Flores Fabra 50th Anniversary of Harts Concep of Law

PROBLEMA
ANUARIO DE FILOSOFA
Y TEORA DEL DERECHO
5
2011
INSTITUTO DE INVESTIGACIONES JURDICAS

UNIVERSIDAD NACIONAL AUTNOMA DE MXICO

Dr. Hctor Fix-Fierro
Director
Dra. Mnica Gonzlez Contr
Secretaria Acadmica
Dra. Elvia Luca Flores valos
Jefa del Departamento de Publicaciones
De las opiniones sustentadas en los trabajos responden exclusivamente

sus autores. El hecho de su publicacin no implica en manera alguna
que esta revista, Problema, Anuario de Filosofa y Teora del Derecho, se
solidarice con su contenido.
Formacin en computadora: Karla B. Templos Nuez

Diseo de portada: Carlos Martn Aguilera Ortiz
PROBLEMA
Anuario de Filosofa y Teora del Derecho
Dr. Juan Vega Gmez
Director
Dra. Elvia Luca Flores valos

Coordinadora editorial
Asistentes de la revista
Sandra Gmora Jurez
Ricardo Reyes Mrquez
Enrique Rodrguez Trujano
Este nmero incluye un cd-rom con el contenido de los nmeros

1 al 5 de Problema. Anuario de Filosofa y Teora del Derecho.
Problema es una revista mexicana que contiene artculos en espaol e ingls sobre diversos temas de filosofa poltica, jurdica y
moral. Esta revista es publicada anualmente por el rea de Filosofa y Teora del Derecho del Instituto de Investigaciones Jurdicas de la Universidad Nacional Autnoma de Mxico.
Problema is a bilingual journal spanish/english that publishes papers relating to legal, political and moral philosophy.
Problema is published annually by the Philosophy of Law and
Legal Theory Department of the Legal Research Institute, at National Autonomous University of Mexico.
Email: problema.unam@gmail.com
Enrique Cceres Nieto / Imer B. Flores / Juan Vega Gmez

Fundadores
CONSEJO EDITORIAL / EDITORIAL BOARD
Jorge Adame Goddard (Instituto de Investigaciones Jurdicas,
UNAM); Zenon Bankowski (Edinburgh Law School); Samantha
Besson (University of Fribourg); Brian Bix (University of Minnesota);
Brian Burge-Hendrix (Quest University); Enrique Cceres Nieto
(Instituto de Investigaciones Jurdicas, UNAM); Tom Campbell
(Charles Sturt University Australia); Jaime Crdenas (Instituto de
Investigaciones Jurdicas, UNAM); Jules L. Coleman (Yale Law
School); Keith Culver (University of New Brunswick); Julie Dickson
(Oxford University); Paulette Dieterlen (Instituto de Investigaciones
Filosficas, UNAM); Ronald Dworkin (New York University/University College London); Timothy Endicott (Oxford University); Imer B.
Flores (Instituto de Investigaciones Jurdicas, UNAM); John Gardner
(Oxford University); Michael Giudice (York University, Canada);
Mark Greenberg (University of California, Los ngeles); Olga E.
Hansberg (Instituto de Investigaciones Filosficas, UNAM); Kenneth
Einar Himma (Seattle Pacific University); Iwao Hirose (McGill University, Canad); Carla Huerta Ochoa (Instituto de Investigaciones
Jurdicas, UNAM); Guillermo Hurtado (Instituto de Investigaciones
Filosficas, UNAM); Duncan Kennedy (Harvard Law School); Matthew
H. Kramer (Cambridge University); Dimitrios Kyritsis (University of
Sheffield); Gerald Lang (Leeds University); Larry Laudan (Instituto de
Investigaciones Filosficas, UNAM); Brian Leiter (University of Chicago); Neil MacCormick ( ) (University of Edinburgh); Andrei Marmor
(University of Southern California); Guillermo J. Man Garibay
(Instituto de Investigaciones Jurdicas, UNAM); Carlos Montemayor
(San Francisco State University); Joseph Raz (Columbia University/Oxford University); Mara Elodia Robles (Facultad de Derecho,
UNAM); Javier Saldaa Serrano (Instituto de Investigaciones
Jurdicas, UNAM); Burkhard Schafer (University of Edinburgh); Frederick Schauer (University of Virginia); Scott J. Shapiro (Yale Law
School); Nicos Stavropoulos (Oxford University); Natalie Stoljar
(McGill University); Rolando Tamayo y Salmorn (Facultad de
Derecho, UNAM); Juan Vega Gmez (Instituto de Investigaciones
Jurdicas, UNAM); Ambrosio Velasco Gmez (Instituto de Investigaciones Filosficas, UNAM); Enrique Villanueva (Instituto de Investigaciones Jurdicas, UNAM); Jeremy Waldron (New York University); Wil Waluchow (McMaster University, Canada); Benjamin
Zipursky (Fordham Law School).
DISTRIBUIDOR:
Universidad Nacional Autnoma de Mxico,
Instituto de Investigaciones Jurdicas
Coordinacin de Distribucin, Promocin y Fomento Editorial,
responsable: arquitecta Carola Lagunes.
Circuito Maestro Mario de la Cueva s/n, Ciudad de la Investigacin en
Humanidades, Ciudad Universitaria, 04510 Mxico, D. F., Tels. 5622
7474 ext. 1703, carola@servidor.unam.mx.
Nmero de reserva al ttulo en Derechos de Autor:

04-2010-102014282600-102.
Nmero de certificado de licitud de ttulo: en trmite.
Nmero de licitud de contenido: en trmite.
Primera edicin: 15 de octubre de 2011
DR 2011, Universidad Nacional Autnoma de Mxico
Circuito Maestro Mario de la Cueva s/n
Ciudad de la Investigacin en Humanidades
Ciudad Universitaria, 04510 Mxico, D. F.
Impreso y hecho en Mxico
ISSN en trmite.
Editores invitados
Guest Editors
Imer B. FLORES
Legal Research Institute, UNAM
Jorge FABRA
Graduate Student, Department of Philosophy,
McMaster University
CONTENIDO
TABLE OF CONTENTS
En torno a El concepto de derecho de Hart en su cincuenta aniversario. Introduccin . . . . . . . . . . . . .
Imer B. FLORES
Jorge FABRA
50th Anniversary of Harts The Concept of Law. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .
Imer B. FLORES
Jorge FABRA
XIII
XVII
DISCUSIN: EN TORNO A EL CONCEPTO DE DERECHO

DE HART EN SU CINCUENTA ANIVERSARIO
DISCUSSION : 50th ANNIVERSARY OF HARTS
THE CONCEPT OF LAW
H. L. A. Hart: Supervisor, Mentor, Friend, Inspiration .
W. J. WALUCHOW
H. L. A. Hart. . . . . . . . . . . . . . . . . . . . . . .
Matt KRAMER
11
Rewriting Harts Postscript: Thoughts on the Development of Legal Positivism . . . . . . . . . . . . . . .

Tom CAMPBELL
IX
23
CONTENIDO
The Simple and Sweet Virtues of Analysis. A Plea for
Harts Metaphilosophy of Law . . . . . . . . . . . .
Pierluigi CHIASSONI
Pulling Off the Mask of Law: A Renewed Research Agenda
for Analytical Legal Theory . . . . . . . . . . . . . .
Keith CULVER
Michael GIUDICE
Descriptive Jurisprudence . . . . . . . . . . . . . . .
Pavlos ELEFTHERIADIS
H. L. A. Harts Moderate Indeterminacy Thesis Reconsidered: In Between Scylla and Charybdis? . . . . . .
Imer B. FLORES
Are Legal Rules Content-Independent Reasons? . . . .
Noam GUR
Towards a Comprehensive Positivist Theory of Legal Obligation . . . . . . . . . . . . . . . . . . . . . . . . .
Kenneth Einar HIMMA
53
81
117
147
175
211
Harts The Concept of Law as a Study in Social Philosophy

Eerik LAGERSPETZ
243
Farewell to the Rule of Recognition . . . . . . . . . . .

Giorgio PINO
265
H. L. A. Hart and the Invention of Legal Philosophy. . .

Dan PRIEL
301
Hartian Positivism as a (Plausible) Error Theory . . . .

Fbio Perin SHECAIRA
325
Hart on Judicial Discretion . . . . . . . . . . . . . . .

Roger A. SHINER
341
CONTENIDO
Lessons from Hart. . . . . . . . . . . . . . . . . . . .
W. J. WALUCHOW
363
Lessons from Hart: A Reply to Waluchow . . . . . . . .

Kenneth Einar HIMMA
385
ARTCULOS
ARTICLES
The Nature of Law and Reasons for Action . . . . . . .
Brian H. BIX
Brian Bix and Mexican Legal Philosophy: a Comment
From the Perspective of the Sociology of Knowledge .
Enrique CCERES NIETO
399
417
RESEAS
HART, H. L. A., Punishment and Responsibility. Essays in
the Philosophy of Law, 2a. ed., Oxford, Oxford University Press, 2008, liii, 277 . . . . . . . . . . . . . . .
Luis NAVARRETE
439
Autores . . . . . . . . . . . . . . . . . . . . . . . . .
471
XI
P ROBLEMA . A NUARIO DE F ILOSOFA Y T EORA DEL

DERECHO, nm. 5, 2011, es una publicacin anual
editada por el Instituto de Investigaciones Jurdicas
de la UNAM. Editor responsable: Juan Vega Gmez. Impresa por Formacin Grfica S. A. de C. V.,
Matamoros 112, colonia Ral Romero, Ciudad Nezahualcyotl, 57630 Estado de Mxico, tel. 5797
6060. Este nmero se termin de imprimir el 15 de
octubre de 2011. En su edicin se emple tipo Bookman Old Style en 6, 8, 8.6, 9, 9.5, 10, 11, 35, 45 puntos; Book Antiqua en 17 puntos, Arial en 9 puntos. En
su edicin se emple papel cultural 70 x 95 de 50 kilos para la pginas interiores y cartulina couch de
162 kilos para los forros. Consta de 500 ejemplares
(impresin offset).
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
H. L. A. HART: SUPERVISOR, MENTOR, FRIEND,

INSPIRATION
W. J. WALUCHOW
Some of the ideas developed and defended in this book originated in my doctoral thesis, Adjudication and Discretion,
submitted at Oxford University. I am delighted once again to acknowledge my
overwhelming debt to Herbert Hart who
supervised the thesis and provided me
with the kind of guidance and encouragement of which doctoral students dreams
are made. Professor Hart has continued to
offer his insightful comments on my written work, and for his ongoing assistance
and generosity I am extremely grateful.
So began the acknowledgments page of my book, Inclusive

Legal Positivism.1 To this day, I continue to feel the force of
Harts influence, both in my philosophical work and, perhaps more importantly, in my relationships with friends,
colleagues, and above all else, my students. We all know,
quite well, the many fine scholarly qualities displayed in
Harts wide-ranging philosophical work: his sheer philosophical insight and clarity of thought; an ease of expres1
W. J. Waluchow, Inclusive Legal Positivism (Oxford: Clarendon Press, 1994),
vi.
W. J. WALUCHOW
sion that often obscured the depth of his understanding;

his appreciation of the immense complexity of the many important issues he addressed; the fairness and open-mindedness with which he dealt with opposing views; and his
willingness to call a spade a spade when he found it necessary to do so. With the publication of Nicola Laceys fascinating biography, we now know much more about Harts
complicated personal life.2 One aspect touched upon by
Lacey is the scholarly relationships Hart shared with his
students, many of whom went on to pursue successful academic careers of their own: Brian Barry, John Finnis, Neil
MacCormick and Joseph Raz, to name just a few. In her final chapter, Lacey discusses Harts complicated relationship with one of his most distinguished students, Ronald
Dworkin, who went on to succeed Hart in the Oxford Chair
of Jurisprudence, a post Dworkin held while I was a student in Oxford. As is well known, a good deal of Dworkins
early work consisted of a full-scale critique of Harts legal
positivism, a theory about the nature of law that Dworkin
hoped to dislodge as the ruling theory of law.3 In my doctoral dissertation, I attempted to defend Hartian positivism
against Dworkins powerful critique, and had the good fortune of being able to consult both scholars. Despite being
officially retired, Hart agreed to supervise my doctoral
dissertation, while Dworkin, while not officially attached to
me in any way, graciously agreed to read some of my work.
I recall visiting Dworkin, on one memorable occasion in
his rooms in New College, to discuss a very long paper I
had written that dealt with his critique of Hart. This paper,
which I had earlier submitted to Hart and which will figure
prominently in the first of the two stories recounted below,
contained the main lines of argument I would later develop
both in my dissertation and, years later, in Inclusive Legal
Positivism. One of Dworkins first questions to me was:
2 Nicola Lacey, A Life of H. L. A. Hart: The Nightmare and the Noble Dream (Oxford, Oxford University Press, 2004).
3 Ronald Dworkin, Taking Rights Seriously (Cambridge, Harvard University
Press, 1978), vi.
H. L. A. HART: SUPERVISOR, MENTOR, FRIEND, INSPIRATION
What does Herbert think of this? to which I replied: He

seems to agree with the line of argument [inclusive legal
positivism] it pursues. Dworkin shook his head in what I
could only interpret as utter dismay. Despite this, he and I
went on to have a very productive (at least for me) meeting.
A few days after this encounter, I arranged to meet with
Hart to get his take on another paper I had written. One of
Herberts first questions to me was: What does Ronnie
think of the paper you had him read? to which I replied:
Not much. He smiled and we too went on to have a very
productive meeting.
In her book, Professor Lacey recounts a bit of this story
and takes it to signify the unfortunate rift that appeared to
have developed between Hart and his distinguished student. Whether that is so is not for me to judge, and is certainly not a topic upon which I care to comment here. What
I wish to do instead is share two further stories that reveal
the kind of relationship Hart and I enjoyed and which, I can
only imagine, he must have shared with most, if not all, of
his other students. These stories will, I hope, serve not only
to acknowledge further the enormous debt of gratitude I
owe my supervisor. I hope they will also help solidify the
conviction among many that in Herbert Hart what one encountered was a truly remarkable human being.
The Nightmare and the Magnificent Dream
When I first began at Oxford, I was assigned a supervisor
who, much to my dismay, did not provide me with the type
of supervision for which I had hoped. R. M. Hare, though
highly accomplished in the field of moral philosophy, had
little background or interest in the philosophy of law, the
subject upon which I had decided to write. Despite this unfortunate situation I persisted, attempting as best I could to
produce work that might someday form parts of a doctoral
dissertation in the subject. But I was clearly floundering
and knew myself that what I was coming up with just didnt
5
W. J. WALUCHOW
pass muster. But I didnt know quite what to do about it.

Well into my second year, I had the good fortune of encountering John Mackie at a legal theory conference in Lancaster, England. Upon hearing of my predicament Mackie
kindly agreed to read my work upon his return to Oxford. I
sent him everything I had written to that point, was duly
summoned to his rooms, and was issued the verdict I had
feared but expected: for the most part, the work just wasnt
very good. There was one exception, however: a short piece
on what was becoming known as the Hart-Dworkin debate. The specific question upon which I had focused was
whether Harts positivism is, as Dworkin maintained, restricted to content-neutral, pedigree tests of legal validity.
My answer was a resounding no and I went on to argue
that Harts famous rule of recognition can easily accommodate conformity with a set of moral values and/or principles as a necessary condition of legal validity, something
Hart himself had acknowledged in a variety of places, and
which he later confirmed in his Postscript to The Concept of
Law.4
Buoyed by Mackies favourable judgment of the piece, I
endeavoured to develop it. [It eventually morphed into the
longish paper mentioned in the preceding section, the one
on which Hart and Dworkin wanted to know the others
views.] I delivered it to Hare who suggested that perhaps we
should solicit Harts views on the paper. I was of course delighted with the idea, and so Hare, right on the spot, tele4 In an early paper, Hart made this suggestion: Bentham indeed recognized,
as Austin did not, that even the supreme legislative power might be subjected to legal restraints by a constitution and would not have denied that moral principles,
like those of the fifth amendment, might form the content of such legal constitutional restraints. (Positivism and the Separation of Law and Morals Harvard Law
Review, Vol. 71, No. 4 (Feb., 1958), 599. See also The Concept of Law, 2nd ed. (Oxford: Clarendon Press, 1994), p. 204 where Hart suggests that In some systems [of
law], as in the United States, the ultimate criteria of validity explicitly incorporate
principles of justice or substantive moral values. In the Postscript to his famous
book (p. 250), Hart comments on Dworkins attribution to him of exclusive or
plain-fact positivism by noting that it ignores my explicit acknowledgement that
the rule of recognition may incorporate as criteria of legal validity conformity with
moral principles or substantive values.
phoned Hart who agreed to read my paper. Soon thereafter,

I deposited a copy in Harts mailbox at University College.
Before hearing back from Hart, whom I had not yet met, I
was summoned to a meeting with Hare to discuss my paper
and made my way to his study in Corpus Christi College.
Upon entering Hares room and preparing to take my usual
place in the chair opposite his, I was, to put it quite simply,
broadsided. Before I was able to take my seat Hare informed me that there was no need for me to sit down. He
had not, he explained, managed to read past page two. I
have to confess, he added, that I found your paper a
frightful bore, at which point he handed it back to me and,
without further ado, immediately went back to work. It is
difficult to explain how distressed this left me. It seemed as
if my scholarly career had just been issued a death sentence. It is one thing to be wrong; it is quite another to have
laboured away for weeks only to come up with something
ones supervisor had no interest in reading. There seemed
nothing left for me to do except leave Corpus, take the bus
home, and begin contemplating whether I should try to
cobble together enough money for a flight back to Canada.
What I had written was the best that I could come up with.
If that wasnt worth reading, then it was obvious that the
time had come to pursue another career.
Upon arriving at Wolfson College, where I had a college
flat, I decided to check my mailbox. In it was a letter that
changed my life and set me on my way to the career that I
now enjoy. It was from Herbert Hart who had finished reading the paper I had submitted to him and which had just
been the subject of Hares less than positive reaction. Hart
began with the following words which are indelibly imprinted on my memory and which I will never forget: My
Dear Waluchow. I am reading with care and fascination
your excellent chapter on Dworkin. Whereas words cannot
capture how distressed I felt less than an hour ago in
Hares rooms, words cannot begin to express the feelings of
utter elation I experienced upon reading this reaction to my
7
W. J. WALUCHOW
work. Not only did H. L. A. Hart read my paper; and not

only did he find it interesting and its arguments plausible.
He found it fascinating. Hart went on, in his letter, to issue an invitation to dinner at University College, and thus
began a scholarly relationship that saw him eventually replace Hare as my doctoral supervisor and which I immensely enjoyed and from which I greatly profited until his
death in 1992.
The Supervisor of Which Graduate Dreams are Made
My second story takes place in the final few months of
my three-year stay in Oxford. Hart had, by this point, replaced Hare as my doctoral supervisor and I was diligently
working away, trying to finish my dissertation by the end of
the current academic year. Rough drafts of the final three
chapters were complete, but much work had yet to be done.
And then tragedy hit. My youngest brother died unexpectedly and I needed to be with family in Canada. My stay in
Canada turned out to be longer than anticipated and upon
my eventual return to Oxford, several weeks later, I faced
what appeared to be an impossible task completing the
dissertation in an even shorter period of time. Many suggested that I abandon that objective and take an additional
year to submit. For a variety of reasons this was an option
that really didnt appeal to me. So I decided to give it my
best shot. I would attempt to finish by the end of the current year, daunting and near-impossible as that task appeared to be. I completed my next chapter and, as per the
usual practice, I submitted it to Hart, expecting to be summoned within the next two to three weeks for discussion
and criticism. The next morning I heard a gentle knock on
the front door of my apartment which, I hasten to add, was
in Kennington, a small village some five miles outside of
Oxford. When I opened the door, I was astonished to find
Herbert Hart, standing beside his bicycle, with my chapter
in hand. As best as I can remember, this is more or less
8
what he had to say: I was out for a ride and thought Id

drop by to see if you wanted to discuss your chapter. I
know youre working under a tight schedule. Despite the
fact that neither I nor my apartment were in any shape to
receive company, I invited Hart in. We met in my study for
several hours discussing the chapter and the ones that
were to follow. I managed to finish on time.
Lessons Learned
There is much we can all learn from the philosophy and
philosophical approach of H. L. A. Hart. His many contributions to legal philosophy and its associated fields are and
continue to be monumental. He stands among the giants of
20th century philosophy of law and helped set in place a set
of questions and modes of addressing them that persist to
this day. About all this much has already been said and I
will have nothing further to add here. But what I do want to
add are some final thoughts on Hart the man, the supervisor, the friend. In reflecting on the two stories recounted
above, I am struck by the enormous influence he had on
my life and, I am sure, all those who had the great fortune
to have known him. Herbert Hart was not only a talented,
accomplished scholar. He was a kind, generous man who
always made me feel like I counted both as a scholar and a
person. When in his presence, I never once felt anything
but complete support and understanding. Hart always conveyed his belief that my views on whatever issues we were
discussing were worth considering, even though he was the
master and I the student. I also learned, from Hart and in
a different way, Hare how a few words can mean so very
much, one way or the other, to a person struggling to make
his mark in the world. Had I not had the great fortune of
receiving Harts letter after my meeting with Hare, I might
never had gone on in philosophy and enjoyed whatever
meager success I have had as an academic. Had I not been
so lucky as to have a supervisor who was willing to drop ev9
W. J. WALUCHOW
erything to help me meet an urgent deadline, I might never

have finished my dissertation. These are the thoughts I entertain when I remember my supervisor and the pivotal role
he played in helping to make me the scholar and person I
am today. These are the thoughts I endeavour to keep
firmly in mind when, in dealing with my own graduate students, I struggle to meet the very high standard he set for
me. Thank you Herbert.
10
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
H. L. A. HART
Matt KRAMER
Herbert Lionel Adolphus Hart (1907-1992) was the foremost
legal philosopher and one of the foremost political philosophers of the twentieth century. Born to a Jewish family in
Yorkshire, England, Hart pursued his undergraduate education at Oxford and went on to qualify as a barrister. After
practicing law for several years, he worked for the British
intelligence service MI5 during the Second World War.
When the War had ended, he returned to Oxford to take up
a Fellowship in Philosophy at New College. He subsequently
became Professor of Jurisprudence at Oxford (and a Fellow
of University College), and still later became Principal of
Brasenose College. During the closing years of his career as
an active scholar, he devoted much of his time to editing
and interpreting the works of Jeremy Bentham. Many former students of his, including Ronald Dworkin, Joseph
Raz, John Finnis, Neil MacCormick, Herbert Morris, and
Wilfrid Waluchow, have been among the most prominent legal philosophers of the next generation.
Hart is best known for his contributions to legal philosophy generally and to legal positivism specifically. While acknowledging his intellectual debts to his great positivist
predecessors Bentham and John Austin, he severely criticized their theories for obscuring the normative dimension
of law (that is, laws orientation toward what ought to be).
At the same time, he emphasized that the normativity of
law is not necessarily moral; throughout his jurisprudential
11
MATT KRAMER
work, he maintained a legal-positivist insistence on the separability of law and morality. In his classic 1961 book The
Concept of Law, and in a number of essays written approximately contemporaneously, he presented a hugely influential account of the ways in which different types of norms
combine to form the structure of a legal system. He laid
particular stress on what he designated as the Rule of Recognition namely, the array of normative presuppositions
that underlie the behavior of legal officials (especially
judges and administrators) as they ascertain the existence
and contents of the laws in their system of governance. Under the prevailing Rule of Recognition in a jurisdiction, legal
officials are both authorized and obligated to follow specific
criteria in determining which norms possess the status of
laws. Those criteria typically fix upon familiar sources of law
such as legislative enactments or adjudicative rulings or
administrative regulations or constitutional provisions.
The Rule of Recognition is one of three main types of secondary norms, which Hart distinguished from the primary
norms of any legal system. Whereas the primary norms are
duty-imposing and liberty-conferring laws that are all directly applicable to the conduct of ordinary members of the
public, many of the secondary norms of a legal system are
juristic standards that are directly addressed to the officials
of the system only. The category of secondary norms comprises not only the Rule of Recognition, but also rules of
change (which confer private powers or public powers to alter the existing legal norms) and rules of adjudication
(which empower and obligate certain officials to deal with
actual or alleged violations of the prevailing laws). The Concept of Law presents the combination of primary laws and
secondary laws as the central structural feature of every legal system.
Hart famously distinguished between certain attitudes or
perspectives that can be adopted in relation to any pattern
of norm-guided behavior such as the operations of a system
of legal governance. A theorist who adopts an extreme ex12
H. L. A. HART
ternal viewpoint will regard such behavior as an array of

sheer regularities akin to the movements of cloud-formation
or of atmospheric pressure. Such a theorist makes no effort
to see peoples actions in the ways in which they are seen
by the people themselves. Much more sensible is a moderate external perspective. A theorist who takes up such a
stance is of course attentive to the observable regularities
that are the preoccupation of the extreme external perspective, but he focuses centrally as well on peoples attitudes
and beliefs. In particular, he attributes normative attitudes and beliefs to people whose behavior indicates that
they accept certain standards as binding upon them.
Though he does not regard himself as bound by those
standards at least insofar as he occupies an external perspective he highlights the attitude of commitment that is
felt by the people whose conduct he is investigating.
Both of these versions of the external perspective differ
from the internal point of view, which is the stance of an
engaged participant in a practice. Somebody who occupies
the internal point of view is seeking to sustain the workings
of an institution or a practice, for moral reasons or for prudential reasons. Such a person takes a censorious attitude
toward violations of the norms of the institution or practice
including her own violations and she cites the norms in
justification of her criticism. She demands compliance with
the norms as such, and she acknowledges the warrantedness of any pertinent demands and criticisms that are directed at her with reference to those norms by her fellow
participants. The attitude of endorsement is what distinguishes this perspective from either of the external perspectives delineated above. To embrace the internal point of
view is to forsake disengagement by committing oneself to
upholding the norms of some practice or institution.
For Hart, the characteristic stance of the officials who
run a legal-governmental system is the internal perspective.
Indeed, he maintained that the adoption of that perspective
by most such officials on most occasions is a necessary
13
MATT KRAMER
condition for the existence of any such system (or, at least,

