Kletzer’s reading of the doctrine that norms operate as “schemes of interpretation,”
incidentally, is questionable in yet another respect. In his hands, this doctrine is taken to
63
64
Id. at 801.
65
Id.
66
Id.
67
Id. at 802.
68
imply that law “has a morality,” 69 or that law is “fundamentally moral in nature,” 70 and so

that morality “cannot by itself contradict the law:” 71

[Law and morality] are related in a way which reveals
the strict impossibility of a moral treatment of “the
law.” As “the law” itself is not a thing, but a relation of
norms and facts, since it itself is an “assessment” of
facts, it is no suitable object of moral assessment. 72

Kletzer’s point is that any supposed moral evaluation of the law is in fact an evaluation,
not, strictly speaking, of “the law,” but of “a specific act of creation of law:” 73 “When we
say, ‘This statute is morally reprehensible’, what we are primarily saying is something like
‘Parliament should not have enacted it.’” 74 And “the problem with this,” he says, “is that
the act in question can only be law if another law evaluates it as being lawful.” 75 Morality
and law, in other words, “both want to do the same thing.” 76 Both want normatively to
assess facts. Accordingly, Kletzer subscribes to Kelsen’s claim that “from the point of view
of positive law as a system of valid norms, morality does not exist as such,” 77 which he
goes on to elaborate by presenting morality as “incomplete law”:

[M]orality is “incomplete law.” The positive law has to
react to expressions of moral opinion with “infinite
indifference” and it has to do so not because the
positive law was somewhat amoral, or in a wrong sense
“pure”, or because it was not interested in morality,
but because the positive law is already the actual
elaboration of what is only subjectively claimed in
private moral utterances. Morality may be the trigger,
69
Id. at 805.
70
Id. at 811.
71
Id. at 813.
72
Id. at 811.
73
Id.
74
Id.
75
Id.
76
Id. at 810.
77
Id. at 811‐812.
the occasion of a legal process, but it cannot by itself
contradict the law. 78

But this claim, I submit, is wrong. It cannot account, e.g., for moral assessments of the
legal organs’ negative acts (such as legislative omissions); for the distinction between
prima facie and all‐things‐considered moral assessments of how legal organs have acted or
ought to have acted or to act; or for the difference between evaluative and prescriptive
moral assessments of the law. But, more generally and more importantly, Kletzer seems
here to make the mistake of supposing that “moral” and “legal” assessments of any given
fact must be similar in content or type simply because in both cases “norms” can be said to
be at play. Yet legal authorization, although it involves a “norm”‐dependent assessment,
does not convey or entail any sort of “moral” assessment of the authorized acts.

Kletzer’s seeming belief to the contrary is perhaps a reflection of Kelsen’s own error—an
79
oft‐diagnosed one —of muddling up under a single umbrella‐notion several dimensions of
legal validity. This hypothesis is supported by Kletzer’s characterization of Kelsen’s
doctrine of the basic norm. Kletzer draws no distinction at all between the presuppositions
involved in “legal cognition” and the presuppositions involved in judgments of legal
validity. This is indeed faithful to Kelsen. But it may easily be shown to be wrong. No
aspect of a description of the law turns on whether the relevant set of materials is in fact
the product of legally authorized acts of law‐creation (let alone of morally justified ones).
Something along the lines of Kelsen’s Grundnorm might arguably be needed in order for us
to represent a given set of materials as expressing a set of valid norms. But we can very
well do without the Grundnorm when determining what norms a given set of materials
expresses if it expresses a set of valid norms.

D.

I am bound to conclude that Kletzer offers in his paper no satisfactory arguments either for
the “cognitivity thesis,” or against its rival view. Moreover, the foregoing discussion has
also hinted that, contra Kletzer, (CT) forms no part of Kelsen’s views.

78
Id. at 813.
79
See Alf Ross, Validity and the Conflict between Legal Positivism and Natural Law, 4 REVISTA JURÍDICA DE BUENOS
AIRES 46ff (1961), reprinted in NORMATIVITY AND NORMS. CRITICAL PERSPECTIVES ON KELSENIAN THEMES 147ff (S.L. Paulson,
B.L. Paulson eds., 1998); Carlos Santiago Nino, Some Confusions around Kelsen’s Concept of Validity, 64 ARCHIV FÜR
RECHTS‐ UND SOZIALPHILOSOPHIE 357ff (1978), reprinted (with minor changes, and under the title El Concepto de
Validez Jurídica en la Teoría de Kelsen) in LA VALIDEZ DEL DERECHO 7ff (1985), and partially reprinted in NORMATIVITY
AND NORMS. CRITICAL PERSPECTIVES ON KELSENIAN THEMES 253ff (S.L. Paulson, B.L. Paulson eds., 1998); Eugenio Bulygin,
An Antinomy in Kelsen’s Pure Theory of Law, 3 RATIO JURIS 29ff (1990), reprinted (under the title Validez y
Positivismo) in ANÁLISIS LÓGICO Y DERECHO 499ff (1991), and reprinted in NORMATIVITY AND NORMS. CRITICAL PERSPECTIVES
ON KELSENIAN THEMES 297ff (S.L. Paulson & B.L. Paulson, eds., 1998).
As tall tales go, however, CT is not without company. Consider some of the claims made by
Kletzer regarding what is supposedly his main subject—the “Gegenstandsproblem,” the
“problem of being an object of cognition in general.” 80 While nowhere given a precise
formulation of this “problem” in connection with the legal domain, we are vaguely told
that “the Gegenstandsproblem deserves a technical term, because the better part of
contemporary jurisprudence takes, as its object, the positive law san phrase.” 81 At first
sight this suggests that the “problem” concerns the “object” of jurisprudence. But this is
not what Kletzer means. “Object,” in this excerpt, refers instead to jurisprudence’s subject
of inquiry as a discipline. In a different passage, Kletzer appears to characterize the
“problem” as concerning “the possibility of rationally relating to the positive law.” 82 This
suggestion, though, must also be ruled away. For, as we have seen, Kletzer’s overall
endeavor is aimed at denying that the “positive law” is in fact an “object” of cognition. At
any rate, and despite such imprecision, the “Gegenstandsproblem” is the subject of many a
bold assertion. The “philosophical shallowness” of the “Anglo‐American strand of legal
positivism”—and particularly of the “tradition of Hartian positivism,” derided by Kletzer as
“a mixture of quasi‐philosophy and quasi‐sociology”—is due, he informs us, to its “lack of a
proper engagement with the ‘problem of being an object of cognition in general.’” 83 How
so? Unfortunately, we are never told. Thus these diagnoses and accusations—which very
often sound wrong‐headed 84 —are baldly delivered in a matter‐of‐fact manner that
nonchalantly dispenses with the support of argument.
80
81
Id. at 786.
82
Id. at 790.
83
Id. at 785‐786. The problem is in turn suggested to be well‐known to Continental theorists; and the Pure
Theory of Law is said to owe its “philosophic sophistication” precisely to its “early engagement” with the
Gegenstandsproblem: see id. at 786, note 1.
84
Take, for instance, Kletzer’s assertion (as well as the ensuing discussing in terms of divided “competences”), at
786, that “Hartian positivism,” “apart from being a theory of the positive law, is so obviously also a theory of the
division of labour between itself and the positive law.” This oddly depicts “Hartian positivism” as if it were
essentially, if only in part, a meta‐theoretical stance. Or take, for another example, Kletzer’s unqualified
endorsement of Cotterell’s account of Hart’s “linguistic empiricism” at 788. In his discussion of ordinary‐language
philosophy, however, Cotterell confuses methods and subject‐matter, saying, e.g., that according to Hart “legal
statements” “represent” (!) the (“observable”) “reality of linguistic practices of people living within a legal
system.” Kletzer, supra note 1, at 789 (emphasis added).

PDF - Vol - 12 - No - 02 - Complete Issue Final

Caricato da

Informazioni sul documento

Descrizione originale:

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

PDF - Vol - 12 - No - 02 - Complete Issue Final

Caricato da

Copyright:

Formati disponibili

GERMAN LAW JOURNAL

Vol. 12 No. 02 Pages 599‐826 01 February 2011

Suppose that a judge sentences me to prison, citing a

Jeremy Waldron, Legal and Political Philosophy, in THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW

It is useful to start that assessment with a clarification of what interpretivism is. 108 I prefer

Odious as retrospective criminal legislation and

tells us to acknowledge as law only in a “pre‐

I leave it to the reader to conclude how much is left of conventionalism here. 123 I should

tell them is that agreements consist of contracts of

Perhaps nothing illustrates the gulf which still

empiricist view would say that a theory’s truth can be

Level 1 Natural Science Legal Science

Ground Level Nature Law

Level 2 Philosophy Legal Science

Level 1 Natural Science Law

Ground Level Nature Legal Facts

determined by the material given to them as the

of nature. Rather, what makes such an event a legal

Level 1 Sociology Law

Each law‐creating act must be a law‐applying act—it

imply that law “has a morality,” 69 or that law is “fundamentally moral in nature,” 70 and so

Potrebbero piacerti anche