for the existence of any straightforward instance of such a
system). By contrast, the characteristic stance of a jurisprudential theorist is the moderate external point of view.
Some legal sociologists might favor the extreme external
perspective, but as Hart argued at several junctures in
The Concept of Law any such approach is far too impoverished for jurisprudential purposes. To attain a satisfactory
philosophical understanding of the operations of any legal
system, a theorist must apprehend the normativity of those
operations.
Though Harts rejection of the extreme external perspective has been followed by virtually every philosopher of law,
his embrace of the moderate external perspective for jurisprudential analysis has been challenged by Ronald Dworkin
for the past four decades. Dworkin contends that an illuminating account of law must be elaborated from the viewpoint of a committed participant. He holds that any such
account has to be constructive in the sense of being favorably disposed toward the phenomenon which it expounds;
it has to seek to portray that phenomenon in the most appealing light. Although this insistence by Dworkin is an element of his longstanding opposition to legal positivisms affirmation of the separability of law and morality, many
opponents of positivism distance themselves from his stand
on this point. Quite a few of those opponents, such as
Finnis, join Hart in taking up a moderate external point of
view when they seek to expound the nature of law.
Still, although Finnis aligns himself with Hart against
Dworkin on the methodological matter just mentioned, his
understanding of the ambitions of jurisprudential enquiry
differs significantly from Harts. Finnis maintains that the
central role of such enquiry is to highlight the morally valuable functions that are performed by the operations of legal
systems. He therefore contends that the moderate external
perspective of the jurisprudential theorist should be oriented toward legal officials whose aims in their activities as
14
H. L. A. HART
officials are morally benign. By contrast, Hart took the central role of jurisprudential enquiry to be theoretical-explanatory. He sought to provide a philosophical explication of
the concept of law which would comprehend all clear-cut
legal systems and which could potentially be extended to
cover all other legal systems as well. He wanted to leave
open the nature of the aims of the officials in particular legal systems, instead of focusing centrally on officials whose
motivations are benign. He recognized of course that the officials in many standard legal systems do act on the basis
of morally salutary considerations, but he endeavored to
come up with an account of law that would cover also the
standard legal systems in which the officials collectively
pursue malevolently self-interested aims. Such an account
can highlight the features of the rule of law that are common to the benevolent regimes and the malevolent regimes,
and can explore how those features are promotive of benign
ends and many malign ends. Harts theoretical-explanatory
project was thus markedly different from Finniss morally
fraught approach to jurisprudential theorizing.
In his opposition to the positions taken by Dworkin and
Finnis, and in his opposition to the positions taken by numerous other natural-law theorists such as Lon Fuller,
Hart upheld legal positivisms insistence on the separability
of law and morality. Though some subsequent legal
positivists have spoken of a separability thesis, any such
language is highly misleading. Legal positivists differ among
themselves concerning what the separability of law and morality involves, and the debates between sundry positivists
and natural-law theorists have been multifarious. The socalled separability thesis is in fact a congeries of theses.
Hart patently recognized as much, for he disentangled a
number of respects in which law and morality have been
perceived as necessarily connected. His initial concern,
which he shared with Austin, was to deny that a norm
must satisfy some threshold test of moral legitimacy if it is
to count as a genuine law. However, he then broadened the
15
MATT KRAMER
scope of his reflections as he pondered a number of other

efforts by natural-law theorists to postulate necessary connections between law and morality. For example, he considered and rejected the thesis that every legal mandate simply by dint of its status as such a mandate imposes on
every one of its addressees a pro-tanto moral obligation of
compliance. One thing that has become apparent as a result of Harts jousting with diverse natural-law theorists is
that some distinct conceptions or dimensions of morality
are varyingly in play when the separability of law and morality is under discussion. The multiplicity of those conceptions or dimensions is partly what accounts for the
numerousness of the disputes between legal positivists and
their foes.
Although The Concept of Law is principally a work of legal
philosophy, it contains some important discussions of topics in political and moral philosophy. Harts first major contribution to political philosophy occurred in his 1955 essay
Are There Any Natural Rights? In that essay, Hart briefly
introduced a theory of political obligation that has come to
be known as the principle of fair play (a principle later
elaborated by his friend John Rawls). That is, he contended
that anyone who benefits greatly from the presence of some
institution is morally required to bear a commensurate
share of the burdens of sustaining that institutions existence. Although the principle of fair play has often come
under attack in the decades since Hart fleetingly propounded it, it continues to be espoused by some presentday political philosophers.
On the basis of some remarks made en passant by Hart
in The Concept of Law, the principle of fair play has also
been developed by quite a few philosophers of criminal law
into a version of the retributivist justification for punishment. Their claim is that somebody who commits a crime
has benefited from the law-abidance of others without fully
bearing his own share of the burdens of law-abidance.
Such a wrongdoer should be subjected to punishment in
16
H. L. A. HART
order to deprive him of his unjust gain. Proponents of this

version of retributivism have sometimes encountered difficulty in specifying credibly the nature of the gain that is
undergone by a criminal. However, the plausibility of the
theory can be salvaged if the gain is understood as the
self-indulgence of a malefactor that consists in his use of
his body and certain external things in forbidden ways. The
extent of that self-indulgence in any particular case is determined not by what the malefactor would be willing to
pay to commit his crime with impunity, but by what the
community would demand in return for allowing his
commission of it.
Hart himself argued in his Punishment and Responsibility
for a deterrence-oriented account of punishment, combined
with retributivist side-constraints. That is, he maintained
that the central purpose of punishment is the reduction of
crime through deterrence; the aim of crime-reduction is
what justifies the use of sanctions and is what determines
the sanctions appropriate levels of severity. Retributivisms
insistence on guilt as a necessary condition for the legitimacy of the imposition of any punitive measures is pertinent
not when we are addressing the purpose of such measures,
but instead when we are pinning down the circumstances in
which they can properly be imposed. Retributivism with its
emphasis on moral responsibility circumscribes the range of
situations in which someone can legitimately be punished,
as it disallows the use of sanctions against anybody who has
not committed the crimes for which the sanctions are levied.
Supplemented in this fashion with retributivistic side-constraints, a deterrence-oriented justification of punishment
will have shed some of its most dismaying implications.
Harts most sustained entry into political disputation occurred in 1963, with the publication of his Law, Liberty,
and Morality. Hart wrote in the liberal tradition of John Stuart Mill by arguing that fornication and homosexual intercourse between consenting adults should not be legally proscribed. Invoking and defending Mills harm principle,
17
MATT KRAMER
which maintains that no mode of conduct can legitimately

be outlawed unless the conduct causes nontrivial harm to
somebody other than the person who engages in it, Hart
submitted that consensual intercourse between adult homosexuals does not cause any detriment that would suffice
to satisfy the harm principle. In particular, the mere fact
that unorthodox sexual practices cause offense to some
people who are aware of their occurrence does not constitute harm of any kind that would render legitimate the prohibition of those practices. Hence, although Harts position
in Law, Liberty, and Morality was broadly utilitarian with a
focus on the misery caused by legal prohibitions that bar
people from behavior which offends some other people but
which is otherwise harmless, his utilitarianism was not thoroughgoing. As a liberal, Hart was not prepared to treat the
disutility of mere offense as a consideration to be weighed
against countervailing considerations in his assessment of
the legitimacy of various legal-governmental regulations.
In several writings included in his Essays on Bentham
and Essays in Jurisprudence and Philosophy, Hart made
noteworthy contributions to debates over the nature of
rights and justice. He joined Rawls and Robert Nozick in rejecting utilitarian rationales for sacrificing the vital interests of some individuals in furtherance of the interests of
others, but he likewise assailed the extreme individualism
of Nozicks libertarian principles of justice. Generally sympathetic to Rawlss ideas, Hart nonetheless challenged
Rawlss remarks about the overriding priority of liberty.
(Among other things, he queried Rawlss unexplained shift
from speaking about the priority of liberty to speaking
about the priority of liberties). What Hart impugned was not
really the priority of certain liberties but instead the claim
by Rawls to have derived that priority from a situation of
pure rational choice the Original Position in which
each choosing agent seeks to promote his or her own interests optimally. Rather, Hart declared, the only tenable basis
18
H. L. A. HART
for Rawlss prioritization of certain liberties is a liberal ideal

of human personality.
Hart further exhibited his liberal allegiances in his
analysis of the nature of rights. He contended that the
holding of a legal right by any person P always involves
the vesting of P with legal powers to waive or demand the
enforcement of the legal duty that is correlated with the
right. Hart adopted this analysis precisely because he believed that no alternative conception of right-holding would
capture the role of rights in enabling individual self-determination. His account of rights, like his other political
stances, was grounded on liberal values.
One additional major achievement by Hart that is well
worth noting here is his analysis of causation in the magisterial volume Causation in the Law which he co-authored
with Tony Honor. Working with an abundance of cases in
tort law and criminal law from several jurisdictions, Hart
and Honor sought to distill the principles which explicitly
or implicitly underlie the causal judgments that are
reached in such cases. Their central analysis of causation,
whereby an event or a fact is causative if and only if it is a
member of a set of conditions that are minimally sufficient
for some result, has influenced not only philosophers of law
but also moral philosophers and metaphysicians and philosophers of action (A set of facts or events is minimally
sufficient for some result if and only if it is sufficient for
that result and it contains no redundant elements in
other words, no elements that are unnecessary for its sufficiency). Hart and Honor went on to adduce some considerations that sway the decisions of courts in singling out certain factors as legally decisive causes. Their key point was
that the factor singled out as the legally responsible cause
in any particular case is something that has disrupted the
normally foreseeable course of events. Such a disruption
occurs if someone has acted voluntarily with the intention
of bringing about a detrimental consequence that is at issue, or if some action or event has taken place in a context
19
MATT KRAMER
where the action or event is abnormal. Unlike the criterion

for causal efficacy recounted above, these considerations
for deeming a cause to be legally decisive are overtly
evaluative. Hart and Honor made clear in their exposition
that each such consideration is indeed broadly ethical. In
their view, legal responsibility within the jurisdictions
which they studied is always implicitly or explicitly a
matter of political morality.
REFERENCES
HART, H. L. A. 1961, The Concept of Law, Oxford, Clarendon
Press.
1963, Law, Liberty, and Morality, Oxford, Oxford University Press.
1968, Punishment and Responsibility, Oxford, Clarendon Press.
1982, Essays on Bentham, Oxford, Clarendon Press.
1983, Essays in Jurisprudence and Philosophy,
Oxford, Clarendon Press.
and HONOR, A. M. 1958, Causation in the Law,
Oxford, Clarendon Press.
SUGGESTED READINGS
ADLER, Matthew and HIMMA, Kenneth (eds.) 2009, The Rule
of Recognition and the U.S. Constitution, Oxford,
Oxford University Press.
CANE, Peter (ed.) 2010, The Hart-Fuller Debate in the
Twenty-First Century, Oxford, Hart Publishing.
COLEMAN, Jules (ed.) 2001, Harts Postscript, Oxford, Oxford
University Press.
DWORKIN, Ronald 1978, Taking Rights Seriously, Cambridge,
MA, Harvard University Press.
20
H. L. A. HART
FINNIS, John 1980, Natural Law and Natural Rights, Oxford,

FULLER, Lon 1965, The Morality of Law, New Haven, CT, Yale
University Press.
GAVISON, Ruth (ed.) 1987, Issues in Contemporary Legal Philosophy: The Influence of H. L. A. Hart, Oxford, Oxford
University Press.
KRAMER, Matthew 1999, In Defense of Legal Positivism,
Oxford, Oxford University Press.
et al. (eds.) 2008, The Legacy of H. L. A. Hart, Oxford,
LACEY, Nicola 2004, A Life of H. L. A. Hart. Oxford, Oxford
University Press.
MACCORMICK, Neil 1981, H. L. A. Hart. Stanford, CA, Stanford University Press.
RAZ, Joseph 2009, The Authority of Law, 2nd. ed., Oxford,
SHAPIRO, Scott 2011, Legality, Cambridge, MA, Harvard University Press.
WALUCHOW, W. J. 1994, Inclusive Legal Positivism, Oxford,
21
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
REWRITING HARTS POSTSCRIPT: THOUGHTS

ON THE DEVELOPMENT OF LEGAL POSITIVISM
Tom CAMPBELL*
Resumen:
Este artculo propone una lectura de la primera edicin de El concepto de
derecho (1961), de H. L. A. Hart, en la cual se recuperan los elementos
de prescripcin moral que acompaan a la tesis descriptiva/explicativa
que Hart explcitamente defendi. De tal modo, la tesis funcionalista del
surgimiento de normas secundarias en sociedades complejas depende de
la importancia de los beneficios sociales, relacionados con la justicia y la
eficiencia, los cuales Hart suscribe. Despus se llama la atencin sobre
la forma en que el Post Scriptum a la segunda edicin de El concepto de
derecho (1994), publicado pstumamente, podra ser reescrito para tomar en cuenta respuestas a Dworkin distintas a las que ofreci el propio
Hart en su Post Scriptum. La alternativa de una forma tentativa de un positivismo jurdico prescriptivo duro conservara de mucho mejor manera los ingredientes prescriptivos que son tan evidentes en la 1a edicin, y
contribuira a la regeneracin de una filosofa del derecho analtica contempornea dentro de la tradicin hartiana.
Palabras clave:
Positivismo jurdico, jurisprudencia analtica, moral, positivismo tico, H. L. A. Hart.
* Centre for Applied Philosophy and Public Ethics, Charles Sturt University,
Australia.
23
TOM CAMPBELL
Abstract:
The article suggests a reading of the 1st edition of H. L. A. Hart, The Concept of Law (1961) which involves bringing to the fore the elements of
moral prescription which accompany the descriptive/explanatory intent
which Hart himself explicitly espouses. Thus, the functionalist account of
the emergence of secondary rules in complex societies draws on the importance of social benefits, relating to justice and efficiency, benefits which
Hart endorses. Consideration is then given to how the posthumous Postscript in the 2nd edition of The Concept of Law (1994) might be rewritten to
take into account responses to Dworkin other than those adopted by Hart
in the Postscript. The alternative of a tentative form of prescriptive hard
legal positivism would better preserve the prescriptive ingredients evident
in the 1st edition and assist in the regeneration of contemporary analytic legal philosophy within the Hartian tradition.
Keywords:
Legal Positivism, Analytical Jurisprudence, Morality, Ethical
Positivism, H. L. A. Hart.
24
REWRITING HARTS POSTSCRIPT
SUMMARY: I. Introduction. II. Re-reading The Concept

of Law. III. Rewriting the Postscript. IV. Conclusion.
I. INTRODUCTION
Legal Positivism comes in many forms: conceptual, descriptive, explanatory and normative (that is, morally evaluative
or prescriptive). Legal Positivism is now generally thought of
as being an descriptive/explanatory theory but there is a
prescriptive or ethical form of legal positivism according to
which one element of a desirable system of law is that it
consists of relatively precise rules that can be identified and
understood without recourse the moral or other controversial opinions of those whose duty it is to follow or apply the
law. In slightly more technical terms, this may be called
prescriptive exclusive positivism according to which a legal
system ought to combine a rule of recognition, that can be
applied without having to make disputable moral judgments, with a set of rules that apply directly to the conduct
of legal subjects that are expressed in similarly amoral
terms.1
H. L. A. Hart, on the other hand, at least in The Concept
of Law,2 is avowedly conceptual and explanatory in a sociological vein, albeit with a hermeneutical bent that seeks to
explain social institutions, such as law, through identifying
the meaningfulness of social relationships for those involved, often using the conceptual methods of ordinary language philosophy according to which attention to linguistic
usages as they arise in working contexts is seen as a fruitful way of uncovering explanatory social meanings and dis-
1 Campbell, Tom, Prescriptive Legal Positivism: Law, Rights and Democracy,

London, Cavendish Publishing, 2004.
2 Hart, H. L. A., The Concept of Law, Oxford, Clarendon Press, 1961, second
edition 1994. Page references in this article relate to the 2nd. edition.
25
TOM CAMPBELL
solving unproductive conceptual puzzles.3 Thus Hart believes that the nature of social and legal obligation can be
illuminated by attending to the difference between our use
of obliged and obligation, which points us in the direction of explaining the authority of law through the interaction of different types of social rules, such rules being
explained in part as generally observed patterns of
conduct accompanied by what may be called a supportive
internal attitude on the part of members of the society or
group in question. This is the renown model of rules that
is intended to express the conceptual cum sociological essence of all developed legal systems as involving the union
of primary rules, which apply directly to conduct of law
subjects and secondary rules, which identify, change and
direct the enforcement of first order rules.4
This model is presented as descriptive and explanatory
rather than evaluative or prescriptive. But there is, methodologically, more to Hart that this, and I am not here referring just to his general corpus much of which is clearly
morally evaluative, but to The Concept of Law itself. There is
much in the actual argumentation and presentation of that
book which can be read as involving the moral views of the
author.5 Indeed it is not difficult to discern elements of
strong normativity in Harts method, elements that contribute to the attraction and persuasiveness of his theory of
law. These normative elements are largely covert and never
pervasive, but may be perceived, nevertheless, as being operative at crucial stages of his analysis of the concept of
law.
3 Hart, H. L. A., Essays in Jurisprudence and Philosophy, Oxford, Clarendon,

1983, p. 346.
4 In fact the label comes from Dworkin, Ronald M., The Model of Rules, University of Chicago Law Review , vol. 35, 1967, pp. 14-46. Hart himself does not confine his theory of legal norms to rules alone.
5 See McCormick, Neil, H. L. A. Hart, London, Edward Arnold, 1981, p. 25: Indeed, as Hart frankly acknowledges at the end of his book [The Concept of Law,
chapter 9, pp. 206-7] the ultimate basis for adhering to the positivist thesis of the
conceptual differentiation of law and morals is itself a moral reason.
26
Such a reading of The Concept of Law comes up against

the difficulty that it is not Harts intention or self-perception to be the constructing a morally prescriptive theory of
law. Famously, he sets out in the book to engage in what he
calls descriptive sociology, albeit through conceptual analysis of legal discourse.6 And in the 1994 Postscript, included in the posthumous 2nd edition of that book, it is emphatically re-affirmed that it is the aim of The Concept of
Law to provide a theory of what law is that is both general
and descriptive.7 My account he writes is descriptive in
that it is morally neutral and has no justificatory aims: it
does not seek to justify or commend on moral or other
grounds the forms and structures which appear in my general account of law, though, he adds, a clear understanding of these is, I think, an important preliminary to any
useful moral criticism of law.8 Ignoring the caveat for the
moment, such explicit statements of denial should surely
put an end to any speculation about the proper interpretation of that work.
However, authors are not the sole authority on the meaning of their utterances, especially when these are released
to the public forum in published form. Certainly we owe it
to authors to make a sincere effort to understand what they
are seeking to say, but in examining what they have published we are entitled to read it in the light of its public
meaning and the way in which the work is likely to be understood by its readers. Bearing this in mind, it is useful
and illuminating to explore the contention that there is
much in what Hart wrote in The Concept of Law that transcends his denial of prescriptive intent. Indeed, I suggest
6 The Concept of Law, vi: Notwithstanding its concern with analysis the book
may also be regarded as an essay in descriptive sociology; for the suggestion that
inquiries into the meaning of words merely throw light on words is false. Many important distinctions, which are not immediately obvious, between types of social
situation or relationships may best be brought to light by an examination of the
standard uses of relevant expressions and of the way in which these depend on social context, itself often left unstated.
7 Ibidem, p. 239.
8 Ibidem, p. 240.
27
TOM CAMPBELL
that this reading of Hart, explains some of the appeal of

Harts theory that arises from a reasonable understanding
of that work as a significantly prescriptive social and legal
theory to the effect that, defeasibly, an organised system of
clear mandatory rules is generally socially and morally beneficial and ought to be encouraged. Of course, the normativity in question assumes the feasibility of the model and
rests on an understanding of how legal systems are and
may possibly become, but this does not mean that the factual content of a theory is not sometimes to be seen as part
of an evaluatively directed theory.
This prescriptive reading of The Concept of Law would be
in line with the tenor of most of Harts other work. Indeed,
Hart constantly adopts prescriptive lines of argument in
most of his writings. He is firmly committed to arguing in
favour of controversial positions in legal and political philosophy, such as the need to retain mens rea as a basis for
excuses in the criminal law,9 and the rejection of paternalistic morality as a basis for the content of criminal law.10 As
a reasonably faithful successor to the Benthamite positivism which upholds the distinction between law as it is and
law as it ought to be, Harts prescriptive activities as an applied ethical philosopher are perfectly legitimate for a positivist as long as they are keep clearly and distinctly separate from evaluatively neutral conceptual and descriptive
analysis. Hart does not have to repudiate the evaluative nature of much of his other work in order to claim that The
Concept of Law is a morally neutral exercise in sociologically oriented conceptual analysis. However, no such bright
line between the prescriptive and the descriptive contributions in Harts work overall. Throughout his many scholarly
undertakings there is certainly a constant endeavour to
achieve conceptual clarification (and much of this is directly related to the search for a general descriptive theory
of law) but his conceptual analysis is frequently closely to
9
10
Punishment and Responsibility, Oxford, Clarendon Press, 1968.

Law, Liberty and Morality, London, Oxford University Press, 1963.
28
repeated efforts to undermine poor evaluative arguments

for unpalatable evaluative conclusions and a willingness to
take the initiative in providing a framework for progressive
liberal thought.
Even if the main thrust of The Concept of Law is providing an understanding of law that is relatively neutral as between different types of legal system, good and bad, it
would be surprising, given Harts admiration for Bentham,
if this were not closely allied to an assumed normative outlook poised to take the field in its own right. Like
Benthams, Harts approach is broadly utilitarian, and is
crucially shaped by a belief in the social utility of rules as
well as their potential as instruments in the causation of
grave harm. This intermingling of prescription and description is not in itself a problem. The two logically distinct
modes must be systematically related in applied philosophy, although, as Hart insisted, there must always be vigilance if we are to avoid illicit moves from is to ought
the much discussed naturalistic fallacy or, as I surmise is more prevalent in Harts work, illicit moves from
ought to is the less noticed normative fallacy in
which beliefs about what ought to be the case improperly
impact on assertions about what is the case.11
In the first edition of The Concept of Law itself, Hart is
quite relaxed about introducing normative considerations to
explain and defend his theory. It is only in the Postscript,
included in the 2nd edition published posthumously in
1994, that he is definitive in rejecting an evaluative interpretation of the book. Perhaps, he was pressured into this
position by Dworkins critical commentary and adopted a
more extreme view than the more eclectic approach of the
first edition. For whatever reason, in the Postscript, which is
based on his unpublished notes, Hart openly rejects any
reading his account of the union of primary and secondary
rules as being in any way a recommendation of legality or
11 Campbell, T. D., The Normative Fallacy, The Philosophical Quarterly, vol.
20, 1970, pp. 368-77.
29
TOM CAMPBELL
the outline of a morally desirable system of law rather than

a morally neutral account of systems of law in general.
Moreover, there he endorses unequivocally soft (or inclusive) positivism, that is the thesis that rules of recognition
can and often do contain moral criteria for the identification
of valid law, and he makes this acknowledgement without
expressing any regret that this is the case, or even any concern that it might undermine his explanatory theory.
Most commentaries on the Postscript gloss over Harts
option of adopting a more prescriptive stance related to the
good consequences that flow, for both social cooperation
and justice, from the instantiation of the governance of
rules. Thus in the collection of essays edited by Jules
Coleman,12 with the notable exception of Jeremy Waldron,
little attention is paid to this possibility. Waldron alone
considers seriously whether Hart may be read as a normative (or ethical) positivist.13 As he explains these terms in
another essay, Waldron notes We need to distinguish,
however, between: first, the value judgments that might be
required in a nonpositivist jurisprudence to identify some
propositions as a valid legal norm; and second, the value
judgments that support the positivist position that evaluations of the former type should not be necessary.14 Perhaps, if Hart been clearer in his own mind about this distinction, and less concerned to protect his original
methodological classification of The Concept of Law, he
would have seen that the arguments he put forward are often in substance evaluative arguments for having, in a
complex society, a rule of recognition that can be applied
12 Jules Coleman, ed., Harts Postscript: Essays on the Postscript to The Concept
of Law (Oxford, Oxford University Press 2001).
13 Although Stephen Perry comes close in discussing Hart as a methodological
positivist (Coleman, op. cit, pp. 311-354) and Liam Murphy discusses the normative reading of Hart, and concedes that Hart does draw on moral grounds to justify
some aspects of his concept of law (Coleman, op. cit., pp. 371-409).
14 Waldron, Jeremy, Kants Legal Positivism, Harvard Law Review, vol. 109,
1996, pp. 1535-66 at 1541.
30
without recourse to evaluative judgments and first order

rules that are similarly operatively morally neutral.
To argue that Hart is a covert prescriptive legal positivist
is not simply to re-read the Postscript, it is to re-write it.
Why should we do this? One excusing reason is that Hart
did not himself publish the notes that have been put together in the Postscript despite many opportunities to do
so. Perhaps he had doubts about the content of the material he had prepared and left unfinished and unpublished
at his death. However, there is no indication that he would
have given any credence to such a rewrite, although he may
have been tempted to make some retractions. He may have
thought that to do so would be to concede too much to his
critics and undermine the distinctiveness and integrity of
his theory which had by that time reached canonical status.15
There is therefore some basis for re-reading The Concept
of Law from a normative perspective. In doing so, it may
bring forward a more convincing basis for many of the
books central theses, which sometimes lack the empirical
bases to lay claim to be an example of descriptive sociology,
and, at the same time, it come to serve more as a source of
inspiration for prescriptive legal philosophies of a positivist
hue which can present a more critical approach to current
legal practice that other contemporary versions of analytical
legal philosophy.
II. RE-READING THE CONCEPT OF LAW
Re-reading The Concept of Law from a morally prescriptive point of view, deploying a certain amount of the hindsight of the later developments in legal theory is an interesting exercise.
The strongest and most evident grounds on which it can
be argued that The Concept of Law is open to an evaluative
15 Lacey, Nicola, A Life of H. L. A. Hart: the Nightmare and the Noble Dream, Oxford, Oxford University Press, 2004, chapter 14, especially pp. 348-55.
31
TOM CAMPBELL
reading lies in his straightforward account of the transition

from a simple society, that is a society with nothing but
first order rules by way of normative ordering, to a complex
society with a combination of primary and secondary rules,
with the secondary rules (that is rules about rules) having
to do with the ways in which primary rules are to be officially identified, interpreted, changed and enforced. In
chapters 5 and 6 of The Concept of Law, the main thrust of
his explanation for the emergence of the ingredients of a developed legal system derives from the social benefits that
follow from the emergence of a legal system in an increasingly complex and changing society. We are taken through
the image of a simple society with customary rules only,
rules that are stable through time, widely held and respected, and enforced in a unorganised way through informal sanctions, to the official organisation of a system of
rules through the activities of specialist officials who make,
mould and apply the rules they officially identify and apply
on behalf of the rest of the relevant social group. And we
are given the clear indication that theses changes actually
do provide the certainty, efficiency and flexibility required
in more complex social environments.
To develop this interpretation, and taking a cue from
Harts commitment to the ordinary language philosophy of
the time, we may attend to the language he uses in presenting this aetiological tale, which might otherwise be
regarded as no more than a value free functionalist explanation. The story unfolds through the progressive identification of defects in a system of customary rules that
turn out, in an increasingly complex and changing society,
to be too static, too ill-defined, too informal and too disorganised to meet the needs of more developed social groups.
Basically, the thesis is that the defects of having only primary rules are overcome through the introduction of the
secondary rules, the rules of the officials whose job it is to
identify, change and apply primary rules, and in particular
through the rule of recognition that sets out the criteria
32
that must be satisfied for a primary rule to be considered

as part of law rather than simply an informal custom. The
defect of everyone enforcing the rules for themselves which
is highly chaotic, and therefore both inefficient and unfair,
is dealt with by setting up a central authority whose task it
is to decide when the rules have been broken and what
should be done about it. The defect of static rules is overcome by providing some mechanisms for change so that the
society as an organised whole can respond effectively to
novel circumstances. So the model of rules is clearly better
than that which preceded it on a number of counts, at least
for the type of societies in which the defects of the simpler
system are manifest. Hart then goes on to talk of the The
pathologies associated with those legal systems that have
not quite made it as legal means of social control, with further talk of failures and breakdowns of such legal systems.16
Further on, where he writes of the protections and benefits provided by the system of mutual forbearances which
underlies both law and morals,17 Hart notes that the step
from the simple form of society, where primary rules of obligation are the only means of social control, into the legal
world with its centrally organized legislature, courts and officials and sanctions brings its solid gains at a certain cost.
The gains are those of adaptability to change, certainty,
and efficiency, and these are immense; the cost is the risk
that the centrally organized power may well be used for the
oppression of numbers with whose support it can dispense,
in a way that the simpler regime of primary rules could
16 The Concept of Law, p.118. Stephen Perry notes that while it is possible to
discern elements of the descriptive-explanatory method in Harts approach to doing legal theory, there are good reasons for believing that he does not in fact employ
that method, or at least that he does not employ it in anything like a pure form. The
most important reasons is that Hart adopts the characterization of law that he
does, expressed in terms of the union of two types of rule, on the basis of evaluative
judgments that have nothing to do with the metatheoretical criteria for assessing
theories. Harts Methodological Positivism in Coleman, Jules, ed., Harts Postscript, Oxford: Oxford University Press, 2001, p. 315.
17 Ibidem, p. 200.
33
TOM CAMPBELL
not.18 While such evaluations are not unequivocal, they

are unequivocally evaluations.
All this can be read as an explanation of the emergence of
legal systems on the model he describes. The assumption
might be that these are intentionally introduced changes
that are designed to and actually do make social outcomes
better than they were before. We can thus understand why
they were introduced and are maintained. This can be read
as an explanation not an evaluation. And where such accounts of social change are couched in more functionalist
terms without reference to the conscious intentions of the
participants they can be rendered explanatory by the assumption of some sort of social natural selection with
changing social equilibria and ongoing historical selection.
Yet, such explanations, whether or not they draw on conscious intentionality, can also be read as prima facie justifications of such change. Of course people may be wrong
about the benefits to be derived from the changes, and we
may challenge their belief that the alleged benefits are indeed benefits, or that they justify the alleged costs. There is
no deductive move here from is to ought, no naturalistic
fallacy. But apparently functionalist explanations are capable of being read and assessed as value judgments which, if
sound, would justify maintaining and where appropriate
changing the systems that have emerged as a result of their
perceived or actual advantages. When this is done, there
may well be a tendency to read back from the justifications
to an idealised picture of the reality the normative fallacy
at work thereby diverting attention from the need for empirical evidence to support even the hermeneutical form of
social explanations.
The symbiosis here between the realm of meaning explanations and that of normative justification is clear. Thus in
his account of the existence and functioning of social rules,
Hart emphasises the internal point of view whereby those
involved in the societies in question have certain beliefs and
18
Ibidem, p. 202.
34
attitudes towards their rules and adherence to them, including certain beliefs about the social utility and importance of such rules. Such beliefs not only explain why the
rules persist but provided a starting point for a discussion
of whether they ought to exist. We can accept that the beliefs and their justification are distinct without undermining the relevance of their justificatory use in deciding
whether legal systems are a good idea and in deciding what
sort of legal systems ought to be preferred.
As Hart himself makes clear in the Postscript it is, of
course, possible to describe and explain the moral judgments of others without endorsing them.19
However, the historical story is reinforced by Hart in
terms of what he sees as the uncontroversial values which
he endorses as a form of minimal natural law. Here Hart
draws on the moral convictions which most of us share20
including the protection of life, bodily security and property
rights that are required to sustain living conditions without
which a tolerable life, even life itself, could scarcely continue. The primary rules themselves that are to be found in
every society are taken to be uncontroversially justifiable if
not in every detail at least in their general thrust. At this
level sociological imperatives are not in tension with moral
ones. This broad awareness of the general aims and advan-
19 Ibidem, p. 243 (in the Postscript): Even ifthe participants internal perspective manifested in the acceptance of law as providing guides to conduct and
standards of criticism necessarily also included in a belief that there are moral reasons for conforming to laws requirements and moral justification of its use of coercion, this would also be something for a morally neutral descriptive jurisprudence
to record but not to endorse or share.
20 Hart, H. L. A., Punishment and Responsibility, Oxford, Oxford University
Press, p. 88. His methodology is set out on page 10. In relation to any social institution 1) state the general aim or value it fosters and 2) enquire after principles limiting the pursuit of that aim or value. Thus the purpose of criminal law is to prohibit
certain types of conductthe primary task of securing society from evilbut there
are limitations as to how this objective may be prosecuted. In the same book, he
notes, on p. 22, the importance in criminal law of fair opportunity and the power
to identify beforehand the particular periods when he will be free of them [sanctions].
35
TOM CAMPBELL
tages of social institutions is a recurring theme of Harts

work.
The endorsement of a minimal morality is clearly distinct
from what Hart calls critical morality which functions to
comment and commend with respect to existing social
norms, but there is no indication that the sort of critical
morality that Hart endorses involves the repudiation of the
basic values identified in his explanatory/normative model.
So what we have are the ingredients of a justification for
developing and assessing legal systems, including the values that are relevant to assessing the advantages and disadvantages of having such systems at all. It is important to
note here that the scarcely concealed generally positive
evaluation of secondary rules is parasitic on the social
value of having primary rules. The attractions of law are
built on the attractions of rule-governance in general. Thus,
Harts picture of the social utility of rules in protecting from
harm and coordinating human conduct is not just a warm
up for his legal theory but an integral part of that theory.
The importance of the secondary rules is unintelligible
without the prior analysis of the value of primary rules. Primary rules are therefore as essential to Harts cryptoevaluative account of law as are secondary ones. This is
not, of course, a claim that rules are beneficial whatever
their content, but that rules are an indispensable technique
for securing certain types of social benefits. Only together
do they provide the basis for a convincing theory of legal
obligation.
Other evidence of Harts evaluations and prescriptive
goals emerge in his discussion of justice in The Concept of
Law, particularly formal justice, where draws attention to
the affinity of law and the justice involved in treating like
cases alike. Here again Hart is careful with his words. He
does not carelessly use justice for the sum of all or all important moral values, but relates it to the discourse of fairness and hence relates it to maintaining or restoring a balance or proportion in the distribution of benefits and
36
burdens, for fairness he notes (with an eye to the context of

ordinary language usage) is an idea closely associated with
shares.21 Moreover, it is clear that he takes it for granted
that such a balance or proportion often requires the use of
rules and their impartial administration, so that The connection between this aspect of justice and the very notion of
proceeding by rule is obviously very close.22 So, while he
emphasises that, however fairly administered, a law may be
unjust in its content, perhaps because it takes a morally irrelevant factor into account, he accepts the close association between having rules, which are administering impartially, and the very idea of justice.
It may be argued that by providing evaluative grounds for
having rules Hart opens the way for critical morality to condemn some first order rules but that this critical morality
does not affect his sociological explanation of legal systems
or have implications for the choice of one style of legal system over another. This would mean that we could accept
that Hart endorsed the role of value judgments in criticising
particular legal rules but not in asserting the value of law
as such. For he does not seem to tinker with the content of
his model in order to commend a type of legal system that
best fulfils its justificatory functions.
This certainly appears to be the position presented in the
Postscript when it comes to the choice between hard and
soft positivism, that is between excluding or including
moral criteria in the rule of recognition. Given, as Dworkin
had pointed out, there are legal systems that have moral
principles in their rule of recognition and given that Hart is
engaged in a descriptive exercise, it seem inevitable that
Hart should agree that rules of recognition may contain
moral criteria, hence his soft legal positivism. And this certainly fits with The Concept of Law in so far as he allows a
role for morality in the exercise of judicial discretion in the
interpretation of open-textured rules.
21
22
Ibidem, pp. 158-59.

Ibidem, p. 161.
37
TOM CAMPBELL
Yet, Hart is not required to stop there and in fact he does

not do so. For instance, he effectively raises the question of
what constitutes a good or effective rule of recognition. The
answer is, in part, a rule that fulfils its role of settling disputes about the content of law. This comes up nicely in his
analysis of legal realism and the role of discretion in regulation. He points out that talk of discretion arises in situations where there are rules to be applied but there is some
leeway as to what they mean and how they should be applied. It is not simply that it makes no sense to speak of
discretion when the scorer can decide what she likes, effectively making up the rules as she goes along (what Hart
calls scorers discretion), but that such a system, taken to
extremes, makes the game unplayable. In this discussion
he happily speaks in terms of the advantages and disadvantages of having rules, of having a degree of discretion
and of having authoritative determinations of what the
rules are and whether they have been broken.23 This enables us to reject a rule of recognition that says what whatever the judge decides is the law is the law. This is the nub
of Harts critique of legal realism. He accepts that a rule of
recognition that does not fulfil its function adequately is either a poor rule or not a rule at all. While within his descriptive approach he does not want to hold that all rules of
recognition are effective beyond the point where a legal system may be said to exist, but he has to hand the values for
making judgments about what is a good or bad rule of recognition and these are the values that are used in his account of why rules of recognition are adopted and sustained.24
In this regard it is interesting to note that the rule of recognition is a social rule. although not in fact a legal one. It
Ibidem, p. 139.
Shapiro, Scott, J.,On Harts Way Out in Coleman, op.cit., pp 149-92 at
177-78 Inclusive rules of recognition do not tell judges which moral rules they
should apply they simply tell judges to apply moral rules. These rules cannot give
epistemic guidance because judges are left to figure out for themselves what those
rules are. Vis--vis such rules, they are like ordinary citizens.
23
24
38
follows that there is an internal attitude attaches to it.25 Assuming that we can accept that a rule of recognition established obligations for the officials, we can surmise that
when those officials criticise departures from that rule they
do so on the basis of their belief in the importance of the
rule, and that this belief in the importance of having a rule
is grounded in an awareness of the value of having some
sort of rule of recognition, provided it fulfils its function.
This gives further support to the idea that the theory assumes and contains the ingredients of a normative legal
positivism that commends having a rule of recognition and
by implication commends rules of recognition that satisfy
certain functional criteria relating to the effective selection
of rules for inclusion in the legal system.
Consider also Harts assumption that each legal system
has one rule of recognition.26 How does he come to that
view? It is feasible, indeed empirically established, that
judges in actual legal systems operate with a range of rules
of recognition, some giving more significance to international law, others to strict statutory construction, others to
the moral principles perceived to underlie the common law.
Yet Hart claim that there is one rule of recognition that is
the unifying feature, indeed the sovereign of each distinct
legal system.
His Postscript response to this is that each legal system
that there is general agreement on its rule of recognition.
He thinks of this as a complex empirical fact.27 Yet he did
not conduct such empirical research and his confidence in
this matter may well be due more to his sense that the effective and just functioning of a legal system depends on
there being consistency between judges with respect to
25 Ibidem, p. 146: To say that at a given time there is a rule requiring judges to
accept as law Acts of Parliament or Acts of Congress entails, first, that there is a
general compliance with this requirement and that deviation or repudiation on the
part of individual judges is rare; secondly, that when or if it occurs it is or would be
the subject of serious criticism and as wrong.
26 Ibidem, pp. 100-12, 246.
27 Ibidem (in the Postscript), p. 292.
39
TOM CAMPBELL
rules of recognition. Judges and others have an obligation

to strive for agreement on what that rule is to be. Only this
can produce the coherence, clarity and decisiveness required by a legal system that is going to decide disputes, reduce harmful conduct and facilitate cooperative activities.
Further evidence of the underlying prescriptive nature of
some of Harts arguments derives from the Hart/Fuller debate as to the practical consequences of judges in an evil legal system adopting legal positivism as a practical philosophy.28 The well known disagreement is whether judges
faced with grossly immoral laws should or should not regard them as laws at all. Fuller takes the view that if judges
are sufficiently attached to a form of natural law that enables them to see that such laws are not properly regarded
as laws and should not therefore be enforced they are less
likely to adjudicate in blind obedience to evil rulers. Hart, to
the contrary, argues that positivist judges, attuned to the
distinction between law as it is and laws as it ought to be,
are more ready to accept that current laws are immoral and
therefore to take a moral stand against them.
Whoever wins this particular debate and it would appear to be matters of fact that mainly divide the protagonists what is going on is a prescriptive argument for or
against the legal theories involved based on their alleged
consequences. Here we appear to have Hart engaged in prescriptive positivism, commending his legal theory on moral
grounds.29 This may be too strong. After all, Hart would insist that he is only saying that legal positivism does not
have the unfortunate consequences it is alleged to have, in
that it does not have the result of increasing adherence to
evil laws. These considerations may not, however, affect his
28 Hart, H. L. A., Positivism and the Separation of Law and Morals, Harvard
Law Review, Vol. 71, 1958, p.593-629; Fuller, Lon, Positivism and Fidelity to Law:
A Reply to Professor Hart, Harvard Law Review , Vol. 71, 195, p. 630.
29 As Fuller suggests, ibidem, at p. 673: Professor Harts essay seems to me to
open the way for such a discussion for it eliminates from the positivist philosophy a
pretense that has hitherto obscured every issue touched by it. I mean, of course,
then pretense of the ethical neutrality of positivism.
40
decision to adopt the theory. Nevertheless, it is difficult to

believe that they do add some weight to his position, and go
some way towards explaining Harts tenacity in defending
the positivist thesis.30
The protagonists in this debate do not call in the empirical evidence that might resolve most of their differences. Instead, they seem to rely more on the logical implications of
the theories in question. Fuller sees legal positivism as implying that judges must apply rules which they do not approve of, and Hart says that natural law requires us to accept that existing law is morally justified. Fullers case does
not seem intuitively compelling unless we assume that
rules of recognition do not include moral criteria, that is,
unless we assume hard legal positivism, and Harts criticism must be directed inter alia against having moral criteria in the rule of recognition, for that is what encourages us
to believe that positive laws have passed moral tests. What
is involved here is effectively a debate about what ought to
be the form and content of the rule of recognition, that is
between hard and soft legal positivism. Hart is defending
both the distinction between law as it is and law as it ought
to be and a legal theory that requires judges and others to
adopt a rule of recognition that is exclusive of moral considerations, and acquiesces in Fullers assumption that this is
precisely what legal positivism entails. It is not simply the
holding of abstract views on the distinction between ought
and is that gives rise to different judges reactions to evil regimes, but views about what sort of rules of recognition
judges ought to follow. This is, therefore, far more than a
marginal dispute about the moral side effects of legal posi-
30 Murphy, Liam, The Political Question of the Concept of Law, in Jules

Coleman (ed.), Harts Postscript, Oxford, Oxford University Press 2001, pp.
371-409 accepts this quietism argument in relation to the population in general,
but only to a limited extent, and denies that it makes Harts general theory prescriptive. See also Liam Murphy, Concepts of Law, Australian Journal of Legal Philosophy, Vol. 30, 2005, pp. 1-19, Dyzenhaus, David, Hard Cases in Wicked Legal
Systems, Oxford, Clarendon, 1991.
41
TOM CAMPBELL
tivism. Rather, it is an issue at the core of the debate about

what constitutes a desirable system of law.
Harts objection to Fullers is not about his general theory
of the internal morality of law in so far as it points to the
benefits of systems of rules that are clear, prospective, general, practicable and stable. His argument is, that such legal systems are not necessarily just or beneficial. Fuller,
Hart thinks, overstates the moral significance of formal fairness, the beneficially efficacious outcome of formally good
laws, and the role of purpose in adjudication. In his words:
The difference between the author [Fuller] and those he
criticises on this matter is the activity of controlling man by
rules and the principles designed to maximize its efficiency
are not valued by the latter or their own sake and are not
dignified by the title of a morality. They are valued so far
as they contribute to human happiness and other substantive moral aims of the law.31
It is not only the phenomenon of evil legal regimes that
flushes out the prescriptive thrust of some of Harts analyses. Another debate, between Hart and Patrick Devlin, also
brings to the surface moral arguments for the adoption of
legal theories with specific implications for what constitutes
a good legal system.32 In that debate Devlin had argued for
the enforcement of morality in the criminal law because a
society has to have one cohesive and consistent pattern of
right and wrong in relation to such matters as sexual conduct and the associated institutions of marriage and divorce. Hart wishes to defend a liberal line on issues such as
the criminalisation of homosexual conduct along the lines
of J. S. Mills harm principle according to which conduct
may not be criminalised if it does not cause harm to others.
Victimless crimes should not be crimes, therefore, even the
general opinion that something is immoral is not in itself a
31 Hart, H. L. A., Essays in Jurisprudence and Philosophy, Oxford, Clarendon
Press, 1983, p.357.
32 Hart, H. L. A., Law, Liberty and Morality, London, Oxford University Press,
1963.
42
reason for prohibiting it in law. To establish this position it

is important for him to show that there is logical space between morality and law. Hart position here may be compared to Neil MacCormicks moral disestablishmentarianism, the principle that nothing should be made law simply
because it is morally approved or disapproved, even (or perhaps especially) by the (moral) majority, a position MacCormick presents as a moral case for amoral law.33
This may look like a straightforward disagreement about
the desirable content of first order rules. However, as many
positivists have pointed out, the de facto distancing of morals and laws cannot be done without the capacity to distinguish the domain of law from the domain of morality. If
moral criteria are introduced to the rule of recognition then
this capacity to restrict law to the prevention of harm and
the promotion of prosperity, for instance, will be in vain to
the extent that such moral criteria play a role in the identification of law. And so, at the very least we can see that
there is a congruity between normative legal positivism and
Harts strong liberal position that he adopted in the
Hart/Devlin debate, which allowed him to argue that moral
approval of heterosexuality is not in itself a reason for
criminalising homosexuality. To operationalise such a system it is necessary to omit from the rule of recognising the
criminal law any criterion along the lines of protecting morals.34
33 Neil MacCormick, A Moralistic Case for Amoralistic Law, 20 Valparaiso University Law Review, 1985, 1-42.
34 It may be argued that Hart himself introduces a moral criterion into the rule
of recognition in the form of the Millean harm principle to the effect that the criminal law may be used to prohibit and deter individuals harming other people. But
there is no evidence that Hart saw the harm principle as a desirable candidate for a
rule of recognition. Remember, the principle says that the law may be used to prevent harm in these sense that it may not be used to prevent anything that is not
harmful to others. It does not say that causing harm to others is a sufficient basis
for rendering a type of conduct unlawful. In fact it is clear that Hart assumed that
the question of whether or not harmful behaviour warrants criminal obligations is
a matter for debate and decision within the political and ultimately the legislative
process.
43
TOM CAMPBELL
This accumulation of evidence demonstrates that it is not

only possible but natural and illuminating to read The Concept of Law as in part an exercise in normative legal philosophy which points to the social value of primary rules and
thence of legal systems constituted through a shared rule
of recognition, and does so in a way which endorses the potential utility of law, and provides reasons that can feature
in the determination of what makes for a good legal system,
in terms of its formal features as well as its substantive
content.
III. REWRITING THE POSTSCRIPT
The Postscript cannot readily be interpreted even as covert prescriptive legal positivism. There Hart explicitly and
sometimes stridently affirms his descriptive/explanatory intent, moves away from an emphasis on the significance of
law as a system of primary rules (as distinct from principles) and appears to repudiate all prescriptive exercises beyond defending the importance of taking a critical attitude
to law.
Here we cannot re-read, but we can re-write. In terms of
the development of legal positivism, the moves Hart made
in response to Dworkins critique of the model of rules may
be unfortunate in so far as they dismiss the prescriptive
and evaluative elements at work kin the legal positivist tradition. From this point of view it would have been better if
he had affirmed the centrality of primary rules within a legal system for reasons of formal justice and administrative
efficiency. He could have noted that admitting moral criteria into the rule of recognition weakens its social utility
for the resolution of disputes, the coordination of conduct
and the democratic authority of legislation. He could have
identified those points at which his theory is not only normative but in conflict with Dworkins ideal of a legal system
dominated by Herculean judges exercising their moral judgments as to what best justifies existing bodies of law and
44
developing that law in accordance with those principles.

And he could have done all this while still maintaining that
there are large elements of purely descriptive analysis in his
work and denying that, in making law the best that it can
be, the substantive moral view of judges must be decisive
in legal process.
The Dworkinian criticisms that Hart is answering in the
Postscript are (a) that law contains principles as well as
rules (b) that the rule of recognition falls foul of the semantic sting whereby it reduces legal disagreement to disagreement about the definition of law, which it clearly is not, and
(c) that legal positivism has to be seen as an interpretive
theory dealing with such issues as the justification of state
coercion. In the Postscript, Harts response to these criticisms is a mixture of capitulation and defiance. Capitulation with respect to the centrality of rules and defiance with
respect to claim that his descriptive theory does not come
into conflict with Dworkins overtly prescriptive approach.
The first and central concession was to declare his theory
compatible with Dworkins model of rules plus principles.
Hart firmly endorses soft or inclusive legal positivism by accepting that there are rules of recognition that contain principles, and that principle are norms that give moral reasons
of various weight for determining the law to be this or that.
This is, perhaps, an inevitable move if Hart is to maintain
his purely descriptive/explanatory commitment. He can
hardly deny the many ways in which it is accepted in many
systems that morality should enter law through the opinions of judges. That seems incontrovertible empirical fact.
Nor can he deny the logical space created for this possibility
by the concept of the rule of recognition, for once that rule
is in operation it can be used to identify as law whatever
satisfies its criteria, and this certainly need not involve simply choosing between putative legal rules.
Yet, Hart could have insisted that effective legal systems
are in large measure rule-based, both with respect to the
criteria in the rule of recognition and the fact that their pri45
TOM CAMPBELL
mary norms are social rules, with principles coming in

largely in complex interpretive issues and permitted judicially initiated rule-change. Frederick Schauer provides a
possible model here with his analysis of rule-based decision
making and its social benefits as something that is by and
large descriptively accurate as well as normatively desirable.35 Better still, Hart could have explained the potential
benefits of such systems and argued for developing systems
of law that enables us to identify the official rules of our
mandatory system of norms in ways that do not involve
moral judgments. That is, he could have given us grounds
for adopting hard or exclusive legal positivism in a system
of rules as an ideal with which to go into battle against
Dworkins scheme of moral reasoning in law.
These moves would have provided Hart with a much
more convincing explanation of legal obligation. His attempt
to distance legal obligation from mere coercion is weak on a
purely descriptive approach. Its attraction lies in adding the
social utility of social obligations to be added to the obligation that comes from the social utility of having a system
that serves the function of identifying and applying a specially important set of those rules. Legal obligation thus
draws on the moral force of primary as well as secondary
rules, for we have need of both. When all the weight of legal
obligation goes on to the secondary rules, with no requirement that these are choosing between primary rules for
which there are social justifications, then the analysis of legal obligation loses the force of the combination value of
primary and secondary rules. Hart would correctly insist
that such primary rules may in fact be evil and the value of
having secondary rules may be entirely negated by that
fact. Nevertheless the fact that rules of recognition are limited to the role of choosing between primary rules that have
potential utility by virtue simply of their being rules rather
than because of their particular content, gives a depth to
35 Schauer, Frederick, Playing by the Rules: A Philosophical Examination of
Rule-Based Decision-Making, Oxford, Clarendon, 1991.
46
the (limited) content that a positivist would want to give to

the notion of legal obligation. Law of a positivist kind gives
rise to stronger prima facie reasons for obeying that law
than would be the case were the rule of recognition be not
constrained by the need to select rules or by the use of
moral criteria in the rule of recognition. The positivist point
still remains that these derive from the moral reasons for
having a system of authoritative rules to govern social life,
moral reasons that may always be outweighed by the content of the first order norms and the ineffectiveness of the
second order norms.
As it is, Hart is left with less support for his earlier claim
that legal positivism provides a basis for developing a critical moral approach to law by institutionalising the separation of law and morality. Only a hard or exclusive positivism can fit his story of the benefits of law as a union of
primary and secondary rules, his emphasis on the affinity
of formal justice and law, his desire to prevent people imposing their morality on the private conduct of others and
the implication that we should be encouraged to adopt a
morally critical view of the content actual laws.
Adopting a laissez faire attitude to admitting moral criteria in rules of recognition greatly reduces the practical significance in his distinction between is and ought in law, as
can bde seen beyond Jules Colemans unadventurous concept of negative positivism: the thesis that there is at least
one possible legal system in which the rule of recognition
contains no moral criteria, as a denial that there is a necessary relationship between law and morality.36 What Hart
does descriptively is to emphasise the de facto overlap of
law and morals in language, function and content. What he
could have done normatively is to encourage a conceptual
distance between law and morals in which his liberal utilitarian aims can flourish.
36 Jules Coleman, Negative and Positive Positivism, Journal of Legal Studies,

Vol. 11, 1982, pp. 139-64.
47
TOM CAMPBELL
At another conceptual extreme, Hart could have argued

for hard or exclusive positivism as a conceptual thesis, an
objective to which Joseph Raz aspires, on the grounds that
only a system that enables us to identify laws without recourse to moral judgment can play a role in guiding conduct in situations where there is moral dispute.37 This
reads remarkably like a normative argument for the utility
of hard positivism, and indeed this is in places how Raz appears to present his sources thesis, that is the thesis that
law must be identified through some sort of social fact,
such as an enactment.38 It is highly unlikely that Hart
would have contemplated such an approach as he eschews
propagating conceptual necessities and would surely want
to say that actual legal systems are better and worse at
guiding conduct, yet it would certainly be closer to the gist
of his own theory than an open ended descriptive pluralism
that has no strong objection to admitting moral criteria into
the rule of recognition and the content of first order rules.
Why Hart did not make some such moves is more a question for biographers than philosophers.39 It is clear from the
content and rhetoric of some of the controversies in which
he was involved that he could take to defending his position
with as much zeal as any academic out to sustain his own
theory. It may be that he was so impressed by the significance of ordinary language philosophy and its repudiation
of first order moral theory as a respectable academic exercise that his methodological loyalties held him back. It may
be that he thought that once he had admitted a measure of
prescription into his theory he would have to abandon the
important and largely feasible objective of providing a general descriptive account of legal systems and accept
37 Raz, Joseph, Authority, Law and Morality, Monist, Vol. 68, 1985, p.195:
its [laws] existence and content can be identified by reference to social facts
alone, without resort to any evaluation.
38 Raz, Joseph, Between Authority and Interpretation, Oxford, Oxford University
Press, 2009, Part II.
39 See Lacey, op. cit., chapter 13.
48
Dworkins invitation to see his theory as essentially interpretive (ie prescriptive).40

Hart was surely wrong in holding that there is no significant conflict between enterprises so different as my own
and Dworkins conceptions of legal theory.41 He may not
have been as narrow as Dworkin in confining his attention
to the need for a theory of law to justify coercive state activity and its moral limits. But his theory does focus on mandatory social rules and he does produce arguments whose
logic is to point up the desirability of government coercion
being exercised through the medium of rules. Less controversially, it is clear that by allowing that his theory is open
to a more Dworkinian form of legal reasoning he deprives
himself of the argument that the second order rules he introduces to solve the defects of a simple normative system
will improve the conditions of uncertainty from which they
are meant to rescue simple societies as they become more
complex.
Interestingly, while the Postscript concedes little to
Dworkin, Hart does let his descriptive guard down when he
is drawn into an evaluative disagreement that brings to the
surface aspects of his underlying motivation. In countering
Dworkins disingenuous suggestion that he (Hart) take his
own theory to be interpretive, that is ultimately an evaluation of what law ought to be, Hart insists that certainty is
by no means the only legal value of importance to him. He
accepts that meeting expectations is a particular moral
merit which law has, not the whole purpose of law,42 and
comments that limited discretion is not unjust because in
interstitial areas there are no expectations to disappoint.43
He also notes the importance of identifying in advance use
of coercive measures and private powers.44 But he takes the
40
41
42
43
44
Dworkin, Ronald M., Lawss Empire, London, Fontana, 1986, p. 52.

The Concept of Law, p. 241.
Ibidem, p. 249.
Ibidem, p. 276.
Ibidem, p. 250.
49
TOM CAMPBELL
certainty to derive from what he calls plain fact positivism

as only part of a system of good legal reasoning: Exclusion
of all uncertainty is not a goal which I considered45 for
there are other aims which law should cherish.46 This refers back to his earlier preference for a limited measure of
judicial discretion in the interpretation and application of
rules where this is required by the open texture of language. In that discussion he clearly sees the value of such
open texture in permitting incremental legal development in
the light of judicial first hand knowledge of actual social
circumstances. Thus, orderly and sensible change is valued
by Hart as well as certainty, clarity and decisiveness.
These comments reveal Hart as something less than a covert hard legal positivist. Perhaps one might call him a firm,
rather than soft, positivist, or, in Schauerian terms, even a
presumptive prescriptive positivist.47 Certainly his apparent
willingness to adopt soft positivism may be seen as in part
a recognition of the value of a certain amount of mushy
law, as advantageous in enabling incremental legal change,
providing opportunity for judicial leadership, and for increasing consistency in the total body of law administered
by courts. So, it seems clear that as a prescriptive legal theorist Hart is not a thorough-going hard legal positivist, but
it does show that he is deeply concerned with prescriptive
legal theory, with what sort of legal system and what concept of law is best in our circumstances, how much judicial
discretion is a good thing, how specific we want our rules to
be, what sort of secondary rules we want to adopt for purposes of adjudication, and so on. This enables him to deny
that he is a single minded hard or inclusive legal positivist
but it does not answer the suggestion that his theory is, in
substantive methodological terms, is in part, morally prescriptive.
Ibidem, p. 252.
Ibidem, p. 273.
47 Frederick Schauer, Playing by the Rules: A Philosophical Examination of
Rule-based Decision-Making, Oxford, Clarendon, 1991.
45
46
50
That leaves us with Dworkins semantic sting, the reductio ad absurdum of theories that seek to define law by a set
of empirical criteria. This can be dealt with briefly, for it is
clearly mistaken to apply this critique to Hart because he
was so very much against providing tight definitions to encompass complex social phenomena. Here his fascination
with Wittgensteins conception of family resemblances and
subtle differences in linguistic usage come into play. Hart
seeks to bring out the similarities and differences between,
for instance, law and morality in a fluid way. He is not interested in a definition of the term law in terms of the necessary and sufficient conditions of its use.
What then of the rule of recognition? Well, it is clear that,
for Hart, on any reasonable interpretation, rules of recognition are not definitions of law. They relate to a variety of
tests with different criteria in different jurisdictions for the
purpose of providing official determination as to which
rules are to be accepted as being laws in that jurisdiction.
By assuming that this is fulfilling the same function as a
sociologists or a philosophers definition of law Dworkin
makes, although he denies he is making, the mistake of
confusing the method of identifying a social phenomena
and the description of the content of that phenomena. Of
course Hart does use rules of recognition to determine
whether or not a mature legal system exists, but Dworkin
confuses the conceptual/sociological question of what constitutes a legal system from the quite different questions as
to what criteria judges do or should use in identifying primary laws.
IV. CONCLUSION
Why should anyone seek to rewrite the Postscript to the
2nd edition of one of the most famous legal philosophy
books of the late modern period, especially if this involves
going counter to the authors clearly stated contentions
concerning his own work? Passing by the point that the
51
TOM CAMPBELL
publication of the Postscript was not authorised by Hart

himself, the main justification for such a presumptuous exercise is that it may help to display the continuity of one
reading of The Concept of Law, in the form of its immensely
influential 1st edition, with current theories of law that are,
in part, morally prescriptive inclusively of the most general
concepts within philosophical jurisprudence, such as legality, the rule of law and law itself. This may help to reconnect legal positivism with its Hartian precursors, such as
Hobbes and Bentham, and open up the possibility of drawing on the insights of Hart himself to develop a contemporary prescriptive version of legal positivism.
52
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
THE SIMPLE AND SWEET VIRTUES OF ANALYSIS.

A PLEA FOR HARTS METAPHILOSOPHY OF LAW
Pierluigi CHIASSONI*
Resumen:
El captulo 1 de El concepto de derecho plantea una cuarta cuestin fundamental, adems de las otras tres bien conocidas: a saber, la cuestin
meta-filosfica relacionada con el objeto, materia y mtodo de la teora
jurdica. El propsito de este artculo es exponer la filosofa del derecho
hartiana en su mejor versin, mediante la referencia a algunas de sus
ventajas tericas, con el fin de defenderla, en la medida de lo posible, de
algunas de las crticas formuladas por los defensores de otros enfoques
(Raz, Leiter y Dworkin).
Palabras clave:
Concepciones de la filosofa del derecho, metafilosofa del derecho, (nueva) jurisprudencia analtica, enfoque hermenutico, naturalismo, interpretativismo, H. L. A. Hart.
* Universit di Genova (pierluigi.chiassoni@unige.it). A first version of this

paper was presented at the workshop on The Fiftieth Anniversary of The Concept
of Law, Jurisprudence Discussion Group, Faculty of Law, Oxford University,
Oxford, May 26, 2011; a second, slightly revised, version was presented at the
Special Workshop H. L. A. Harts The Concept of Law Reconsidered, organized
by prof. Imer Flores for the 25th IVR World Congress, Goethe Universitt, Frankfurt a.M., August 18, 2011. I wish to thank participants at both events for their
questions and comments.
53
PIERLUIGI CHIASSONI
Abstract:
Chapter I of The Concept of Law raises a fourth, capital, issue, besides the
three well-known ones: i.e., the meta-philosophical issue concerning the
point, the matter, and the method of legal theory. The paper purports to
present Harts philosophy of jurisprudence in its best light, also by referring to some of its theoretical pay-offs, and to defend it, so far as possible,
against a few criticisms by supporters of different outlooks (Raz, Leiter,
and Dworkin).
Keywords:
Conceptions of Jurisprudence, Metaphilosophy of Law, (New)
Analytical Jurisprudence, Hermeneutic Approach, Naturalism,
Interpretivism, H. L. A. Hart.
54
nous ne deviendrons pas Philosophes, pour avoir lu

tous les raisonnements de Platon et dAristote, sans
pouvoir porter un jugement solide sur ce qui nous est
propos. Ainsi, en effet, nous semblerions avoir
appris, non des sciences, mais des histoires.
R. DESCARTES
[Analytical philosophy] is suspicious of grand theory
if it comes along too soon and obscures valuable distinctions.
H. L. A. HART
SUMMARY: I. The Fourth Issue. II. The Model of Ordinary

Analysis. III. Theoretical Claims. IV. Is the Ordinary Analysis Model Inadequate for the Concept
of Law? V. Is the Ordinary Analysis Model at
Odds with Jurisprudential Naturalism? VI. Who is
the Wiser Hedgehog? VII. Concluding Remarks.
VIII. Bibliography.
I. THE FOURTH ISSUE

Chapter I of The Concept of Law, Persistent Questions,
concerns the recurrent problems which puzzle jurisprudential inquiries about the law. As we know, Hart singles out
three such issues as foremost, which he considers to be
buried together under the elusive question What is law?.
First, whether, and to what extent, the law is a matter of
rules; second, why, and how, the law has to do with coercion; third, whether there is any necessary connection between law and morality.1
1 Hart, H. L. A., The Concept of Law, 1961, 2nd. ed., With a Postscript edited by
P.A. Bulloch and J. Raz, Oxford, Great Britain, Oxford University Press, 1994, p.
13: Here then are the three recurrent issues: How does law differ from and how is
related to orders backed by threats? How does legal obligation differ from, and how
is related to, moral obligation? What are rules and to what extent is law an affair of
rules?.
55
PIERLUIGI CHIASSONI
Surely, these are the central issues on which The Concept

of Law turns in its several chapters.
Nonetheless, the whole book and indeed, the whole
work of Herbert Hart, since its beginnings appears to be
preoccupied with a fourth, paramount, puzzling, persistent
issue: namely, with the meta-philosophical issue, concerning
the point, the matter, and the method of legal theory.2
In the province of learning, revolutions are the outcomes
of dramatic, comprehensive changes in philosophical outlook and method. Philosophy of law (legal philosophy, jurisprudence) is no exception. Herbert Harts The Concept of
Law represents the cornerstone of a revolutionary, analytic-philosophy view concerning the path philosophical inquiries upon the law should take in order to be a socially
worthwhile undertaking.
Harts philosophy of legal theory and some of its theoretical offsprings have been contested during Harts lifetime by several critics, Ronald Dworkin being perhaps the
most radical one. After Harts death, further criticisms appeared, suggesting the whole enterprise Hart advocated
since the early 1950s to be either outfashioned, or, from its
very outset, a wrong new detour from older, safer, and
more valuable philosophical approaches.
My purpose in this paper is vindicating, so far as possible, Harts view about the proper point, matter, and method
of legal theory as still valuable here and now, on the benchmark both of the way of posing questions it suggests, and of
the sort of answers it promotes.
In the first part of my paper, I will provide a survey of the
central tenets of the Hartian model of analytical legal theory
( 2) and recall some of the theoretical claims Hart makes
2 The point is set out with crystal-clear determination since the incipit of Harts
1953 inaugural lecture: In law as elsewhere, we can know and yet not understand.
Shadows often obscure our knowledge, which not only vary in intensity but are
cast by different obstacles to light. These cannot all be removed by the same methods, and till the precise character of our perplexity is determined we cannot tell
what tools shall we need (Hart, H. L. A., Definition and Theory in Jurisprudence,
1953, in Hart, H. L. A., Essays in Jurisprudence and Philosophy, Oxford, Great Britain, Clarendon Press, 1983, p. 21).
56
by applying it ( 3). In the second part, I will consider in

turn a few criticism to the Hartian model, set forth by the
supporters of hermeneutical conceptual analysis (Raz),
naturalized jurisprudence (Leiter), and interpretivism
(Dworkin), and suggest that they may be overcome ( 4-7).
II. THE MODEL OF ORDINARY ANALYSIS
Harts philosophy of legal theory Harts view about the
point, matter, and method of legal theory may be recounted in terms of five central tenets: 1) the Clarification
Principle, 2) the Conceptual Analysis Principle, 3) the
Not-Upon-Other-Books Principle; 4) the No-Mystery Principle;
5) the Soft-Tone or Nirvana Principle.
The first tenet concerns the purpose and basic features of
legal theory: the core, so to speak, of its disciplinary charter. The latter four tenets complement the former by identifying the tools and standpoints legal theory must adopt in
its inquiries. Together, the five principles build up a model
of general expository jurisprudence, within the joint tradition of Benthams and Austins analytical jurisprudence
and Oxbridges ordinary language philosophy. In the following, I will refer to this model as the Ordinary Analysis
Model.
2.1 The Clarification Principle
According to the Clarification Principle the first tenet of
Harts Model of Ordinary Analysis the purpose of legal
theory should be the clarification of the general framework
of legal thought, by focussing on its central piece, represented by the concept of law, but also on other related concepts such as the concepts of legal obligation, legal right, legal power, legal rule, legal validity, legal sanction, nullity,
etc., in their connections with the concepts of rule, existence of a rule, having an obligation, being obliged, coercion,
morality, moral obligation, etc.
57
PIERLUIGI CHIASSONI
Legal theory in Harts view does not exhaust jurisprudence as a general, uncompromised, label for the philosophical study of law. It is just one branch of it, side by
side to legal policy, which is concerned instead with the
criticism of law.
So conceived, legal theory and legal policy (critical jurisprudence) are in turn complementary enterprises to legal
science (doctrinal study of law), legal history (historical jurisprudence), and legal sociology (sociological jurisprudence, descriptive sociology of law).
Legal policy is a normative venture, the Hartian version of
Benthams Censorial Jurisprudence and Austins Science of
Legislation. It purports to evaluate existing legal systems,
or some parts thereof, surely from the standpoint of some
previously selected moral or political philosophy (Utilitarianism, Liberalism, Moral-Majority Perfectionism, Free-Market Darwinism, Catholic Fundamentalism, etc.), but also
from the standpoint of instrumental rationality and efficiency. Indeed, as Hart makes clear, the criticism of law
does not necessarily amount to the moral criticism of law.
Contrariwise, legal theory should stick to three basic
standards: generality, structure, and description.3
First, legal theory should be general: it should be about,
and elucidate, the law in general, namely, law as a widespread social and historical phenomenon.
Second, legal theory should be a structural inquiry about
the law: namely, it should be concerned with the conceptual apparatus and institutional arrangement common to
developed legal systems and legal cultures (the general
framework of legal thought, the distinctive structure of a
municipal legal system), not with the peculiar contents of
the legal rules of this or that legal system which is the
matter of the doctrinal study of law.
Third, legal theory should be descriptive. Hart entertains
a complex view about the standard of description, which
embraces both a negative and a positive characterization.
3
Hart, H. L. A., op. cit., n. 1, pp. v, 17, 239 ff.
58
On the negative side, legal theory is descriptive if, but only

if, it is not a justificatory, morally value-laden, morally committed, enterprise. Accordingly, legal theory should stick instead to Max Webers ideal of scientific Wertfreiheit which
requires: on the one hand, a transparent commitment to
certain epistemic values (truth, clarity, evidence, logical and
terminological consistency, empiricism); on the other hand,
the programmatic refusal to pass any open or undercover judgment whatsoever upon positive law on the basis of normative-ethical values (justice, freedom, equality,
might makes right, etc.). In this way, as Hart suggests, legal theory may work as the clear and honest prologue to
any reasonably informed (rational) criticism of law.
On the positive side, legal theory is descriptive if, but only
if, it is a philosophical investigation upon the law along the
lines of P. F. Strawsons descriptive metaphysics. Indeed, it
must be a venture of explanatory elucidation (in Harts own
terms) of the conceptual apparatus in the field of law,
where analysis in a narrow sense, with its therapeutic and
constructive (systematic) sides, combines with thought experiments in the way of philosophical imagination, with its
explanatory and inventive sides (on which I will say a few
more words below).
Descriptive metaphysics bears a two-ways, mutual dependence, relationship with cultural sociology, or, in Hartian
terms, descriptive sociology. On the one hand, descriptive
sociology provides descriptive metaphysics with the rough
data to be explained (the explicanda). On the other hand,
descriptive metaphysics provides descriptive sociology with
an elucidated framework of concepts and tools (the
explicata) to be used for fresh sociological inquiries.4
The Clarification Principle provides a clear view not only
about the point, but also about the matter, of legal theory.
In a truistic, tautological sense, the matter of legal theory
can be nothing else but law, whatever the law is and/or we
4 See Sugarman, D., Hart Interviewed: H. L. A. Hart in Conversation with David Sugarman, in Journal of Law and Society, 32, 2005, pp. 267-293, at pp. 289 ff.
59
PIERLUIGI CHIASSONI
mean by that. Going beyond truism and tautology, however, the Clarification Principle suggests to regard the law
as a linguistic phenomenon: if not strictly speaking, at least
in the looser sense of law being, in the main, a matter of
sentences, words and concepts. Such a linguistic conception of law which is fully in tune with Harts double philosophical allegiance: to Benthams analytical jurisprudence,
on the one side; to the Oxbridge ordinary language philosophy, on the other side opens the way to the second tenet
of the Hartian Model of Ordinary Analysis: the Conceptual
Analysis Principle, to which I turn now.
2.2 The Conceptual Analysis Principle
According to the Conceptual Analysis Principle, clarification of the general framework of legal thought, so as to provide an improved analysis of the distinctive structure of
municipal legal systems, requires a careful analysis of the
law-talk. This requires in turn philosophical definitions and
explanatory elucidations that must be brought about not
only by the plain, lexicographic, record of linguistic conventions, but also, as I said before, by means of a reconstructive approach including thought experiments in the way of
philosophical imagination.
The several tools Hart singles out for his Ordinary Analysis Model of legal theory are well-known. Nonetheless, they
make up a tool-box that is so sophisticated in kind and
number, to be worth of a survey.
The tools of the Hartian Model of Ordinary Analysis fall,
very roughly speaking, into three main groups of related
and conspiring items: 1) Linguistic tools; 2) Hermeneutic
tools; 3) Philosophical tools. In fact, from a genealogical
standpoint, all these tools are philosophical: they are all the
outputs of (mostly) analytical forms of philosophizing. Accordingly, the distinction I draw holds, in the main, from the
standpoint of their respective use: the analysis of natural
languages and natural discourses in view of therapeutic
60
and systematic goals; the analysis of law-talk and other

sorts of norms- or rules-talks in a society; the working out
of insightful explanations connecting our conceptual apparatus, as identified and tentatively established by means of
linguistic and hermeneutical tools, to structural, persistent,
features of the human condition.
Keeping this warning in mind, lets have a look at the
tools.
Linguistic Tools
Linguistic tools mirror Harts familiarity with the major
achievements of Oxbridge, post WWII, ordinary language
philosophy combined with his early-bird awareness of the
powerful insights that Jeremy Bentham embodied in the
very heart of his program for a new, revolutionary, jurisprudence. They encompass three basic elements: a) a theory of
natural languages; b) a theory of definition; c) a theory of
concepts.
Harts theory of natural languages the building-blocks of
which are such ideas as meaning-as-use, the variety of linguistic uses or functions (with a special focus on the operative, prescriptive, and justificatory functions), the criticism
of the descriptivist, objectivist, and deductivist fallacies,
language-levels, etc. has its core in the claim that natural
languages are fairly, but not throughly, efficient tools of human communication, provided their words and sentences
are characterized by a physiological fringe of indeterminacy
in the forms of vagueness (from ordinary gradual,
combinatory, analogical, etc. vagueness to open texture)
and ambiguity.
Harts theory of definition along the lines of Benthams
path-breaking theorizing about fictitious entities and the
method of paraphrasis turns upon three ideas: first, the
variety of definitions; second, the double-instructive virtue of
definitions; third, the limits of definitions. The first idea
the variety of definitions emphasizes that definition by
61
PIERLUIGI CHIASSONI
genus and specific difference (per genus et differentiam

specificam) needs to be complemented by further, different,
forms of definition, to wit, by contextual definition and central-case definition, that are more suitable to definienda for
which either no clear genus, or no closed set of necessary
and sufficient properties, common to all the referred items,
is available. The second idea the double-instructive virtue
of definitions is meant to contrast the view according to
which definitions are just about words. This is done, following J. L. Austin, by pointing to the fact that a (good) definition of a term does provide useful instruction not only
about the ways the term is or may be used, but also about
the very things it does refer to. The third idea the limits of
definition, for which a clear statement may be found in
John Austins On the Uses of the Study of Jurisprudence, a
work Hart edited in 1954 as an appendix to The Province of
Jurisprudence Determined is tantamount to the rejection
of what may be called the definitional fallacy: the methodological blunder which consists in the pretence of solving
some complex theoretical problem by way, and in the short
space, of the few sentences making up the definition of a
phrase.5 Surely, theories have at their heart some definition
of their key-terms;6 but they cannot be altogether superseded by definitions.
Finally, Harts theory of concepts is likewise characterized
by three backbone ideas. First, concepts are either a matter
of convention, or a matter of stipulation. There are no true
concepts outside of the realm of the ordinary uses of words.
There are no true concepts in some rarefied dimension of
5 Austin, J., The Uses of the Study of Jurisprudence, in Id., The Province of
Jurisprudence Determined, 1832, with an Introduction by H. L. A. Hart, New York,
The Noonday Press, 1954, pp. 370-371; Hart, H. L. A., Definition and Theory in
Jurisprudence, 1953, in Hart, H. L. A., op. cit., n. 2, pp. 25-26, 26 ff., 47 fn 28.
6 According to Hart, adequate descriptive theories of social phenomena like
law are made by three basic ingredients: definitions; empirical statements about
passing features of the world (ordinary statements of fact); empirical statement
about constant features of human beings and their world (statements the truth of
which is contingent on human beings and the world they live in retaining the salient characteristics they have). Hart, H. L. A., op. cit., n. 1, pp. 199-200.
62
real essences, as legal conceptualists, like the great German jurist Rudolph von Jhering (in his first period), maintained.7 Second, stipulative concepts are neither true, nor
false, but only liable to pragmatic justification. Their value, if
any, depends on the goal(s) they are meant to serve, and on
how they have been worked out in view of those goals.
Third, theoretical concepts, like those worked out by legal
theory, are stipulative concepts informed by an overall explanatory goal.8
Hermeneutic Tools
Hermeneutic tools mirror Harts interest not only for sociology, but also, I may venture to say, for anthropological
philosophy and philosophical anthropology.9 The former
deems it worthwhile making use of anthropologists (supposed) tools and perspectives for philosophizing. The latter
assumes philosophy may help in the working out and
sharpening of tools for anthropological inquiries. At this
crossroad of perspectives and suggestions, fueled by the perennial fascination educated men feel for their primitive fellows, Hart lays down the well-known distinction between
the internal and the external point of view as to the normative system(s) of any given society. In so doing, he is careful
to reject a purely behaviouristic conception of the external
standpoint, in favour of an hermeneutic external standpoint,
where the observer does not only record the participants
non-linguistic behaviours, but also takes into account participants own norms-talk, and, what is more, puts herself in
participants shoes to see their norms and their actual
7 Hart, H. L. A., Jherings Heaven of Concepts and Modern Analytical Jurisprudence, 1970, in Id., op. cit., n. 2, pp. 265-277.
8 Hart, H. L. A., op. cit., n. 1, pp. 213-214.
9 See Hart, H. L. A., op. cit., n. 1, p. 289, where, besides P. Winch, Hart also
quotes an essay by R. Piddington on B. Malinoswkis theory of needs, and p. 291,
where works by Malinowski, A.S. Diamond, K.N. Llewellyn and W. Hoebel are
quoted.
63
PIERLUIGI CHIASSONI
working as they do see them.10 The observer/user distinction is connected in turn to the key-difference, the theory of
natural languages made somehow easier to perceive, between statements about a normative system (external statements), on the one hand, and statements using a normative
system to do things like making or rejecting claims, ascribing rights, duties and responsibilities, evaluating, judging,
criticizing and justifying behaviours (internal statements), on
the other.11 In this way, the observer/user distinction
points to, and emphasizes, the epistemic value of awareness, for it prompts each of us, who are most of the time
both users and observers of normative systems, to ask at
any circumstance the capital questions: What am I doing
now?, What is she doing there?.
Philosophical Tools
Philosophical tools, the last set in the present survey of
Harts jurisprudential tool-box, mirror his deep commitment
to refounding jurisprudence as a worthwhile, sophisticated,
genuinely philosophical enterprise, not confined to the writing down some of tiresome linguistic spicilges, but providing
useful elucidations of the structure of legal thought and legal
institutions. Here, we find three leading ideas: the principle
of methodical ignorance; the antireductionism principle; the
method of philosophical imagination.
10 Hart, H. L. A., op. cit., n. 1, pp. 88 ff., 239 ff.; see also Id., Theory and Definition in Jurisprudence, in Proceedings of the Aristotelian Society, 29, 1955, pp.
247-249; Id., Essays on Bentham. Jurisprudence and Political Theory, Oxford, Great
Britain, Clarendon Press, 1982, pp. 106-161; Id., Introduction, in Hart, H. L. A.,
op. cit., n. 2, pp. 13 ff. In his Pragmatische Antropologie, Immanuel Kant draws a
like distinction between knowing the world and having use of the world: the former amounts to the knowledge of a game to which we assist; the latter amounts
instead to participating to the game (Kant, I., Antropologia dal punto di vista
pragmatico, in Id., Critica della ragion pratica e altri scritti morali, ed. by P. Chiodi,
Torino, Italia, Utet, 1970, p. 542).
11 Hart, H. L. A., op. cit., n. 10, p. 248: We can contrast the external standpoint
of the observer of a legal system who is thinking about its rules and their present
and future operation with the internal standpoint of one who is using the rules of
the system either as an official or private person in various ways; see also Id., Definition and Theory in Jurisprudence, 1953, in Hart, H. L. A., op. cit., n. 2, p. 27.
64
The principle of methodical ignorance advocated as a basic methodological tenet for jurisprudence by Bentham and
Austin is a safety device against knowledge by acquaintance. Familiarity with words and objects, far from providing deep and reliable knowledge about social phenomena, is
rather likely to lead to confusion and delusion. Accordingly,
jurisprudence must be the art of systematically ignoring
what people at large (pretend to) know, starting from peoples linguistic and conceptual certainties.12
The anti-reductionism principle is a safety device against
unwarranted theoretical reductionism. The works of many
legal philosophers exhibit the tendency to reduce the complexity of legal systems in order to show them, so far as
possible, as structurally simple phenomena, characterized
by a pleasing uniformity of pattern. Contrariwise, the
anti-reductionism principle suggests to preserve the complexity of legal systems by means of a network of adequately articulated concepts, so far as it is conceptually
warranted by functional and other practical differences.13
As I said before, Hart regards the method of philosophical
imagination as a major tool in the game of descriptive metaphysics. In Harts understanding, philosophical imagination
requires the working out of thought experiments meant to
explain how our actual conceptual and institutional structures are, and why, by comparing them with alternative
imaginary situations. Three thought experiments lay at the
core of Harts legal theory: 1) the simple model of law as coercive orders; 2) the idealized picture of a primitive, pre-legal, society ruled only by a discrete set of unconnected primary norms of obligation, somehow preluding to Nozicks
invisible-hand explanation for the rise of the state out of a
Lockean state of nature; 3) the theory of the minimum content of natural law: the empirical theory of natural law
12 Remember: In law as elsewhere, we can know and yet not understand (Hart,
H. L. A., Definition and Theory in Jurisprudence, 1953, in Hart, H. L. A., op. cit.,
n. 2, p. 21, italics added).
13 Hart, H. L. A., op. cit., n. 1, pp. 38 ff.
65
PIERLUIGI CHIASSONI
Hart opposes both to traditional natural law theorists and

to Kelsenian legal positivism.
So far for the Principle of Conceptual Analysis. The three
latter principles of Harts Model of Ordinary Analysis may
be recounted in shorter terms.
2.3 The Not-Upon-Other-Books Principle
The third tenet of Harts Ordinary Analysis Model is the
pedagogically aimed Not-Upon-Other-Books Principle.14
In its positive side, the principle prescribes legal theory to
be about the law and related social phenomena, not about
legal theory books.
In its negative side, the principle purports to rule out jurisprudential theology: the way of doing legal philosophy
where jurisprudents identify some very broadly formulated
truth (or likely true intuitions) about the law as the focus of their investigations (like, e.g., law is fact, law is
convention, law is force, law is norm, law is morality,
law depends on social sources, law is interpretation,
etc.), and then spend their time discussing competing interpretations of such assumed truths, in a never-ending series of argument, counter-argument, counter-counter-argument, and so on.
2.4 The No-Mystery Principle
The fourth tenet of Harts Ordinary Analysis Model is the
No-Mystery Principle.
Due to the influence of natural law thinking, jurisprudence has been conceived for a long time, and is still being
conceived now, as an investigation upon the nature or essence of law: assuming, in so doing, that the law does have
one true nature, one true essence, liable to be discovered by
means of adequate inquiries.
14
Ibidem, p. vi.
66
In Harts view, however, such naturalist or essentialist

conceptions of jurisprudence are metaphysically suspect.
They build around the law an unnecessary, misleading halo
of mystery, since the very way they pose their starting
questions (What is the nature of law?, What is the essence
of law?). For, as Hart suggests, there is no such a thing as
the one true nature or the one true essence of law. All we
have in fact, is a general social phenomenon we are used to
call law (derecho, diritto, droit, Recht), so that any philosophical inquiry upon such phenomenon may simply be
conceived as purporting to provide an answer to plainer,
metaphysically safer, questions like What is law? or
What is the concept of law? to be read, of course,
against the background of Harts Clarification and Conceptual Analysis principles. Accordingly, if we nonetheless do
like using such phrases as the nature of law or the essence of law, we should at least surround them with the
sanitary belt provided by inverted commas.15 Indeed, as
Hart suggests, surely things do have properties; but they do
not have any essential property, outside of some determinate conceptual frame.
2.5 The Soft-Tone Principle
The fifth, and (for the present survey) last, tenet of Harts
Ordinary Analysis Model is the Soft-Tone or Nirvana Principle.
Legal philosophers frequently present their (presumed)
discoveries about the law by way of immoderate forms of
speech. They tend to great exaggerations, surely pour pater
les juristes, but with a price: for they raise unnecessary
paradox and puzzlement among jurisprudents and educated public opinion alike, which call for adequate deflationary analysis. In Harts view, on the contrary, legal theorists should carefully avoid exaggerations, and prefer
15
As Hart himself does. See, e.g., Hart, H. L. A., op. cit., n. 1, p. 155.
67
PIERLUIGI CHIASSONI
instead (what may be regarded as) a more modest,

craftsmanlike, Nirvana approach, aiming at cool definitions:16 it goes without saying, to be regarded as tentative
and always revisable.
Harts own work as a whole provides a clear evidence of
his unflinching abiding to the Soft-Tone principle.
III. THEORETICAL CLAIMS
So far, I have stayed with Harts philosophy of jurisprudence. It is time now, before proceeding to consider a few
rival views, to recall the answers Hart offered to legal theorys persistent questions by employing the methodological
apparatus of the Ordinary Analysis model. We all (presume
to) know those answers, so a very swift account should go.
The law and rules issue. Hart sets forth a paradigm-case
definition of the concept of law (law as the union of primary and secondary rules), warning it should not be used
as a strict definition of expressions like law, legal, or legal system ruling upon linguistic uses, but simply as a theoretical device (a concept) whose value must be measured
on the rod of its ability to promote clearer theoretical inquiries and moral deliberation. There is, accordingly, nothing
odd, nor self-contradictory, in Harts entitling his major
book The Concept of Law, provided one has a clear view
about Harts theory of concepts and overall Ordinary Analysis model.17
Ibidem, p. 2.
Dealing with the law and rules issue, Hart also defends a normativist conception of law (law is made of rules) against radical rule-scepticism, by resorting to
the linguistic dependence of legal rules (legal rules work like sentences in a natural
language). This view, so far as its bearing on the theory of intepretation is concerned, was criticized for overlooking the actual practice of written-law interpretation. Hart apparently accepted the criticism, conceding that interpretive methods
may make the idea of there being rules which pre-exist to their judicial application
troublesome (see Hart, H. L. A., Introduction, in Hart, H. L. A., op. cit., n. 5, pp.
8-9). In this way, however, his view about rules, and law being made of rules, becomes quite similar to Kelsens and soft realists frame-theory of rules.
16
17
68
The law and coercion issue. Hart outlines an explanation

of why coercive sanctions are a constant feature of the social phenomenon of law, to be taken into account within
our concept of law here and now, by appealing to a
Hobbesian-Humean model of the human condition, where
sanctions are needed as a guarantee to voluntary cooperation against free-riding.18
The law and morality issue. Hart defends a quite sophisticated view on this issue. Such a view, however, has been
somehow obscured by the there is no necessary connexion
between law and morals shibboleth. Three basic claims
make up the Hartian theory of the relationships between
law and morals: first, the multiple contingent connexions
claim; second, the conceptual neutrality claim; third, the
no-committment claim.
It is worthwhile emphasizing that these three claims,
jointly considered, have in themselves a strong pedagogical
import, if I may say so. For they make clear that the law
and morality issue is not, really, one issue, but a set of heterogeneous problems, which are very often dealt with in a
wholesale, confused way, but calling for separate standpoints, separate approaches, and separate answers. It goes
without saying that such a problematization of the issue
represents a valuable pay-off following from the adoption of
the Ordinary Analysis approach.
The multiple contingent connexions claim points to the several ways positive legal systems may be related to systems
of social and/or critical morality (content, motivation, existence, validity, interpretation, criticism, instrumentality)
as they may also be related, in fact, to religious outlooks,
economic systems, rules of etiquette, etc. It suggests that
any idea of a special, necessary, relationship of conceptual
subordination of law to morality (Which morality, by the
way?) is, so to speak, in the eye of the beholder.19
18
19
Hart, H. L. A., op. cit., n. 1, pp. 191 ff.

Hart, H. L. A., op. cit., n. 1, pp. 185 ff., 200 ff.
69
PIERLUIGI CHIASSONI
The conceptual neutrality claim defends the virtues of a

broader, morally-neutral, concept of law as compared with
the narrow, morally-laden, concepts advocated by nonpositivists (Natural lawyers & C., like Gustav Radbruch
and, presently, John Finnis and Robert Alexy), since such a
broader, neutral concept would be preferable to theoretical
and practical purposes alike.20
Finally, the no-committment claim maintains that legal
theory should keep itself out of the controversial philosophical issue concerning the ontological status of moral values.21 This claim plays a key role in shaping Harts own version of legal positivism vis vis to inclusive (or soft) and exclusive (or hard) positivism alike.
Soft positivists make two basic claims. First, the validity
and content of legal norms may depend on moral criteria: it
is not necessarily the case that validity and content of legal
norms do not depend on morality (contingent moral validity-clauses thesis). Second, whenever a legal system includes a moral-validity clause (such as, e.g., the dignity
clause of the European Constitution or the due process
clauses of the American Constitution), such a system incorporates i.e., converts into pre-existing law, from the standpoint of adjudication both the moral norms referred to by
the clause, and all the full-fledged norms the content of
which may be derived from them (incorporation thesis).
Hard positivists, as we know, claim roughly the opposite.
First, the validity and content of legal norms cannot depend
on moral criteria: it is necessarily the case that the identification and content of legal norms do not depend on moral
argument (social sources, or no moral-argument, thesis). Second, whenever a legal system includes a moral-validity
clause, such a clause is tantamount to delegating to judges
and other officials the power of making new law according
Ibidem, pp. 207 ff., 213-214.
Ibidem, pp. 250 ff., at pp. 253-254: I still think legal theory should avoid
commitment to controversial philosophical theories of the general status of moral
judgments and should leave open [...] the general question of whether they have
what Dworkin calls objective standing.
20
21
70
to certain moral criteria (law-making-power delegation thesis).

Surely, Hart claims against Dworkin to be not a
plain-fact positivist, but, rather, a soft positivist. As a soft
positivist, however, Hart does not go all-the-way through
along with his fellow softers. He stops at the first claim,
concerning the possibility of moral validity-clauses. As to
the second claim, the incorporation thesis, he takes a different view. This ultimately makes of Harts legal positivism
a third theory, au dessus de la mle: above the contest between inclusive and exclusive positivists.
Harts reasoning runs, very roughly, as follows. One: soft
positivist provisions (moral validity-clauses) may safely be
interpreted as incorporation devices if, but only if, the morality they refer to has an objective standing. Two: unfortunately, the objective standing of moral norms and values is
controversial. Three: consequently, soft positivists may
safely adhere to the moral-validity clause claim, but not
also to the incorporation claim, since, qua legal theorists,
they cannot rule out that moral-validity clauses may in fact
work as power-conferring rules, i.e., as directions to courts
to make law in accordance with morality. Four: likewise,
hard positivists, qua legal theorists, must leave open the
possibility of moral-validity clauses working as incorporation devices. Five: the objective standing of moral values,
whatever we mean by that, however, makes no practical difference from the standpoint of the judges called to apply
moral-validity clauses. In any case, their duty will be the
same: namely, to make the best moral judgment they can,
on any moral issue they have to decide.
As I said at the beginning, I think Harts philosophy of legal theory does represent, still here and now, a valuable
outlook. Surely, like every outlook, it is to be regarded as
open to continuous refinement and amendment. But its
central ideas seem to outline a useful and reliable model of
philosophical inquiry upon the law. Surely, we may not
agree with Harts theoretical claims, and think better ones
71
PIERLUIGI CHIASSONI
may be set forth and argued for. But these shortcomings,

by themselves, do not involve the repudiation of the Ordinary Analysis model which, by the way, is a model for legal theory: neither for the whole philosophical investigation
upon law, which includes the working out of grand justificatory theories of law, nor for the whole study of law.
Nonetheless, Harts way has been the target of criticisms,
even radical ones. I will consider now a few of them, raised
by Joseph Raz, Brian Leiter, and Ronald Dworkin, respectively.
IV. IS THE ORDINARY ANALYSIS MODEL INADEQUATE
FOR THE CONCEPT OF LAW?
According to Eugenio Bulygin, conceptual analysis may
be conceived as aiming at the explication or rational reconstruction of concepts:
Es una vieja tradicin analtica llamar el proceso que conduce de un concepto a otro mejor, i.e., ms exacto, explicacin o reconstruccin racional (Carnap). Las diferentes teoras
del derecho se esfuerzan por formular un concepto de derecho ms exacto y apropiado de acuerdo con algn criterio
terico, como la simplicidad, la fecundidad e incluso la elegancia de la presentacin.22
Joseph Raz, however, objects that such a way of understanding conceptual analysis may work for the concepts of
the natural sciences, which are descriptive, explanatory
concepts, but it cannot work for social concepts, like the
concept of law, legal right, gift, property, marriage, duty,
etc. This is so, Raz claims, because social concepts are hermeneutic concepts: they are concepts we use to understand
ourselves, other people, and our position in the world. They
are not simply explanatory tools; they also contribute to
22 Bulygin, E., Raz y la teora del derecho. Comentarios sobre Puede haber
una teora del derecho? de Joseph Raz, in Raz, J., Alexy, R., Bulygin, E., Una
discusin sobre la teora del derecho, Madrid-Barcelona, Spain, Marcial Pons, 2007,
pp. 107-108.
72
shaping the very social world we try to understand and do

belong to.23
Bulygin shares the same view of conceptual analysis as
Hart. As a consequence, Razs criticism to Bulygin also applies to Harts Ordinary Analysis model.
Is Raz right? Does the Ordinary Analysis model work for
natural concepts only, but is really inadequate to cope with
social, constitutive concepts, like the concept of law?
I do not think so. We have seen that Harts tool-box includes a sophisticated, hermeneutic, conception of the external point of view, which requires observers both to analyse users norm-talks, and also to put themselves in users
shoes, in order to grasp the way they understand the concepts they use. Obviously, such an hermeneutic external
point of view may be usefully adopted, by each participant
to a social institution, to get a detached understanding of
her own, and her fellows, committed understanding of such
an institution, for here too holds the description principle, so
elegantly formulated by Hart in his Postscript: Description
may still be description, even when what is described is an
evaluation.24 Now, this is precisely what Razs conceptual
analysis for social concepts seems to require. Accordingly,
Harts conceptual analysis, far from being outdated and unfit to deal with social concepts, appears perfectly in tune
with the hermeneutic standpoint adopted by Raz.
V. IS THE ORDINARY ANALYSIS MODEL AT ODDS
WITH JURISPRUDENTIAL NATURALISM?
One of the most powerful reform-proposals concerning
the methodology of legal theory that has been advanced in
23 Raz, J., Teora y conceptos. Rplica a Alexy y Bulygin, in Raz, J., Alexy, R.,
Bulygin, E., op. cit., n. 22, pp. 119-120; see also Raz, J., Can There Be a Theory of
Law?, 2004, in Id., Between Authority and Interpretation. On the Theory of Law and
Practical Reason, Oxford, Great Britain, Oxford University Press, 2009, pp. 17 ff.
24 Hart, H. L. A., op. cit., n. 1, p. 244; see also Id., Comment, in R. Gavison
(ed.), Issues in Contemporary Legal Philosophy. The Influence of H. L .A. Hart, Oxford, Great Britain, Clarendon Press, 1987, pp. 35-42.
73
PIERLUIGI CHIASSONI
recent years, is the call for a Naturalized Jurisprudence

launched by Brian Leiter.25
Apparently, Leiter criticizes Harts jurisprudence on three
counts.
First, Harts theory is committed to soft or inclusive positivism. Unfortunately, naturalistic inquiries upon the law,
like e.g. the Attitudinal Model developed by Segal and
Spaeth, show the hard positivists concept of law to be
preferable.
Second, Harts theory does not provide an adequate account of the influence judges own ideological attitudes
plays on judicial decision-making, being focussed, instead,
on the problem of the normativity of law and judicial acceptance of legal rules.26
Third, Harts theory, committed as it is to conceptual
analysis, faces the following dilemma, which I propose to
call Leiters Dilemma: either it is just a piece of glorified
lexicography, and so it is a pointless enterprise; or it endorses an immoderate, unwarrantedly ambitious, view
about the virtues of conceptual analysis, and so it necessarily misses its target. For Leiter claims conceptual
analysis cannot in fact illuminate the reality, i.e., the nature of law; it can, and do, illuminate rather, the nature of
our talk about law.27
25 Leiter, B., Naturalizing Jurisprudence. Essays on American Legal Realism and
Naturalism in Legal Philosophy, Oxford, Great Britain, Oxford University Press,
2007.
26 I draw the criticism from the following passage of Leiter, B., op. cit., n. 25, p.
188: the best causal explanation of decision, the Attitudinal Model, is one that relies centrally on Hermeneutic Concepts: for it is supposed to be the attitude of
judges towards the facts that explains the decision, and attitudes are clearly
meaningful mental states that are assigned a causal role in accounting for the outcome (the decision). But a judges favorable moral attitude towards, e.g., privacy in
the home which might be the attitude explaining some of his votes in
search-and-seizure cases is not the same as the kinds of Hermeneutic Concepts
that H. L. A. Hart treats as central to the phenomenona of modern legal systems:
for example, that official accept some rules from an internal point of view, that is,
as imposing obligations on them of compliance.
27 Leiter, B., op. cit., n. 25, p. 196: But on Farrells (more plausible) rendering
of conceptual analysis, we do not illuminate the reality, i.e., the nature of law, we
illuminate, rather, the nature of our talk about law.
74
If we take into account the picture of Harts philosophy of

legal theory and theoretical claims I outlined before, Leiters
criticisms seems open to rejection.
As to the first criticism, one may reply that, so far as the
incorporation claim is concerned, Hart is neither a soft, nor a
hard positivist. Indeed, Hart endorses and defends a more
sophisticated version of legal positivism, which, on the one
hand, suggests legal theorists, qua legal theorists, should
leave the controversial issue about the objective standing of
moral values open, and, on the other hand, claims that
judges, when applying moral validity-clauses, must decide,
in any case, by making their best moral judgment on the issue at stake. This last point, in turn, suggests that Harts
theory of law may be regarded as being in fact in tune with
the outcomes of the Attitudinal Model of adjudication
sponsored by Leiter.
As to the second criticism, one may notice that, according
to naturalized jurisprudents, what makes the law work is
judges attitudes: like, for instance, their favourable moral
attitude towards [...] privacy in the home. Now: what does
it mean having a favourable moral attitude towards [...]
privacy in the home? Hart would say that it means roughly
the following: that (some or most) judges accept, for moral
reasons, the principle of privacy as a paramount normative
standard that should guide social behaviours and provides
overwhelming justification to judicial decisions sanctioning
conducts trespassing on privacy. Accordingly, contrariwise
to Leiters suggestion, judicial ideologies and moral attitudes are liable to be conceptualized in terms of Harts legal
theory.
As to the third criticism, two replies are in order. First, that
it is really surprising to find a naturalized jurisprudent assuming that there is something, representing the nature of
law, beyond the universe of words and sentences making
up legal norms and legal-norm-talk for law, as Hart
rightly suggests, is basically a linguistic social phenomenon.
Second, Harts conceptual analysis is neither the ambitious
75
PIERLUIGI CHIASSONI
enterprise described by Leiter, nor glorified lexicography. It

instantiates, rather, a third genus of conceptual analysis:
the prudent, illuminating, rational-reconstruction, explanatory, imaginative, conceptual analysis, so well described by
Bulygin and elegantly advocated by Strawson.
VI. WHO IS THE WISER HEDGEHOG?
The fox knows many things, but the hedgehog knows
one big thing. In his last book, Justice for Hedgehogs, Ronald Dworkin adds a new link to his criticism of Hartian jurisprudence, by appealing to an interpretivist theory of law
characterized both by a one-system, integrated view, of the
relationships between law and morals (law is a branch of
political morality), and by an integrated view of values (the
several values in the several dimensions of human life really make up a unitary whole).28
Leaving aside Dworkins substantive claims which, by the
way, appear open to criticism on several counts,29 what
about the methodological outlook of interpretivist jurisprudence?
A tentative survey suggests the following principles to be
afoot: 1) Rely on self-evidence; 2) Take norm-talk and justificatory-talk at face value; 3) Do not waste your time distinguishing between the external and internal standpoint, internal and external statements, observers and users, but
just go to the heart of the matter; 4) Do not care about
charitable interpretation; 5) Do not care about inconsistencies; 6) Do not care about establishing carefully the meaning of the key-terms and key-distinctions you employ in
your argument.
28 Dworkin, R., Justice for Hedgehogs, Cambridge, Mass., London, England,

The Belknap Press of Harvard University Press, 2011.
29 Is the one-system view really different from the interpretivist two-systems
view? By what test are the implicit principles which would fit and justify positive legal materials moral principles? Which sort of argument, if, any, does support the
claim that law is a branch of political morality?
76
Following these tenets, Dworkin regards it to be an obvious connection between law and morality, that when a
community decides what legal norms to create, it should be
guided and restrained by morality. He considers legal positivism and interpretivism as two theories competing to provide an answer to the same question, as rival normative
political theories. He claims, accordingly, that the best interpretation of legal positivism makes it tantamount, though
with nuances of course, to the notorious view criticized by
natural-lawyers like Radbruch: i.e., ideological positivism or
Gesetz ist Gesetz positivism.30 He suggests, furthermore,
that a legal positivist like Bentham should be regarded,
paradoxically, as a closet interpretivist.31 He seems to reject and to accept the is/ought distinction, at the same
time, apparently, on different understandings of such distinction.32 He seems to associate Hart to the jurisprudence that has traveled from some declaration about the
essence or very concept of law to theories about rights and
duties of people and officials.33
Lets go back to the hedgehog. In fact, there are many
hedgehogs around, each one knowing, by definition, just
one thing but a big one.
Now, it is unlikely that all these hedgehogs around do
know the same one big thing. Indeed, if we cast a cursory
glance at the world of jurisprudence, we see to stay with
the Anglo-American world two competing hedgehogs. The
Dworkin, R, op. cit., n. 28, pp. 409-410.
Ibidem, p. 486, fn 6.
32 One thing is the is/ought distinction within the Humean tradition sponsored
for jurisprudence by Bentham and his analytical followers; another thing is the
is/ought distinction as a distiction, within an interpretive, value-laden, committed
practice, between de iure condito (or de moribus conditis) considerations and de iure
condendo (or de moribus condendis) considerations. But Dworkin overlooks the
point, though it is crucial for understanding in which way the two-systems picture, keeping law and morality separate, is, in Hartian terms, right. See Dworkin,
R., op. cit., n. 28, pp. 407-409, and 410 ff., where he actually, though perhaps
unawarely, defends a Radbruchian two-tiered solution to the evil law puzzle similar to Gustav Radbruchs. See Radbruch, G., Statutory Lawlessness and Supra-Statutory Law, 1946, in Oxford Journal of Legal Studies, 26, 2006, pp. 1-11.
33 Dworkin, R., op. cit., n. 28, p. 407.
30
31
77
PIERLUIGI CHIASSONI
first, self-proclaimed, hedgehog is, of course, Ronald

Dworkin, who pretends to know the one big thing of the
unity of law and morality, pointing in turn to the unity of
value. The second hedgehog, if I may venture to say so, is
Hart. However, the one big thing Hart knows, and urges us
to know, is not a matter of substantive theory but, rather, a
matter of philosophical methodology. The Ordinary Analysis
Model, with its several principles, tools, and caveats, is
Harts one big thing.
How big such a thing is, may be gathered from a comparison with Dworkins methodology, even on the basis of the
swift remarks above. On the whole, the Ordinary Analysis
Model preludes both to a better (clearer, more precise, candid) legal theory, and to a better (clearer, more precise, candid) political philosophy. The Hartian outlook is wider, and
richer, and subtler, than the Dworkinian outlook, which is
poorer, confused, made of arguments frequently ignoring
the sophistications of (Hartian) positivism. So: who is the
wiser hedgehog, after all? I think there should be no doubt
about the right answer to such a question.
VII. CONCLUDING REMARKS
Interpretivism (Dworkin), naturalized jurisprudence
(Leiter), and hermeneutic conceptual analysis (Raz) represent, perhaps, the three most powerful post-Hartian philosophies of jurisprudence.
Their critical import upon the Ordinary Analysis model,
however, appears to be modest, if not misguided or missing
the target.
We may draw something of a moral, out of this story:
methodologically-aware jurisprudents would do better taking into account the Hartian Model, if only to refine and
complement it, also in the years ahead.
78
VIII. BIBLIOGRAPHY
AUSTIN, J., The Province of Jurisprudence Determined, 1832,
with an Introduction by H. L. A. HART, New York,
The Noonday Press, 1954.
BULYGIN, E., Raz y la teora del derecho. Comentarios sobre
Puede haber una teora del derecho? de Joseph
Raz, in RAZ, J. et al., Una discusin sobre la teora del
derecho, Madrid-Barcelona, Spain, Marcial Pons,
2007.
DWORKIN, R., Justice for Hedgehogs, Cambridge, Mass., London, England, The Belknap Press of Harvard University Press, 2011.
HART, H. L. A., Definition and Theory in Jurisprudence,
1953, in Id., Essays in Jurisprudence and Philosophy,
Oxford, Great Britain, Clarendon Press, 1983.
, Theory and Definition in Jurisprudence, in Proceedings of the Aristotelian Society, 29, 1955.
, The Concept of Law, 1961, Second edition, With a
Postscript edited by P. A. Bulloch and J. Raz, Oxford,
Great Britain, Oxford University Press, 1994.
, Jherings Heaven of Concepts and Modern Analytical
Jurisprudence, 1970, in Id., Essays in Jurisprudence
and Philosophy, Oxford, Great Britain, Clarendon
Press, 1983.
, Essays on Bentham. Jurisprudence and Political
Theory, Oxford, Great Britain, Clarendon Press,
1982.
, Introduction, 1983, in Id., Essays in Jurisprudence
and Philosophy, Oxford, Great Britain, Clarendon
Press, 1983.
, Comment, in R. Gavison (ed.), Issues in Contemporary
Legal Philosophy. The Influence of H. L. A. Hart,
Oxford, Great Britain, Clarendon Press, 1987.
79
PIERLUIGI CHIASSONI
KANT, I., Antropologia dal punto di vista pragmatico, in Id.,

Critica della ragion pratica e altri scritti morali, ed. by P.
Chiodi, Torino, Italia, Utet, 1970.
LEITER, B., Naturalizing Jurisprudence. Essays on American
Legal Realism and Naturalism in Legal Philosophy,
Oxford, Great Britain, Oxford University Press, 2007.
RADBRUCH, G., Statutory Lawlessness and Supra-Statutory
Law, 1946, in Oxford Journal of Legal Studies, 26,
2006.
RAZ, J., Teora y conceptos. Rplica a Alexy y Bulygin, in
RAZ, J. et al., Una discusin sobre la teora del derecho,
Madrid-Barcelona, Spain, Marcial Pons, 2007.
, Can There Be a Theory of Law?, 2004, in id., Between Authority and Interpretation. On the Theory of
Law and Practical Reason, Oxford, Great Britain,
Oxford University Press, 2009.
SUGARMAN, D., Hart Interviewed: H. L. A. Hart in Conversation with David Sugarman, in Journal of Law and Society, 32, 2005.
80
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
PULLING OFF THE MASK OF LAW: A RENEWED

RESEARCH AGENDA FOR ANALYTICAL LEGAL
THEORY
Keith CULVER
Michael GIUDICE
Resumen:
El objetivo de este artculo es identificar y defender una parte de la renovada agenda de investigacin de la teora jurdica analtica: un renovado
enfoque relacional sobre la manera de caracterizar el concepto de derecho siguiendo la lnea trazada por Hart de investigar la relacin del derecho con la moral, la coercin y las reglas sociales. Nosotros defendemos una investigacin adicional de carcter descriptivo-explicativo sobre
la relacin que tiene el derecho con la seguridad, el medio ambiente y las
tecnologas de la informacin, en el contexto de disposiciones jurdicas
estatales y extra-estatales. Esta investigacin responde a fenmenos jurdicos recientes del modo en que fueran identificados por la tesis inter-institucional de la legalidad desarrollada en nuestro reciente libro Legalitys Borders (Oxford University Press, 2010), y asimismo responde a
los intereses y a la perspectiva de una versin modernizada del ciudadano ordinario de Hart. Un renovado enfoque relacional proporcionar
los medios para un anlisis ms profundo de las formas de legalidad que
surgen al interior y alrededor del cada vez ms inestable estado municipal, en la era moderna de una mayor interdependencia e interconexin
entre las disposiciones jurdicas estatales y extra-estatales.
Palabras clave:
Jurisprudencia analtica, teora inter-institucional del derecho, legalidad, positivismo jurdico relacional, anlisis conceptual contextualizado, ciudadano ordinario.
81
KEITH CULVER / MICHAEL GIUDICE
Abstract:
This article identifies and advocates one part of a renewed research
agenda for analytical legal theory: a renewed relational approach to characterization of the concept of law, following the lead set by Harts exploration of laws relation to morality, coercion, and social rules. We advocate
further descriptive-explanatory investigation of laws relation to security,
environment, and information technology, in the context of state and extra-state legal orders. This investigation is responsive to emerging legal
phenomena as identified by the inter-institutional account of legality developed in our recent book Legalitys Borders (Oxford University Press, 2010),
and is further responsive to the interests and perspective of a modernized
version of Harts ordinary citizen. A renewed relational approach will provide a means to deeper characterization of the forms of legality arising
within and around the increasingly unsteady municipal state in the modern
era of increased interdependence and interconnection amongst state and
extra-state legal orders.
Keywords:
Analytical Jurisprudence, Inter-Institutional Theory of Law, Legality, Ordinary Citizen, Relational Legal Positivism, Contextualised Conceptual Analysis.
82
PULLING OFF THE MASK OF LAW
SUMMARY: I. Introduction. II. Analytical Legal Theory: Functions, Methods, Objects and Evidence. III. Contextualised Conceptual Analysis. IV. Conclusion.
V. Bibliography.
I. INTRODUCTION
This article is a contribution to an unfortunately neglected
genre of analytical legal theory: sympathetic yet comprehensive renovation of its research agenda. This renovation
is motivated by the fact that analytical legal theory is in the
midst of an identity crisis or, at least according to its critics, it certainly ought to be. In the aftermath of Ronald
Dworkins celebrated response in The Model of Rules I to
Harts magisterial The Concept of Law, analytical legal theory appears to have lost track of its reason for being. The
historical task of analytical legal theory is perhaps best expressed by its most colourful forefather, Jeremy Bentham.
In a characteristically vivid use of metaphor, Bentham advises that we must draw aside that curtain of mystery
which fiction and formality have spread so extensively over
the Law and in criticism of Blackstone suggests that while
Law shews itself in a mask...this mask our author instead
of putting off has varnished.1 Bentham decried, for example, Blackstones approving report of the fiction contained
in the doctrine that there was no need to publicise widely
the acts of Parliament, since every man in England is, in
judgment of Law, party to the making of an Act of Parliament, being thereat by his representative.2 This might
make good legal sense, Bentham writes, but it is an affront
1 Bentham, Jeremy, A Comment on the Commentaries, Chap. II, s. 1 in The Collected Works of Jeremy Bentham, London, Great Britain, Clarendon Press, 1968, p.
124. This particular setting of Benthams remarks is of course used in an essay by
H. L. A. Hart, out of which some of our argument grows: The Demystification of the
Law in Essays on Bentham, Oxford, Great Britain, Clarendon Press, 1982, p. 23.
2 Bentham, Jeremy, A Fragment on Government, in Burns, J. H. and Hart, H. L.
A. (eds.), Introd. Harrison, R., Cambridge, Great Britain, Cambridge University
Press, 1998, p. 17.
83
to common sense. The job of legal theory is to confront the

affront: to clarify what remains obscure about law, using
morally neutral methods. While analytical legal theory continues to accept as the kernel of its task the job of pulling
off the mask of law, the reasons why this goal is sought
have varied, as have the methods deployed in pursuit of the
goal. Disputes regarding the proper motivations and methods of analytical legal theory have taken so much effort that
analytical legal theory gives the appearance of being in a
state of perpetual preparation to reach its goal, with little
demonstrable progress.
Some analytical legal theorists, such as Leslie Green, see
a mixed record of progress in response to persistent questions of legal theory: some broadening of understanding in
investigation of the relation between law and social rules,
some narrowing, some deepening of understanding in investigation of the relation between law and morality, and
some paralysis3 and a depressing point of stasis4 in
study of the relation between law and coercion. Yet Green
does not advocate any great departure from past methods
or question, and is seemingly content for analytical legal
theory to continue work on its persistent questions, no
matter the reactions of the plain lawyers who might never
be convinced of the interest of those questions.5 Few, however, are persuaded that this will result in the needed
breakthroughs. David Dyzenhaus, for example, has called
analytical legal theory a stagnant research programme,6
3 Green, Leslie, General Jurisprudence: A 25th Anniversary Essay (2005),
25 Oxford Journal of Legal Studies 4, 565-580, p. 575.
4 Ibidem, p. 573.
5 As Green explains at the conclusion of General Jurisprudence: A 25th Anniversary Essay, There have always been jurisprudence books that didnt discussion much law; now there are jurisprudence books that dont discuss much jurisprudence. Will plain lawyers find them interesting? Perhaps not; but we should
remember that the editors of the Quarterly Review urged Charles Darwin to abandon the manuscript of The Origin of Species and instead write about pigeons, because everyone is interested in pigeons. And perhaps we should be glad the advice
was ignored. Ibidem, p. 580.
6 Dyzenhaus, David, Positivisms Stagnant Research Programme (2000), 20
Oxford Journal of Legal Studies 4, 703-722.
84
whose increased precision in specification of claims and

methods has generated increasingly sophisticated defence
of increasingly limited theoretical ambitions. As Jeremy
Waldron put it, Most of those who bask in Harts glory prefer to fiddle with issues about exclusive and inclusive legal
positivism, a discussion that has been following the law of
diminishing returns since Jules Coleman initiated it in
1988, rather than address more urgent and compelling
issues.7 Others are less kind in their assessment of the sophisticated parochialism they see, charging that the enterprise of analytical legal theory risks collapsing into itself,
leaving little more than a goal and endless quibbles over
method. As Brian Simpson memorably assesses the postConcept progress of analytical legal theory,
In a sense it has been downhill all the waydownhill, that
is, from the lucidity and elegance of Herberts writing to the
unattractive elaborations of some of his critics and defenders, downhill from Herberts direct analysis of law and legal
institutions to writings about what other people have said
about what other people have written about law and legal institutions. In British military circles there was, in my time, a
bawdy monologue, much recited in pubs, which took the
form of a bestiary. One of the creatures featured in it was the
Fu-Fu Fly, which was said to fly in ever diminishing circles
until it finally vanished up its own bottom, from which secure if unsanitary location it looked out at the world with
scorn and derision. That, leaving on one side scorn and derision, is more or less the present picture in relation to much
of the secondary literature on The Concept of Law.8
Simpsons despairing picture leads us to a simple question: what next for those intrigued by analytical legal theory
yet frustrated by its offering deflections and defences rather
7 Waldron, Jeremy, Hart and the Principles of Legality, in Kramer, M. H. et.
al. (eds.), The Legacy of H. L. A. Hart: Legal, Political, and Moral Philosophy, Oxford,
Great Britain, Oxford University Press, 2008, p. 69.
8 Simpson, Brian, Herbert Hart Elucidated (2006), 104 Michigan Law Review
6, 1437-1460, pp. 1455-6.
85
than advances? One option is to adopt an entirely new research agenda, perhaps taking a direction like that pioneered by Brian Tamanaha in his socio-legal positivism.9 A
rather more promising option, we shall argue, lies in a line
of argument we began in our recent Legalitys Borders. The
inter-institutional theory of law elaborated there is an attempt to marshall overlooked insights of analytical legal
theory in support of sympathetic, yet ground-up renovation
of analytical legal theory sufficient to enable explanation of
legality in diverse forms within and beyond the state. The
argument of Legalitys Borders contributes, we believe, to
the extrication of analytical legal theory from stagnation,
and re-orientation toward pressing problems whose salience is rooted in the experience of ordinary citizens encountering diverse forms of legality in a world in which the
law-state appears prominent, yet no longer dominant. This
article continues the methodological amplification started
in Legalitys Borders, turning from the expanded range of
legal phenomena we investigated in Legalitys Borders, to
an attempt to deepen the analytical account of legality as it
is found in those diverse phenomenal contexts. Where our
inter-institutional theory was developed and deployed to
enable characterisation of prominent forms of legality in addition to the law-state, this next, complementary step is an
attempt to better understand the nature of legality in some
of those extra-state situations. This approach is not, however, straightforwardly a continued investigation of intrastate, trans-state, supra-state, and super-state legal phenomena.10 Rather, it is an attempt to deepen understanding
of prominent manifestations of legality, in whatever state or
non-state form they occur, by expansion of an approach begun by Hart, yet left behind as enthusiasms led elsewhere.
What has been called Harts relational approach to the
9 Tamanaha, Brian, Socio-Legal Positivism and A General Jurisprudence
(2001), 21 Oxford Journal of Legal Studies 1, 1-32.
10 See Culver, Keith, and Giudice, Michael, Legalitys Borders, New York,
United States, Oxford University Press, 2010, pp. 148-171.
86
concept of law sought to explain that concept by way of its

relation to social rules, morality, and coercion. The results
of this investigation are well-known, as are the results of its
newer champions. What is less well-recognized is the fact
that Hart did not suppose his choice of these particular relations was in any way exhaustive, and certainly not exclusive. Rather, these relations appeared salient from the point
of view of the ordinary educated person to whom his theory
was nominally addressed. Hart might have done otherwise.
He might, for example, have chosen the relation between
law and religion, or he might have added the relation between law and liberty, which he treated later in the celebrated Hart-Devlin debate. His choice of the focal relations
of The Concept of Law was just that: a choice, made in light
of the interests of a particular audience engaged from a
particular explanatory and historical perspective, in which
the relations explored matter not just individually, but collectively. We aim here to take up Harts approach and insight to further develop the relational approach, renewed in
the context and perspective of the citizen in the age of globalization. In that context we turn from Harts chosen relations to those relations individually and collectively salient
to our citizens perspective as a member of a state made
unsteady by dynamic mixtures of social and natural
pressures: the relations between law and security, law and
technology, and law and environment. The precise way in
which we engage those relations will produce a contextualised concept of law capable of serving as a bridge
between descriptive-explanatory approaches to legal theory,
and the empirical phenomena those theories engage.
Before we explore those relations and their product, however, we must take up the prior questions of the nature and
benefits of a renewed relational approach to the concept of
law, and the still earlier question of how such an approach
might be divorced from the dismal mis-steps which have led
to analytical legal theorys stagnation. We should aim, at
very least, to see where analytical legal theory went wrong,
87
so we might attempt to replace present stagnation with a

renovated research agenda giving fresh life to Benthams
clarity of purpose and ambition.
II. ANALYTICAL LEGAL THEORY: FUNCTIONS,
METHODS, OBJECTS AND EVIDENCE
It is probably impossible to conduct a kind of social archaeology which might reveal precisely the chain of events
leading from Benthams confident goal-setting, through
Harts inspiring re-statement of legal positivism, and on to
the present state. Fortunately, for our purposes, no such
archaeology is necessary. We are seeking less to isolate a
point of error from which the correct path might be rejoined, and rather more to understand where the debate
lost its way as a dialogue amongst scholars. With this purpose in mind, let us explore a suggestion visible in Brian
Simpsons remark that analytical legal theory has become
too little concerned with direct analysis, and too much a
debate regarding what other people have said about what
other people have written. What Waldron and Simpson see
as a distracting preoccupation with words about words,
rather than words about things, is perhaps nobler than
suggested by the Fu-Fu flys backward trajectory. It may be
that analytical legal theory has been something of a victim
of its own openness to criticism, launching too eagerly into
dialogue regarding questions adjacent to its main job. This
possibility is visible in the way debate has been drawn into
response to questions it should perhaps have rejected as
secondary questions to be answered after further efforts at
theory building, not in medias res. Entanglement in these
questions has resulted, we argue below, in a dialogue which
has left analytical legal theory simultaneously too narrow
and too broad, in need of a substantial middle we aim to
supply. More specifically, and as we have argued elsewhere,11 analytical legal theory has been distracted by at11
88
tempted response to three sometimes concurrent demands

it demonstrably need not engage. The practicality demand
asks that analytical legal theory become a kind of practical
philosophy aimed at least partially at giving tools of understanding to lawyers and private citizens a demand associated with Greens plain lawyers. The practicality demand
sometimes arrives in the company of the normativity demand, the demand that any theory of law capture laws
distinctive normative function in a wholly theoretical way,
independent of any particular system. Both practicality and
normativity demands are sometimes found in conjunction
with a third demand, the insistence that analytical legal
theory must somehow be total, exhaustive and utterly
self-sufficient, a contender for the title of best comprehensive understanding of law as a whole. These demands are
hardly unreasonable as representations of the diverse interests inquirers may bring to the activity of understanding legality; yet from the fact that these are reasonable interests,
nothing follows about whether a valuable contribution to
our understanding of legality must satisfy any one or all of
these interests. These are not the only justifiable interests
in legality a valuable contribution to understanding of legality might address altogether different interests. More importantly, there is little reason to suppose that response to
these interests must come from analytical legal theory. In
fact, as we aim to show below, the greatest value in these
demands may lie not in their value as demonstrations of
the shortcoming of analytical legal theory, but in their
pointing the way to needed renovations of analytical legal
theory sufficient to enable it to connect appropriately with
other data, arguments, and insights into legality in its
many forms.
In suggesting that analytical legal theory has become
both too narrow and too broad, we are identifying coincident problems whose origin is more an epiphenomena of a
Culver, Keith, Leaving the Hart-Dworkin Debate (2001), 50 University of Toronto Law Journal 4, 367-397.
89
dialogue than a consequence of theory-building goals or

methodological commitments and operations. The roots of
these problems in dialogue is a key to understanding how
analytical legal theory has strayed from its motivating goal
and descriptive-explanatory methods, and in turn key to
understanding how renovation of the approach might take
up and extend its original insights. In this context, what is
perhaps most surprising about the crippling narrowness
identified by Green, Waldron, Simpson and others is the
fact that the criticism is not new. In fact, it is a danger explicitly acknowledged by analytical legal theorists, particularly in handling of the phenomena of adjudication in response to the practicality demand. As Hart lists the
accusations evident in the demand, the sin imputed to
positivist conceptions of the judicial function was formalism, conceptualism, mechanical or slot machine jurisprudence or an excessive belief in the use of logic in reaching decisions.12 In early response to these charges, legal
realists took up the new methods available from psychology
and sociology to throw light on the all-too-human dimension of judges as flesh-and-blood reasoners capable of misinterpretation, political bias, and other motivations with the
potential to limit the likelihood of impartial application of
rules with determinate meaning. A few of these realists took
theorists to task for omitting other dimensions of courtroom life, as Jerome Frank did in chiding his colleagues
for forgetting juriesprudence13 which might additionally
take account of the role of juries. Yet for the most part, response to the practicality demand led to an unquestioned
focus on the pivotal role of the judge, and surrounding investigation of the nature of legal concepts treated judicial
use of legal norms as the crucible for theories of law. Later
in the century H. L. A. Harts attempt to avoid undue nar12 Hart, H. L. A., Law in the Perspective of Philosophy, in Essays in Jurisprudence and Philosophy, Oxford, Great Britain, Clarendon Press, 1983, 145-158, p.
152.
13 Frank, Jerome, Preface to the Sixth Printing in Law and the Modern Mind,
Gloucester, Massachusetts, Peter Smith, 1970. First published in 1930.
90
rowness by means of a broader theory of social rules soon

collapsed into a similar fate. Harts pioneering use of the
idea of social rules accepted and applied by generically
identified officials became enmeshed in the deep but narrow Dworkin-initiated debates regarding judicial discretion.
Here our purpose is limited to diagnosis of just where the
dialogue fell away from the purportedly central goals of analytical legal theory, so we will not fall back into discussion
of theoretical characterisation of adjudication. It is worthwhile nonetheless to quickly illustrate the deficiencies of
the judge-centred approach to the adjudicative enterprise,
beyond Franks pointed observation about the importance
of juries.
It seems to be easily forgotten that judges do not operate
in isolation. There are other players within and without the
courtroom who can influence very significantly the context
in which judges carry out their work. With the advent of the
Charter of Rights and Freedoms in Canada, for example, interveners play a role in Supreme Court decision-making
which is only now beginning to be examined by scholars
based mostly in university departments of Political Science.
The most frequent interveners are the Womens Legal Education and Action Fund and the Canadian Civil Liberties Association. It is perhaps still unclear just what effect interventions by these groups have, but evidence is growing that
there is some effect rather than none, leading sometimes
controversial scholars Ted Morton and Rainer Knopf to
identify a so-called Court Party of diverse political interests
which seek political change through politicization of the judicial process.14 Sometimes these interveners carry out
their intervention for reasons unconnected to the case at
bar, as, for example, when LEAF intervened in Andrews v
Law Society of British Columbia15 to satisfy its own interest
in bringing the court to interpret equality provisions of the
14 Morton, F. L. and Knopff, R., The Charter Revolution and the Court Party,
Peterborough, Ontario, Broadview Press, 2000, Chapter 1.
15 Andrews v. Law Society of British Columbia [1989] 1 S.C.R. 143.
91
Charter. There may be still other factors only a little beyond

the courtroom which directly influence judges task. Prosecutorial discretion, for example, determines which laws are
in fact applied and brought before judicial interpreters. Unpopular laws may become functionally obsolete as prosecutors consciously cease relying on them, all without judicial
attention to the nature of the norm expressed in the obsolete law. (We might call this desuetude in fact as opposed to
desuetude by rule of positive law, as is permitted in, e.g.,
Scotland). None of these facts are novel or unavailable to
analytical legal theorists, and these sorts of facts were top
of mind in Franks and Harts response to accusations of
formalism. Later analytical legal theorists entanglement in
narrow questions of judicial activity is almost incomprehensible in light of the availability of earlier warning against
just this sort of narrowness. Yet seen from the point of view
of theorists peppered by versions of the practicality demand
by plain lawyers and their sympathizers, it is entirely understandable that these theorists have paused to extinguish
this kind of flaring up of criticism, before pressing on with
pulling off the mask of law in a way more sensitive to the
wide range of phenomena properly examined by a descriptive-explanatory theory. The derailing of a research agenda
by the practicality demand is interesting in itself as a tale of
changing enthusiasms and misplaced effort, but what is far
more interesting about the drain of philosophical energies
toward questions about judges comes in the effect of this
work on arguments addressing more general questions
about the nature of the concept of law and legal system, the
nature of legal obligation, acceptance and fidelity to law,
and associated concepts of duty, permission, right, and so
forth.
There is a curious knock-on effect from analytical legal
philosophys deep but narrow response to accusations of
arid conceptualism. The general-level picture of the nature
of law and legal system built on the back of the Anglo-American judge-focussed picture of brute phenomena
92
overemphasises the importance of judges as legal officials

at the cost of losing touch with legal phenomena purportedly captured by these pictures, particularly with the addition of the inclusive vs. exclusive positivism debate and its
devolution into a debate over the logic of intra-theory differences. The key features of the general picture are quite simple: a legal system may be said to exist when officials accept and private citizens at least obey certain authorised
norms as non-optional standards of conduct. The validity of
authoritative norms is expressed by the collective commitment and practice of officials, whose recognition of certain
norms as authoritative provides a master rule of validity for
the system. This picture is placed under serious theoretical
strain by two empirical factors: the increasingly elusive
nature of the class of officials, and the increasingly porous
nature of nominally discrete legal systems.
Harts famous sketch of a primitive legal society under
the rule of Rex I was meant as a notional construction, yet
at some points in history reality bore a substantial resemblance to theory. The most famous of the Icelandic sagas,
Njals Saga, tells of a rudimentary legal order in which authoritative legal norms were verified by the law-speaker,
who could be consulted by disputing parties to confirm the
existence of a particular legal norm.16 The rule of recognition was maintained and passed on by the requirement that
the law-speaker provide an annual recitation of all of the
laws, so the system was evidently quite simple, or the
law-speaker possessed remarkable skills of memorisation.
Modern municipal systems of law are significantly different
from this isolated island society and its slim set of authoritative norms. As we have already observed, a wide range of
complex forces some local, and some apparently universal bear on judges making determinations of law. These
forces are poorly represented by the kinds of model officials,
nearly always model judges, depicted in general theories of
16 Njals Saga, trans. Bayerschmidt, C. F. and Hollander, L. M., Ware, Herts,
Wordsworth Editions, 1998, Ch. 97 n. 5.
93
law delivered by analytical legal theory and its interlocutors. Dworkins Hercules,17 and Jules Colemans recent development of the Swede,18 seem to resemble more closely
the kind of official envisioned in the Icelandic sagas than
the diverse range of officials evident in modern municipal
systems of law. Both Dworkins Hercules and Colemans
Swede function as model authorities at a crucial nexus
point, Dworkins in the context of a court, and Colemans
as a kind of special epistemic authority regarding the identity of legal norms in a system which is not the object of the
Swedes normative allegiance. These choices of illustrative
character each refer to a nexus point of testing for normative authority which is under threat by empirical change
from within the fragile borders of discrete municipal systems, and from without. As William Twining has argued,
globalisation and interdependence challenge black box
theories that treat nation states or societies or legal systems as discrete, impervious entities that can be studied in
isolation either internally or externally.19 Familiar examples of penetration of municipal legal systems can be seen
in the effect of international trade law on domestic legal
practices, and the demands placed on states by international law-governed human rights violations in other states.
Recently, the collapse of state isolation was illustrated in
the conduct of the United States, the global leviathan historically tempted toward an isolationist stance. The BBC
World Service reported that freshly declassified US documents show that senior American officials were warned in
plain terms of impending genocide in Rwanda, yet chose
nonetheless to vote with the remainder of the UN Security
17 Dworkins fictional judge of superhuman powers is most prominently used in
his Laws Empire, Cambridge, Massachusetts, Belknap Press of Harvard University
Press, 1986.
18 Coleman, Jules, The Practice of Principle, Oxford, Great Britain, Oxford University Press, 2001.
19 Twining, William, Globalisation and the Legal Theory, London, Great Britain,
Butterworths, 2000, p. 51. See also Twining, William, General Jurisprudence: Understanding Law from a Global Perspective, Cambridge, Great Britain, Cambridge
University Press, 2009.
94
Council to withdraw troops from Rwanda. According to the

BBC, In May [1994], a Pentagon memo cautioned against
using the term genocide to describe the Rwandan massacres, as under the 1948 Genocide Treaty, this could legally
oblige the UN, and by extension the US, to act. Genocide
finding could commit USG (US Government) to actually do
something, the memo warned.20 This example illustrates
neatly the extent to which identification of officials of a particular discrete legal system is increasingly difficult, and
perhaps more importantly, how judge-focussed accounts of
officials in general-level analytical pictures of law as an inhabitant of discrete legal systems omit demonstrably important phenomena. In the instance cited, the US plainly
feared a kind of external legal source and legal official
triggering a significant change to its legal position, and in
turn its policy position. Of course, it is entirely possible for
a critic sympathetic to a rule of recognition plus official acceptance account of the concept of law to respond in various ways including insistence that the events of the example are interesting not because they show the
interdependence of the American legal system with other
sources of legality, but because the talk of a trigger to action shows precisely how that system remains autonomous
to the extent that there is some final choice to be made by
officials in the event that the trigger is tripped. Or perhaps
one might say there was some antecedent recognition-carrying choice to join a treaty which foresaw just this sort of
event. Yet as we argued at length in Legalitys Borders,
while this style of response might be effective against occasionally encountered challenges to the explanatory power of
the orthodox view, the rise of intra-state, trans-state, supra-state and super-state legality is so extensive and so
pervasive that a theory which fails to engage those novel
phenomena risks losing any claim to being meaningfully
20 BBC World Service Tuesday 21 2001 August, Africa page, US Warned about
Rwanda Genocide. http://news.bbc.co.uk/2/hi/africa/1502708.stm Accessed 14
December 2011.
95
descriptive-explanatory. Instead of pulling off the mask of

law, such an approach risks becoming an apologist for fictions.21 Worse yet, a state-centric rule of recognition plus
official acceptance approach risks embracing a fiction
which Hart saw and railed against in Austin the misleading view that law-states are or ever have been autonomous
in the sense required by Austins uncommanded commander. The deep but narrow gains made in exploration of
social rules appear to have given rise to a general jurisprudence whose attempts at response to comprehensiveness
and normativity demands lead them to a picture of law so
broad and abstract that it is no longer clear just how they
are descriptive-explanatory, and still less clear how these
investigations are connected to the remainder of Harts
original investigation of the relations between law, social
rules, morality and coercion, all in aid of a richer picture of
law developed from the perspective of the ordinary citizen.
The results may be satisfying to some,22 bravely holding
their analytical course and ignoring calls to do otherwise.
Yet it seems to us that this amounts to bravely holding to a
course which has been subtly altered by the force of a dialogue which has left its original goal, motivation and audience behind, forgetting Benthams and Harts urging clarity
in understanding law as a social institution, practice and
experience. We have tried to indicate above some ways in
which narrowness and overbreadth are evident in the careers of devices such as Hercules, the Swede, and the Uncommanded Commander. Still more pressing shortcomings
are evident in the fact that even if all were well with investi21 We are not alone in this assessment. See, for example, Tamanaha, Brian, A
General Jurisprudence of Law and Society, Oxford, Great Britain, Oxford University
Press, 2001; Twining, William, General Jurisprudence: Understanding Law from a
Global Perspective, Cambridge, Great Britain, Cambridge University Press, 2009;
Von Daniels, Detlef, The Concept of Law from a Transnational Perspective, Surrey,
Great Britain, Ashgate Publishing, 2010; and Santos, Boaventura de Sousa, Toward a New Legal Common Sense: Law, Globalization, and Emancipation, 2nd edn.,
London, Great Britain, Butterworths, 2002.
22 Perhaps those who suppose they are the Darwins in Greens metaphor. See
Green, Leslie, op. cit., n. 5.
96
gation of law and social rules and morality, there is still the
stasis observed by Green with respect to coercion, and
worse yet, as Green writes, What we have not had is convergence.23 The absence of convergence in investigation of
social rules, morality and coercion is particularly troubling
in light of the contingent relevance of these relations to an
understanding of law satisfactory to inquirers interests.
The relation of law to those particular relations was judged
important by Hart in his deepening of a picture of those
features of legality particularly salient to ordinary citizens.
We have already suggested that Hart might have chosen
other relations to investigate, and in other work did so.
What does it say about the progress of analytical jurisprudence that it has made little headway toward an integrated
picture of just three of the contingent relations relevant to a
citizen whose interests may well change? As we suggested
in Legalitys Borders, we would do well to conduct legal theory from the perspective and interests of the ordinary citizen who travels in a globalizing world whose effects are felt
at home and when travelling. The interests of that citizen
are at best partially satisfied by the partial advances of Hercules, the Swede and friends in a subset of the relations between legality and its wider social context. The citizen of Legalitys Borders demands more: a wider relational approach,
with better integration, resulting in a concept of law whose
claim to be descriptive-explanatory is grounded in a clear
account of the responsiveness the theory owes to social evidence, and fidelity to its goal of clarification such that practicality, normativity, and comprehensiveness demands can
each be given their proper space and no more. In a phrase,
the ordinary citizen of the 21st century demands a
contextualised conceptual analysis whose deployment of
Harts relational approach renews analytical jurisprudence
in a way which overcomes its narrowness and overbreadth
while acknowledging the limits of analytical jurisprudence
23
Green, Leslie, op. cit., n. 3, p. 575.
97
and the value of other theoretical and empirical approaches

to law.
III. CONTEXTUALISED CONCEPTUAL ANALYSIS
In this paper we limit ourselves to sketching the nature
and role of the kind of contextualised conceptual analysis
we suppose needs to be developed, leaving its full development to future work. This analysis is a continuation of the
approach begun in Legalitys Borders. There we sought to
develop a bottom-up phenomena-sensitive explanation in
which our account of legality as agglomerations of institutions of law operated as kind of radar or detection device.
Our inter-institutional account is capable of detecting manifestations of legality in both familiar areas of social life
such as the law-state, and in less familiar areas of transand super-state legality. That account is intended as one
kind of contribution to the broadest possible picture of law,
surfacing for further examination several instances of legality given insufficient attention by contemporary analytical
jurisprudence. The difference between our inter-institutional theory and our present effort is usefully expressed by
way of a spatial metaphor: where our inter-institutional
theory enables detection of peaks and promontories of legality on the terrain of social life, this paper and
contextualised conceptual analysis aim to enrich our account of those promontories by beginning to examine the
relations between legality as found there, and other social
forces salient from the perspective of our ordinary citizen
who travels and finds legality in the law-state, intra-state,
trans-state, and super-state situations we have previously
characterised. The resulting conceptual analysis is
contextualised in three ways. First, in our advancing the relational approach toward previously underexamined relations between law and other social phenomena, we aim to
be responsive to social phenomena salient from the point of
view of the ordinary citizen who travels, undertaking what
98
in Legalitys Borders we called re-socializing a descriptive-explanatory view of law. Contextualised conceptual

analysis aims in this way to fill the hole left by overly narrow theorizing generating a distorted general jurisprudence
whose claims to descriptive-explanatory relevance are at
best difficult to credit, as discussed above. Second, our use
of an expanded relational approach is intended to contribute to contextualised conceptual analysis by conducting a
history-sensitive style of analysis, not just narrowly sensitive to legal history or the history of salient social phenomena in relation to legality, but to the fact that our concepts
have histories as they have evolved in our understanding
and as reflections of practice. Third, our expanded relational approach is contextualised in the sense that it follows and is complementary to our inter-institutional analysis, and is situated as one component of a broader
conceptual-explanatory context, in which we claim we are
not liable for direct response to practicality, normativity
and comprehensiveness demands, yet we accept that we
are reasonably held liable for the intersection of our view
with theories responding directly to those demands. It will
be characteristic of this expanded relational approach that
it produces a concept of law which is vague, yet tolerably
vague, as the story of laws relations to salient social forces
is a story of a dynamic concept constantly under renewal in
response to novel phenomena.
The remainder of this paper is devoted to exploration of
these three aspects of contextualised conceptual analysis,
motivated by the shortcomings of analytical theory, identified as a kind of narrowness in engagement of social phenomena, mistaken inference to generality from that limited
set of data, and failure to situate analytical theory relative
to other reasonable demands made of legal theory considered in toto and not just in its pre-censorial, clarificatory
and expository mode. The very last part of this paper will
discuss the particular relations we propose to engage in future work as a way to broaden the phenomenal basis of an99
alytical legal theory. It will be useful to approach that discussion via further consideration of the sense in which
contextualised conceptual analysis is historically sensitive
and self-consciously a partial contribution to the broadest
possible picture of legality, by way of beginning to respond
to the problem of unselfconscious abstractness evident in
analytical legal theory to date.
3.1 History
Analytical jurisprudence has long attempted to take history seriously in various ways, but those attempts have
been arguably unselfconscious: history has mattered as an
input to jurisprudential reasoning, not as an aspect of it.
We propose a broader role for history in contextualised conceptual analysis, understanding that the concept of law
and subordinate concepts have a more than etymological
history, and that this history matters to the way we understand them now, particularly when aiming at something
like a universal concept of law applicable to all societies at
all times. Let us explain this claim in situ. One of the problems which has vexed analytical legal philosophy is understanding of the nature of legal obligation how laws obligate, why subjects obey laws, and so forth. A typical
analytical approach to the question of the nature of legal
obligation is to examine first the meaning of obligation, and
then to rely on that rich understanding of the meaning of
the concept to try to understand the place of that concept
in the surrounding web of legal concepts such as duty and
permission. Lon Fuller, for example, writes of fidelity to law
as a kind of translation of the attitude of acceptance of law,
and builds the notion of faith, the Latin fide, into the account of how law presents itself to subjects in a way which
secures their acceptance.24 Hart contributes the observation
that obligation seems to have its roots in the Latin ligare,
24 Fuller, Lon, The Morality of Law, 2nd edn., New Haven, United States, Yale
University Press, 1969, pp. 38-41.
100
to bind, and works from there toward ideas developed in a

better known, and likely better worked fashion by Joseph
Raz, in his understanding of legal norms as a special variety of exclusionary reasons.25 Yet all of this seems to be a
half-step where a full step is needed. Some concepts have
deeper historical roots than others, and the changes some
concepts undergo over the course of their history may reflect special characteristics of that concept. Those special
characteristics may in turn require special handling in any
understanding of that concept and the conditions under
which it is applied. Legal obligation seems to us to be one of
these concepts whose historical career presents variations
in meaning which are more than surface deep.
Understanding obligation as rooted in ligare, a binding
tie, may capture a subtly yet importantly different dimension of the practice thought to be represented by descriptive-explanatory accounts of lived experience which inform
the concept. To understand choice of conduct consistent
with legal obligation as a choice to which one is bound is to
invoke a metaphor which captures a sense of helpless involuntariness, and a constraint imposed from without. It
may well be true that this understanding of legal obligation,
built out of Roman use of obligatio generated from ligare,
represents adequately the historical development of one use
of the term and one contribution of experience to the
shared concept. Yet this understanding may be unduly narrow, and a poor basis for an understanding of the concept
of legal obligation as part of a broader understanding of adjoining legal concepts of permission, duty, and so forth. In
25 See Hart, H. L. A., The Concept of Law, 2nd edn., Oxford, Great Britain, Clarendon Press, 1994, p. 87; Commands and Authoritative Reasons in Essays On
Bentham, Oxford, Great Britain, Clarendon Press, 1982, 243-268. See also Raz, Joseph, The Authority of Law, Oxford, Great Britain, Clarendon Press, 1979; Authority, Law and Morality, (1985) 68 The Monist, pp. 295-324; Authority and Justification (1985) 14 Philosophy and Public Affairs 3, pp. 3-29; Postscript to the Second
Edition: Rethinking Exclusionary Reasons in Practical Reason and Norms 2nd
edn., Princeton, New Jersey, Princeton University Press, 1990, 178-199; Explaining Normativity: On Rationality and the Justification of Reason (1999) 12 Ratio,
354-379; Engaging Reason, Oxford, Great Britain, Oxford University Press, 2000.
101
Bertram Wyatt-Browns essays on the ideal of honour in the

antebellum American South,26 he remarks on the arguments made by pre-revolution Americans seeking some
form of self-rule that taxes sought by the English crown
were supposed to be considered as free gifts, and as such
could be offered only by the elected representatives of those
who possessed the property thus given.27 As Edmund Morgan reads Wyatts argument, In depriving them of the sole
right to tax themselves, the right to offer whatever they gave
in taxes as a free gift to their king, Parliament deprived
them of honour28 and collapsed the relationship which
served as the precondition of a legal obligation to pay tax.
These pre-revolution Americans seem to have understood
legal obligation as something quite different from an externally imposed tie that binds. For these Americans, legal obligation sprung from an internally generated sense of honour as the autonomously chosen, internally generated
entering into a liberty-limiting obligation. Any externally
coerced compliance to some norm was to be resisted as
strongly as possible.
We do not aim to suggest that either of these examples
ligare or free gifts illustrates the uniquely correct understanding of the concept of legal obligation, yet at the
same time we do not wish to discard either one as an historical anomaly. It may be possible to integrate these ideas
into some part of the contemporary analytical account of legal obligations as a special kind of reason, excluding other
reasons for action, and accepted in that role for any reason
at all, as Hart supposes, or because the exclusionary reason expresses some normally justified course of action, as
Raz supposes. Yet even if these ideas cannot be incorporated into the contemporary account, it is far from clear
26 Wyatt-Brown, Bertram, The Shaping of Southern Culture: Honour, Grace, and
War, 1760s-1880s,Chapel Hill, North Carolina, University of North Carolina Press,
2000. See also Edmund Morgans review, The Price of Honour, New York Review
of Books XLVIII, No. 9, May 31, 2001, pp. 36-38.
27 Morgan, ibidem, p. 37.
28 Idem.
102
that we are justified in discarding one as anomalous or an

outlier and simply picking the other, perhaps ligare, as the
etymologically nearest cousin to what we mean by legal obligation today. At least two options are opened by the historical wariness we think ought to be characteristic of an
appropriately contextualised conceptual analysis. We might
choose to modify the account of legal obligations as
exclusionary reasons accepted by legal subjects for certain
kinds of reasons, choosing to understand historical accounts of the binding force of legal obligations as indications of contextually variable perceptions of what counts as
a normal reason for action, capable of bearing special force
once recognised as a legal reason for action. Or we might
choose to revisit the adequacy of this account of legal obligations. We might recognise that the contemporary social
situation reflected in doctrines such as Razs exclusionary
reasons approach to legal obligation is itself a historical
product concerned with a concept which is evolving
through its use. Or we might imagine still further options
for response. Yet doing nothing about the deep historical
roots of these varying ideas about legal obligation would be
a mistake which we ought to resist on the grounds that doing so would lead into acceptance of an undesirable descriptive-explanatory dissonance: we ought not to accept
those pictures of legal obligation which capture some aspect of contemporary practice while making it difficult or
impossible for us to understand at the same time the practice of legal obligation in these historical contexts. Laws are
conventional products, and the conventions which produce
them have histories which are part of a full understanding
of the product. Historically-sensitive, perhaps historically
inclusive theorising may result in an understanding of the
concept which is capacious and vague as it incorporates
quite different societies and times, but that is scarcely an
objection if possession of that concept enables us to
understand better the very wide variety of ways of living life
under law, avoiding the well-known trap of undue
103
narrowness giving rise to Hercules, the Swede, and the

Fu-Fu fly.
3.2. The Contextualised Concept
in Legal Theory as a Whole
Having gone this far to emphasise the importance of history in understanding the meaning of concepts, we must be
clear that our contextualised approach to the concept of
law requires further methodological innovation beyond recognition that concepts do not arrive history-free. The second element of contextualised conceptual analysis is derived from an argument Culver made in a characterisation
of H. L. A. Harts picture of the concept of law.29 That argument attributed to Hart what Culver called relational legal
positivism, a method of concept-reconstruction which begins from what might be called a folk understanding of the
concept of law, and builds a dynamic, changeable concept
out of its relations to other phenomena. This concept of law
is a kind of reflection of changes beneath it, a concept modified according to the results of ebbing and flowing changes
in life under law. This approach, which Hart once recognized as hermeneutic,30 has been taken up by some analytical theorists, including, for example, Jules Coleman,
whose support of a pragmatic method for analytical legal
philosophy extends to acceptance of the need for our concepts to be revisable, and continuous with the social sciences, while not being supplanted by them.31 Colemans
powerful pragmatism is, however, hampered by a lack of
clarity about its place relative to the normativity, practicality, and comprehensiveness demands. His assertion that
Culver, Keith, op. cit., n. 11.
See MacCormick, Neil, H. L. A. Hart, 2nd edn., Oxford, Great Britain, Oxford
University Press, 2008, pp. 46 and 184; and Hacker, P. M. S., Harts Philosophy of
Law, in Hacker, P. M. S. and Raz, Joseph (eds.), Law, Morality, and Society: Essays
in Honour of H. L. A. Hart, Oxford, Great Britain, Clarendon Press, 1977.
31 Coleman, Jules, op. cit., n. 18, p. 201. See also Leiter, Brian, Naturalizing Jurisprudence, Oxford, Great Britain, Oxford University Press, 2007, Part II.
29
30
104
the pragmatic method is always responsive to all of these

concerns seems to leave his theory aiming to capture absolutely everything, which is plainly at odds with the nearly
total absence of empirical data in his analysis, and his use
of abstract devices such as the Swede, nominally independent of particular legal systems. Contextualised conceptual
analysis aims to improve upon Colemans pragmatism by
being clear about its sources, the limits of its aims, and
clear about the reasons why those aims are properly sought
using the descriptive-explanatory relational approach. This
analysis recognises the incompleteness of the relational approach as a way of understanding social phenomena, and
rests its value on the clarificatory power it can bring to our
otherwise potentially confused understandings of the concept of law and subordinate legal concepts, all contributing
to the widest possible understanding of the promise and
limits of law as a mode of organising social life. This limits
the relational approach and contextualised concept of law
to a role which is admittedly impractical and non-comprehensive. Contextualised conceptual analysis recognises that
it is not an attempt to provide the best comprehensive theory of law as a whole, and remains agnostic regarding the
character of non-philosophical investigations best suited to
carrying out empirical and local-level theoretical investigations which inform the project of conceptual analysis. Investigations in sociology and anthropology of law, for example, are welcomed by contextualised conceptual analysis,
yet this analysis holds no particular view regarding the
methodology of sociology or anthropology. This analysis is
nonetheless cautiously universal about the concept of law,
in the sense of attempting to clarify the concept as a universally applicable concept. The clarificatory power of the
contextualised concept of law is derived from two dimensions of its method, together with a careful admission
about the limits of that analysis. Contextualised conceptual
analysis understands through its sensitivity to history that
while legal philosophy can be concerned with general or
105
universal concepts, life under law is fundamentally local.

Historical sensitivity allows contextualised conceptual analysis to recognise the historical variability of legal practices.
This might be called the vertical richness of contextualised
conceptual analysis, matched by the horizontal reach of the
relational method, whose focus in Harts hands on relations
between law and coercion, morality, and social rules required us to look broadly at the function of what is said to
be law and persuades us away from excessive reliance on
observed data from particular situations which are local
both to a particular legal culture and to a particular time.
The relational method reminds us to look beyond the massive figures of judges in Anglo-American legal life to other
officials and other experiences of life under law, and to
avoid moving too quickly from local experience, particularly
judicial experience and constitutional experience, to general
conclusions about the nature of the concept of law. In this
way contextualised conceptual analysis occupies a middle
ground which admits the importance of local experience,
yet attempts also to widen our understanding of that experience. This analysis reminds us of the importance of history to concept-formation, and the importance of working
broadly and thoroughly with local investigations to
constantly re-inform the concept of law. The resulting
concept is contextualised because it is inclusive of a wide
range of experience of life under law, yet it is admittedly
theoretical and practically thin because it is a reflection of
empirical investigations and locally situated conceptual
investigations (into, e.g., interpretive principles accepted by
judges in the context of Canadian Charter adjudication).
The historically-sensitive, partial contribution to general
jurisprudence made by the contextualised concept may ultimately be a vague concept whose edges are not easily delineated. This vagueness is, however, both tolerable and expected of a concept emerging from relational study, and
serving as a constantly revised conceptual backdrop for understanding of life under law, a prelude or constant accom106
paniment to normatively committed and social scientific investigations of law. Unsurprisingly, then, given the
emergent, reconstructed character of this clarificatory concept, it is not the only one. It is the particular result of one
method calculated to reveal law as it is for the sake of the
ordinary citizen who travels, pulling off the mask of law not
by being practical or comprehensive, but by filling the gap
between practically oriented, often judicially-focussed legal
philosophy, and general accounts of the logic of norms.
Contextualised conceptual analysis is then truly a middle
way, its virtues identifiable only through its co-operation
with accompanying investigations. It is nonetheless a powerful middle way, since it opens the possibility of analytical
legal philosophy coming to terms with changes in social life
judged by our observers perspective in a world where the
importance of laws morality, social rules, and coercion is
changed, and changing. In the pluralist world inhabited by
our citizen who travels, laws relation to morality is far more
varied and complex than the relation claimed by Lord
Devlin in the Hart-Devlin debate as he celebratedly claimed
that a sense of sin in a society where morality and religion
are conjoined is essential to effective administration of
criminal law.32 The social rules to which our ordinary citizen is subject may be no more logically complex than those
of Harts generation, but the diversity of their sources and
their force is ever-more bewildering in the globalizing world.
And while the relation between law and coercion remains of
fundamental interest as a persistent question, the way in
which the ordinary citizen experiences this question has
likely changed, as our citizens mobility enables both
evasion and sudden meetings with coercion.
32 As a judge who administers the criminal law and who has often to pass sentence in a criminal court, I should feel handicapped in my task if I thought that I
was addressing an audience which had no sense of sin or which thought of crime
as something quite different. Devlin, Patrick, The Enforcement of Morals, Oxford,
Great Britain, Oxford University Press, 1965, p. 4.
107
3.3. New Relations

The contextualised concept of law is in some ways a surprisingly thin concept. In aiming to be historically sensitive and universal, it will be very broad; and in aiming to
be constantly responsive to new phenomena so constantly
under revision, its edges are likely to be vague and it is
unlikely to be usefully put to service as a test for legality.
Fortunately it need not be concerned with providing a test
for legality, because such a test is provided by the complementary or companion inter-institutional view. The contextualised concept of law can rest its claim to an audience on the quality and relevance of its elucidation of relations between legality and social phenomena relevant to
the inquirers. The choice of relations to investigate individually and in their convergence is accordingly crucial to the
final value of the contextualised analysis of the concept. At
the same time, the choice of relations to investigate may
have unanticipated effects as the goal of clarification in aid
of normative reform may once again take pride of place
amongst the virtues of an analytical theory of law, generating fresh debates on method, application, data, and inference, pulling us away from worn-out debates, and toward
new questions.
Since this paper is largely agenda-setting, we will confine
ourselves in this closing section to a few words of justification regarding the relations we propose to investigate in future work. The relations we have chosen are those which
are relevant to our ordinary citizen who travels, and in addition to travelling in a touristic way, sometimes migrates,
and sometimes invests in various ways without migrating.
This observer looks to legality as a tool and as an indicator
of a certain kind of desirable stability, an increasingly valuable characteristic as social and environmental forces give
rise to what we will call the era of the unsteady state.
We propose to begin satisfaction of this observers interests at a point of overlap with Hart, in the relation between
108
law and coercion, but at a particular point of special relevance to our citizen who travels: the relation between law
and security, in the particular context of the law-state and
the question of how to characterize the identity of a legal
system or other forms of legality over time. Familiar examples of the challenge facing characterisation of law and security are easily found, from the decades-long absence of a
law-state in Somalia, to the uncertain legal status of
Kosovo, and on to the lengthy and varied period of uncertainty seen in the set of uneven regime and system-transitions in the context of the Arab Spring of 2011. Questions
regarding the relation between law and security became
particularly pronounced in that context as the rapid
changes seen in countries such as Egypt and Tunisia in
spring were not followed by equally rapid changes in other
countries. Spring wore on into summer, and with it the Libyan insurrection, whose leaders at length gained recognition as states recognized the former rebels as the legitimate
government of Libya.33 More recently questions of state
identity, legality, and legitimacy have arisen in Syria as ongoing unrest verges on civil war.34 At the same time a different kind of security threat to the identity and continuity of
the state has become visible as the European sovereign
debt crisis has threatened the viability of Greece and other
states with an unsustainably high debt-to-GDP ratio,35 rais33 For general analysis see Anderson, Lisa, Demystifying the Arab Spring:
Parsing the Differences between Tunisia, Egypt and Libya (2011), 90 Foreign Affairs 3, 2-7.
34 Journalistic analysis of developments in Syria includes, for example, Syrias
Uprising: Growing Steadily Less Peaceful The Economist online, December 19,
2011, available online at http://www.economist.com/node/21541978. Accessed
January 14, 2012.
35 Eurostat, the Directorate-General of the European Commission responsible
for provision of European-level statistical information, reports that In 2010, 14
Member States reported debt to GDP ratios over the reference value of 60%. Greece
recorded the highest debt ratio with 144.9%, followed by Italy with118.4%.
Structure of government debt in Europe in 2010, Eurostat Issue number
68/2011. Available online at: http://epp.eurostat.ec.europa.eu/portal/page/portal/product_details/publication?p_product_code=KS-SF-11-068 . Accessed January
14, 2012.
109
ing questions regarding the identity and continuity of the

law-state, potentially especially far-reaching questions as a
new European treaty may be negotiated to further mutualise economic planning, coincidentally reducing legal
autonomy.
The ordinary citizens concern with the relation between
law and security is increasingly accompanied by concerns
regarding the relation between law and environment. From
the Asian tsunami of 2004 to the European heatwave of
200336, and on to news of pervasive overfishing of global
fish stocks,37 the ordinary citizen is inundated by news of
environmental disaster and advocacy of substantial response on a state, supra-state and super-state basis. From
climate change response to fish stock management, it is increasingly clear that humans conduct in their environments has effects which cannot be contained within the
boundaries of states of which those humans are citizens.
The urgent need for effective response to environmental
challenges raises fresh questions regarding the meaningful
independence of states as a kind of super-state environmental law may become a precondition for further existence
of states. Further questions are arising regarding the identity and continuity of states, as some island states such as
the Republic of Kiribati may soon disappear beneath the
waves under rising ocean waters.38 We are familiar with the
phenomenon of conquest but immersion seems entirely
new.
Finally, we intend to investigate a third relation, between
law and technology, a relation with profound implications
for conceptions of agency and autonomy in the context of
36 Stott, P. A. et al., Human contribution to the European heatwave of 2003,
(2004) 432 Nature 2, pp. 610-614.
37 Anticamara, J. A. et al., Global fishing effort (1950-2010): Trends, gaps, and
implications, 107 Fisheries Research, No. 1, 131-136.
38 Sanderson, Katharine, Too late to save Pacific island nation from submersion, Nature news online June 6, 2008. Available online at: http://www.nature.com/news/2008/080606/full/news.2008.880.html. Accessed Janaury 14,
2012.
110
the law-state and other forms of legality. Limiting ourselves

for the moment to the rise of the new information communication technologies, we note frequent discussion in the
press regarding cyber-attacks supported by one government
against the banking, military, and social information systems of other governments, all occurring within the supernational infrastructure of the internet.39 Is the internet now
its own form of legality, perhaps like we might want
super-state environmental law to be, a kind of substrate for
all further legal activities by law-states? In this context we
may also see virtual personalities and living wills, potentially creating artificial legal agents, perhaps unworrying at
the level of the individual citizen, but rather more interesting from the point of view of legal theory as we consider the
implications of virtual life. Consider, for example, the possibility that the Republic of Kiribati, submerged beneath the
waves as a consequence of global failure to mitigate climate
change, might emerge again as a virtual state operating as
a law-state in all but territorial fashion, a persistent legal
culture even in the absence of geographic boundaries,
much like the social culture of the forcibly evicted and still
angry former inhabitants of Diego Garcia in the British Indian Ocean Territories.40 Might our future see the
impossibility of the demise of law-states, instead simply
their transformation?
IV. CONCLUSION
It is almost misleading to offer a conclusion to this paper,
since it is above all a re-launch of a renewed research
agenda for analytical jurisprudence, advocating a return to
an expanded version of Harts descriptive-explanatory, rela39 See, for example, the New York Times series on cyberwar. Available online at
http://topics.nytimes.com/topics/features/timestopics/series/cyberwar/index.html.
Accessed January 14, 2012.
40 Vine, David, Island of Shame: The Secret History of the U.S. Military Base on
Diego Garcia, Princeton, New Jersey, Princeton University Press, 2011.
111
tional approach to characterisation of the concept of law. In

bringing this approach to bear on laws relation to security,
environment and technology, we aim simultaneously to offer a contextualised concept of law relevant to the interests
of contemporary inquirers, and to redirect explanatory efforts away from an excessively narrow focus on the phenomena of adjudication, toward a broader range of phenomena from which a broad concept of law may be inferred,
adequately reflective of the full range of relevant empirical
phenomena. This contextualised concept of law may additionally put old problems in fresh perspective, encouraging
us to see, for example, the questions of continuity and
identity as pressing matters. A contextualised concept of
law responsive to the concerns of our modern citizen may
escape from the illusion fostered by stated-centred jurisprudence that the continuous, identity-stable state is legalitys standard manifestation, whose interior needs further
investigation while its exterior shell its continuity of identity may be taken for granted. As we begin to develop a
contextualised concept of law via our investigation of new
relations, the question of identity and continuity of legal orders gains a renewed prominence it has not had since the
early 1970s, when John Finnis investigated identity and
continuity in decolonization of Africa, in his Revolutions
and the Continuity of Law.41 Pessimists might view this as
analytical legal theory coming full circle in its enthusiasms,
looking once again to the phenomena of revolution as a reason to engage the persistent question of the identity and
continuity of legal order. Optimists like us see the possibility of new and better answers to the persistent question,
which is not just persistent but of new importance in the
new era of the state made unsteady not just by revolution
but by other developments as well. The relation of law to
coercion, explored by Finnis as the relation of law to secu41 Finnis, John, Revolutions and Continuity of Law, in Philosophy of Law, Collected Essays: Volume IV, Oxford, Great Britain, Oxford University Press, 2011,
407-35.
112
rity, remains important to our worldly citizen, albeit increasingly at the scale of a given legal order, and not just
legal obligation by legal obligation in a single-law state. As
the states role as a law-giver is increasingly shared with
other legal orders, and those orders and the states role in
the community of legal orders is affected by changing relations between law and environment, and law and technology, the question of the identity and continuity of legal orders is justifiably regarded as demanding attention at least
equal to that historically given the persistent questions of
the nature of municipal legal systems and legal obligation
within them, and the relation of legality in that context to
morality, coercion, and social rules. If we succeed in carrying out the research agenda we have advocated, via a renewed relational approach taking the persistence of legal
orders as a major challenge to descriptive-explanatory legal
theory, we may contribute to the destruction of the fiction
of the centrality of the state to the nature of legality, and in
that way, contribute one step to pulling the mask off law.
V. BIBLIOGRAPHY
ANDERSON, Lisa, Demystifying the Arab Spring: Parsing the
Differences between Tunisia, Egypt and Libya
(2011), 90 Foreign Affairs 3.
ANTICAMARA, J. A. et al., Global fishing effort (1950-2010):
Trends, gaps, and implications, 107 Fisheries Research, No. 1.
BENTHAM, Jeremy, A Comment on the Commentaries, Chap.
II, s. 1, in The Collected Works of Jeremy Bentham,
London, Great Britain, Clarendon Press, 1968.
, A Fragment on Government, in BURNS, J. H. and
HART, H. L. A. (eds.), introd. Harrison, R., Cambridge,
Great Britain, Cambridge University Press, 1998.
COLEMAN, Jules, The Practice of Principle, Oxford, Great Britain, Oxford University Press, 2001.
113
CULVER, Keith, Leaving the Hart-Dworkin Debate (2001),

50 University of Toronto Law Journal 4.
and GIUDICE, Michael, Legalitys Borders, New York,
United States, Oxford University Press, 2010.
DEVLIN, Patrick, The Enforcement of Morals, Oxford, Great
Britain, Oxford University Press, 1965.
DWORKIN, Ronald, Laws Empire, Cambridge, Massachusetts, Belknap Press of Harvard University Press,
1986.
DYZENHAUS, David, Positivisms Stagnant Research Programme (2000), 20 Oxford Journal of Legal Studies 4.
FINNIS, John, Revolutions and Continuity of Law, in Philosophy of Law, Collected Essays: Volume IV, Oxford,
Great Britain, Oxford University Press, 2011.
FRANK, Jerome, Preface to the Sixth Printing, Law and the
Modern Mind, Gloucester, Massachusetts, Peter
Smith, 1970. First published in 1930.
FULLER, Lon, The Morality of Law, 2nd edn., New Haven, United States, Yale University Press, 1969.
GREEN, Leslie General Jurisprudence: A 25th Anniversary
Essay (2005), 25 Oxford Journal of Legal Studies 4.
HACKER, P. M. S., Harts Philosophy of Law, in HACKER, P.
M. S. and RAZ, J. (eds.), Law, Morality, and Society:
Essays in Honour of H. L. A. Hart, Oxford, Great Britain, Clarendon Press, 1977.
HART, H. L. A., Commands and Authoritative Reasons, in
Essays on Bentham, Oxford, Great Britain, Clarendon
Press, 1982.
, Law in the Perspective of Philosophy, in Essays in
Jurisprudence and Philosophy, Oxford, Great Britain,
Clarendon Press, 1983.
, The Concept of Law, 2nd edn., Oxford, Great Britain,
Clarendon Press, 1994.
114
, The Demystification of the Law, in Essays on Bentham, Oxford, Great Britain, Clarendon Press, 1982.
LEITER, Brian, Naturalizing Jurisprudence, Oxford, Great
MACCORMICK, Neil, H. L. A. Hart, 2nd edn., Oxford, Great Britain, Oxford University Press, 2008.
MORGAN, Edmund, The Price of Honour, New York Review
of Books, XLVIII, No. 9, May 31, 2001, pp.
MORTON, F. L. and KNOPFF, R., The Charter Revolution and
the Court Party, Peterborough, Ontario, Broadview
Press, 2000.
RAZ, Joseph, Authority and Justification (1985), 14 Philosophy and Public Affairs 3.
, Authority, Law and Morality (1985), 68 The Monist.
, Explaining Normativity: On Rationality and the Justification of Reason (1999), 12 Ratio.
, Engaging Reason, Oxford, Great Britain, Oxford University Press, 2000.
, Postscript to the Second Edition: Rethinking Exclusionary Reasons, in Practical Reason and Norms 2nd
edn., Princeton, New Jersey, Princeton University
Press, 1990.
, The Authority of Law, Oxford, Great Britain, Clarendon Press, 1979.
SANDERSON, Katharine, Too late to save Pacific island nation from submersion Nature news online June 6,
2008. Available online at: http://www.nature.com/
news/2008/080606/full/news.2008.880.html.
Accessed Janaury 14, 2012.
SANTOS, Boaventura de Sousa, Toward a New Legal Common
Sense: Law, Globalization, and Emancipation, 2nd edn.,
London, Great Britain, Butterworths, 2002.
115
SIMPSON, Brian, Herbert Hart Elucidated (2006), 104 Michigan Law Review 6.
STOTT, P. A. et al., Human contribution to the European
Heatwave of 2003 (2004), 432 Nature 2.
TAMANAHA, Brian, A General Jurisprudence of Law and Society, Oxford, Great Britain, Oxford University Press,
2001.
, Socio-Legal Positivism and A General Jurisprudence (2001), 21 Oxford Journal of Legal Studies 1.
TWINING, William, General Jurisprudence: Understanding
Law from a Global Perspective, Cambridge, Great Britain, Cambridge University Press, 2009.
, Globalisation and the Legal Theory, London, Great
Britain, Butterworths, 2000.
VINE, David, Island of Shame: The Secret History of the U.S.
Military Base on Diego Garcia, Princeton, New Jersey,
Princeton University Press, 2011.
VON DANIELS, Detlef, The Concept of Law from a Transnational Perspective, Surrey, England, Ashgate Publishing.
WALDRON, Jeremy, Hart and the Principles of Legality, in
KRAMER, M. H. et al. (eds.), The Legacy of H. L. A. Hart:
Legal, Political, and Moral Philosophy, Oxford, Great
WYATT-BROWN, Bertram, The Shaping of Southern Culture:
Honour, Grace, and War, 1760s-1880s, Chapel Hill,
North Carolina, University of North Carolina Press,
2000.
116
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
DESCRIPTIVE JURISPRUDENCE
Pavlos ELEFTHERIADIS*
Resumen:
Hart sostuvo que la jurisprudencia analtica es descriptiva y general pero
no elabor una distincin explcita entre las preguntas conceptuales que
son principalmente tericas y otras que son principalmente prcticas. En
la explicacin de Hart, la jurisprudencia analtica busca cierto tipo de
claridad acerca de la nocin de derecho y las dems ideas bsicas tal
cual ellas ocurren en nuestra experiencia ordinaria. En el Post Scriptum,
l explica que la jurisprudencia es el estudio terico o cientfico del derecho como un fenmeno social, el cual puede, en ocasiones, basarse en la
deliberacin moral, pero tpicamente no lo hace. Esto puede contrastarse
con las preguntas prcticas acerca de como las personas dirigen su
propia vida. Llamo esta tesis descriptivismo. Esta ha sido una posicin
muy influyente en la jurisprudencia analtica y ha sido defendida por los
seguidores de Hart hasta hoy. En este ensayo despierto dudas sobre su
coherencia. Primero, no creo que Hart defendiera esta tesis de forma
consistente. Segundo, no creo que la posicin haya sido defendida de forma adecuada, ni por Hart ni por sus seguidores. An existe una gran
ambigedad acerca de la posicin exacta de la jurisprudencia entre la
razn terica y la razn prctica.
Palabras clave:
H. L. A. Hart, jurisprudencia analtica, jurisprudencia descriptiva, razn terica, razn prctica.
Fellow in Law, Mansfield College, Oxford.
117
PAVLOS ELEFTHERIADIS
Abstract:
Hart said that analytical jurisprudence is descriptive and general but did
not draw an explicit distinction between conceptual questions that are in
the main theoretical and others that are in the main practical. In Harts account, analytical jurisprudence searches for some kind of clarity about the
idea of law and the other basic legal ideas as they occur in our ordinary experience. In the Postscript he explains that jurisprudence is the theoretical
or scientific study of law as a social phenomenon, which may, on occasion,
but typically does not, rely on moral deliberation. It is to be contrasted to
practical questions as to how to live ones life. I shall call this view
descriptivism. It has been a very influential position in analytical jurisprudence and it is being defended by Harts followers even today. In this essay I raise some doubts about its coherence. First, I do not think Hart defended this view consistently even though he did so in the Postscript.
Second, I do not think that the position has been adequately defended, either by Hart or by his followers. There is still a great deal of ambiguity as
to the precise position of jurisprudence between theoretical and practical
reason.
Keywords:
H. L. A. Hart, Analytical Jurisprudence, Descriptive Jurisprudence, Theoretical Reason, Practical Reason.
118
SUMMARY: I. Generalisations. II. Semantic Recovery. III. Inductive Inference. IV. Conclusion.
Hart said that analytical jurisprudence is descriptive and

general.1 It is descriptive in that it is morally neutral and
has no justificatory aims.2 It is general in the sense that its
conclusions are not specific to any particular legal system
or legal culture, but aim to give an explanatory and clarifying account of law as a complex social and political institution with a rule-governed (and in this sense normative)
aspect.3 Hart does not draw an explicit distinction between
conceptual questions that are in the main theoretical and
others that are in the main practical. But he unequivocally
rejects the view that analytical jurisprudence deals with
practical questions. In his account, analytical jurisprudence searches for some kind of clarity about the idea of
law and the other basic legal ideas as they occur in our ordinary experience. In the Postscript he explains that jurisprudence is the theoretical or scientific study of law as a
social phenomenon, which may, on occasion, but typically
does not, rely on moral deliberation.4 It is a theoretical subject, part of theoretical as opposed to practical reason.
Harts theory is therefore interested in improving our view
of the world as it is and not saying how it should be. I shall
call this view descriptivism. Harts contemporary followers
take these views to lay the ground for their analytical pursuits. In this essay I wish to challenge this view. First, I do
not think Hart defended this view consistently even
though he did so in the Postscript. Second, I do not think
that the position has been adequately defended, either by
Hart or by his followers. There is still a great deal of ambi1 H. L. A. Hart, The Concept of Law, second edition (Oxford: Oxford University
Press, 1994) 239-40 (henceforth CL). For a forthright and clear defence of this view
see Andrei Marmor, Legal Positivism: Still Descriptive and Morally Neutral 26 Oxford Journal of Legal Studies (2006) 683.
2 CL, 240.
3 CL, 239.
4 CL, 209.
119
guity as to the precise position of jurisprudence between

theoretical and practical reason.
I. GENERALISATIONS
Jurisprudence for Hart is descriptive but not interested
in knowing everything about the law. It is interested in
suitable generalisations and general connections between
its various parts. Hart says that we should give a an explanatory and clarifying account of law, rising above trivial
specifics. Jurisprudence aims at a deep and illuminating
understanding of general truths about the law. Here is a
preliminary account of general jurisprudence that remains
faithful to the descriptive ambition, which I shall call (DJ):
(DJ) Analytical jurisprudence is the systematic and rational
inquiry aiming at true theoretical generalisations (explanans)
about the institutions and practices of law (explanandum),
arrived at through the accurate collection of relevant material facts.5
Approaches to law will differ both according to the type of

explanation they offer (explanans) and in the subject matter
they take to be exploring (explanandum). The sociology of
law, for example, is interested in law as a social phenomenon in its most directly empirical sense. Its subject matter,
its explanandum, is the whole range of conduct, beliefs and
intentions of persons who are implicated in legal practices.
The relevant generalisations of the sociologist concern the
regularity and other connections between distinct events
and actions, beliefs and intentions of persons. The main
aim is to establish causal links in terms of general laws or
their equivalent. This is because the typical explanation in
social science, as in all scientific explanation, is the causal
5 I borrow the terminology of explanans explanandum from Carl G. Hempel,
Studies in the Logic of Explanation in his Aspects of Scientific Explanation and
Other Essays in the Philosophy of Science (New York: The Free Press, 1965)
245-295. I intend to use this terminology as a neutral device, without intending to
endorse or reject any of Hempels views on the philosophy of science.
120
connection. This is the case even if the causal link rests on

an implicit theory of deeper motivating reasons or dispositional traits. We may observe, for example, that as a rule
when interest rates go up, inflation goes down and may
then adjust fiscal policy accordingly. It is obvious that jurisprudence, unlike economics and sociology, is not interested in causal connections of this kind. The most cursory
glance at the history and the continuing practice of legal
philosophy shows that it is a narrower project. It has a distinct subject-matter or explanandum.
Unlike social science, jurisprudence is not concerned
with the necessary and sufficient conditions of a prediction
that X and Y will act in certain ways. Jurisprudential theories of law and rights are concerned with legal practices
and legal doctrine. They are interested in the correct formation of propositions of law. This is the sense in which, for
example, Hart tells us that the idea of validity contributes
to jurisprudence. Validity tells us what is a rule of the system in the sense of an internal rule that constitutes a
ground for compliance with its contents: For the word
valid is most frequently, though not always, used, in just
such internal statements, applying to a particular rule of a
legal system, an unstated but accepted rule of recognition.6
Within such a framework jurisprudence seeks to understand propositions of law in more or less systematic ways.
Hart for example concludes that [w]herever a rule of recognition is accepted, both private persons and officials are
provided with authoritative criteria for identifying primary
rules of obligation.7 Sociologically informed theories of doctrinal law are of course possible, but they are not exactly
part of sociology. Such theories seek to explain law with the
help of generalisations concerning events, actions, beliefs
and intentions. They may find the criteria for the truth of
the propositions of law in some kind of pattern of conduct
or belief. Austin, for example, finds that in the idea of com6
7
CL, 103.
CL, 100.
121
mands of the Sovereign. But such theories do not offer explanations of conduct or belief by, say, pairing the existence
of commands by the Sovereign with any kind of conduct or
belief. Austins theory addresses the traditional questions of
jurisprudence, namely how to distinguish law form morals
and how to account for the idea of a legal system. As is well
known, Austin argued that what makes legal propositions
true is a certain pattern of behaviour, consisting in the existence of a sovereign, a habit of obedience and the threat
of sanctions and he used empirical observations in order to
explain legal doctrine and reasoning. But the link was not
causal. It was conceptual. The facts of power operated as
necessary and sufficient criteria for the existence of law and
a legal system of law. They are not offered as causes. They
are criteria for the correct application of the term law.
Analytical jurisprudence is therefore about propositions of
law, not the conduct or beliefs or dispositions such
propositions may bring about.
Descriptivism says that the correct application of such
terms depends on criteria that can be successfully described. The appropriate subject matter of legal theory is
not, therefore, every fact of current law but the inferences
we make about propositions of law on the basis of criteria.
The explanandum of jurisprudence, what stands to be explained by the best theory, is the fact that such inferences
are regularly and effortlessly made and applied by courts
and officials in the ordinary course of life. We conclude that
the aim of all ambitious theories of jurisprudence is the
generalisations that explain how propositions of law are
possible.
We can imagine three stages of this process. A descriptive
theory should first provisionally fix the domain of relevant
propositions of law, at least approximately. It should then
propose, at a second stage, a set of principles that explain
how legal propositions work within such a domain. It must
finally, at a third stage, have the principles tested against
our observations. We fix the domain, propose principles
122
and then test these principles against the facts of the

domain.
When descriptivism is put in this way it invites the
thought that it is actually paradoxical. We just said that we
need to test and improve all our explanations against observation. But a complete description may be self-defeating.
Let us say we wish to have a more careful description of our
subject-matter, the successful articulation of propositions
of law. We fix our domain somehow and then proceed with
the explanans of the domain. We may select an explanans
that includes more detail (e.g. that legal offices are held by
employees of the state) but narrows down the domain
somewhat. We exclude jurisdictions, say, where legal offices
are held by private individuals, not state officials. They are
not relevant to the inquiry. We narrow our domain in order
to offer a more complete analysis. But there is no way of
limiting this process. Under the descriptivist framework, it
seems that any added detail will be an improvement, even if
it covers a narrower domain. More accuracy is better than
less. We are always moved to adopt an explanatory generalisation that is closer to the facts. Inconsistencies will suggest that we must adopt a somewhat narrower scope. But
once we do so, we have undermined our theorys generality.
If adding any one more true feature of the analysandum, is
an improvement to our theory, then generality is bound to
disappear. Any explanation we offer will succumb to a more
detailed version of it. All possible features, properties and
stories of the analysandum will eventually become potentially relevant. The fact that the police wear uniforms becomes relevant. The fact that judges may or may not wear
wigs become relevant. And so on ad infinitum. If accuracy
is our ideal, we are led to a meaningless compilation, a
hopeless muddle.8 The supposed accuracy of a theory turns
into a paradox.
8 It would be perhaps something approaching the perfect map in Borges story,
where the cartographers of an imaginary Empire draw up a map so detailed that it
ends up covering exactly the territory. See Jorge Luis Borges, Collected Fictions,
123
It must be evident that descriptivism does not require

this. But how does it stop it? There ought to be two corrections. First, we cannot make the domain too narrow. If we
are offering a theory of law in general, we cannot limit our
observation to law in Britain or in France or in public international law. All relevant domains are to be included. We
cannot limit our observation to private law or to criminal
law. All areas are included. But the way we do that requires
a theory of the scope of our theory. Second, we do not seek
perfect, one to one accuracy. Our explanations are not
aimed to be perfectly informative. We say then that we seek
generalisations over a given domain. Both such aims go beyond accuracy and invite the consideration of other factors.
We must replace accuracy with a more complex set of
standards.
One such idea available to Hart at the time is that of conceptual explication. Quine described it as follows:
We have, to begin with, an expression or form of expression
that is somehow troublesome. It behaves partly like a term
but not enough so, or it is vague in ways that bother us, or it
puts kinks in a theory or encourages one or another confusion. But also it serves certain purposes that are not to be
abandoned. Then we find a way of accomplishing those same
purposes through other channels, using other and less troublesome forms of expression. The old perplexities are resolved.9
translated by Andrew Hurley, (London: Penguin 1999). An earlier suggestion of the
1:1 map was made by Lewis Carrolls Sylvie and Bruno (London: Macmillan, 1893)
vol. 2, p 169, in the section The Man in the Moon:
What do you consider the largest map that would be really useful?
About six inches to the mile.
Only six inches! exclaimed Mein Herr. We very soon got six yards to the mile.
Then we tried a hundred yards to the mile. And then came the grandest idea of all!
We actually made a map of the country, on the scale of a mile to the mile!
Have you used it much? I enquired.
It has never been spread out, yet, said Mein Herr: the farmers objected: they
said it would cover the whole country, and shut out the sunlight! So now we use the
country itself, as its own map, and I assure you it does nearly as well.
9 W. V. O. Quine, Word and Object (Cambridge, Mass.: The M.I.T. Press, 1960)
260.
124
The explanation we require is not a synonym or a paraphrase of the troublesome expression but an elimination,
Quine says, of the puzzles it causes.10 In this sense, analysis is not paradoxical. It is not meant to reproduce every aspect of our beliefs and practices, confusions, inconsistencies and all. Accuracy is not our guiding ideal. We seek to
create consensus, where there appears none at first sight.11
Generality is thus more important and accuracy is built
around it. This account of analysis suggests why we are not
allowed to tinker with the existing domain for the sake of
accuracy. We are not to limit it for the sake of convenience,
in order to discover a theory that fits. The task is harder.
We are offering an analysis of this domain, not another. So
at the second stage, the stage of the articulation of explanations, a successful theory of law will provide explanations of
inferences the conscious application of which will produce,
for the relevant domain, approximately the same propositions of law that we have as a matter of observable fact,
other things being equal. The consistency of such conclusions with the conclusions reached by active legal practitioners is the confirmation that the descriptive explanation is
correct. Our observations must confirm the theory. If they
do, we have offered an explication of the troublesome term.
Explication is meant to achieve this fit without perfect
accuracy.
Harts theory of law provides perhaps a good example of
an analytical explication of this type. As is well known,
Harts theory set out to answer three questions of law and
legal reasoning, namely how does law differ from orders
backed by threats, how does legal obligation differ from
moral obligation and what are rules and to what extent is
Word and Object, 260.

Quine writes (Word and Object, 272) that the strategy of semantic ascent is
that it carries the discussion into a domain where both parties are better agreed on
the objects (viz., words) and on the main terms concerning them. The strategy is
one of ascending to a common part of two fundamentally disparate conceptual
schemes, the better to discuss the disparate foundations.
10
11
125
law an affair of rules.12 In order to provide an answer Harts

theory rejected the argument for the command theory of
law. The explanations it offers are not a reduction of propositions of law to the facts of power, obedience, sovereignty
etc. Instead, we are presented with the argument that the
rule of recognition and the internal point of view help us
outline both the truth conditions for propositions of law
and the conditions for the existence of a legal system.
Harts theory replaces the salient features of law that any
educated man would be able to identify at least in some
skeleton way, with an explicit account of law on the basis
of the idea of a social rule, the rule of recognition and the
internal point of view.13 These are the abstract generalisations that explain law. When we apply them to our domain,
we see that they match more or less the inferences made by
ordinary lawyers in the practice of their trade. They track
the already known instances of propositions of law, but
without the confusion and uncertainty. Unlike Kelsen, our
assertion of the truth of the propositions of law is not just
hypothetical but is based on the social facts of a given legal order. Such facts are the background to the correct theory of the union of primary and secondary rules under a
rule of recognition. This is the sense in which Hart says
that the rule of recognition helps us assert the truth of
propositions of law here and now. The theory and the associated ideas of rule of recognition, validity and internal
point of view are offered as less troublesome expressions,
as both an explanation and a correction to our practices of
law. Harts theory offers thus a general explanation for the
inferences of legal practitioners and scholars.
It must be clear, therefore, that descriptive jurisprudence is concerned with the facts of experience but is more
complex than the description of actions and intentions, the
12
13
CL, 13.
CL, 3.
126
subject matter perhaps of a descriptive sociology.14 But here

we run into problems. The idea of explication does not allow
us to assume that a successful generalisation over a given
domain is always possible. Quine says that we use the explication to achieve our purposes through other channels.
Our traditions and practices may be so varied and conflicting that such no single theory may be able to capture them
all. If so, we may have to say that there cannot be a general
explication of this domain. If the relevant practices are too
confused, no amount of general theory will be able to account for them all. This may well be the case for law, where
theories and theorists have been divided for a very long
time.15 Hart himself recognises that the domain of law is full
of obscurities and disagreements. There are different theories of law and different theories of rights and other important legal concepts. Such disagreements do not concern
only borderline cases but also the very foundations of the
terms they seem to explain. Such conflicts and inconsistencies give rise to the persistent questions of jurisprudence
that occupy Hart at the start of his argument.16 How does
Hart overcome the pluralism of the theories and functions
of law? There are reasons to be cautious about the prospects of general explanations, even if accuracy is not the
only value.
II. SEMANTIC RECOVERY
Some theorists draw a distinction between the beliefs, attitudes and dispositions about law that people may have
and the meaning of law independent of such beliefs. They
14 CL, v. Hart writes as follows: Notwithstanding its concern with analysis the
book may also be regarded as an essay in descriptive sociology; for the suggestion
that inquiries into the meaning of words merely throw lights on words is false.
15 For a similar line of thought see Stephen R. Perry, Harts Methodological
Positivism in Coleman (ed.), Harts Postscript: Essays on the Postscript to the Concept of Law (Oxford: Oxford University Press, 2001) 311, at 328: Because a primary
goal of description is presumably accuracy, one would have thought that the external observer should simply describe what is there, confusions obscurities and all.
16 CL, 13-17.
127
argue that our explanatory generalisations should be tested

against the meaning, not against the beliefs. The argument
proceeds through a semantic argument about the term law
and its cognates: it offers a correct explication of the meaning of law, not of the beliefs or psychological states of any
persons. If the argument works, the correct explanatory
generalisations about law may disregard dissenting beliefs,
however widespread. So the beliefs of the legal realist or the
natural lawyer may not be troublesome. They are conceptual errors.
A semantic explication of this kind may defend a theory
of law as follows. First, we say that the limits of the relevant
domain are the limits set by the relevant language. Second,
we articulate appropriate explanatory principles that we
imagine may coincide with the underlying semantic standards for the application of law. We then test the theory
against our observations of current linguistic practice. The
required generalisations emerge thus inductively. They are
revealed little by little, in the process of uncovering underlying meanings from each instance of proper application.
There is good textual evidence that this is what Hart had in
mind as general and descriptive jurisprudence.17 He said
that by looking into the meaning of concepts we are looking
not merely at words but also at the realities we use
words to talk about. We are using a sharpened awareness
of words to sharpen our perception of the phenomena.18
The belief that there is a link between semantic questions
and appropriate jurisprudential explanations is also appar17 In addition to The Concept of Law, see also Hart, Definition and Theory in
Jurisprudence in H. L. A. Hart, Essays in Jurisprudence and Philosophy (Oxford:
Clarendon Press, 1983) 21. For these issues see generally Nicos Stavropoulos,
Harts Semantics in Coleman (ed.), Harts Postscript 59 and Timothy A. O.
Endicott, Law and Language in Jules Coleman and Scott Shapiro (eds.), The Oxford Handbook of Jurisprudence and Philosophy of Law (Oxford: Oxford University
Press, 2002) 935.
18 CL, 14. The quotation is from J. L. Austins A Plea for Excuses. For Harts
endorsement of the ordinary language school of philosophy see Hart, Jherings
Heaven of Concepts and Modern Analytical Jurisprudence in Hart, Essays in Jurisprudence and Philosophy (Oxford: Clarendon Press, 1983) 265, at 274-277.
128
ent in Harts pursuit of the meaning of expressions such as

command, obedience, being obliged, having an obligation and a legal system in the course of the argument on
the concept of law. In the preface to his Essays in Jurisprudence and Philosophy Hart described how the linguistic philosophy pursued by J. L. Austin in the 1950s and 1960s
appeared to him to be in a position to resolve age-old philosophical confusions. It achieved this by pointing out that
longstanding philosophical perplexities could often be resolved not by the deployment of some general theory but by
sensitive piecemeal discrimination and characterization of
the different ways, some reflecting different forms of human
life, in which human language is used.19
It would be instructive to compare Harts suggestions
with Austins own. J. L. Austin believed that the analysis of
ordinary language could yield significant results in many
types of philosophical inquiry, including political and legal
philosophy. In his well-known statement on philosophical
method Austin employed an example from the law (excuses) and defended the analysis of ordinary language on
the following grounds:
First, words are our tools, and, as a minimum, we should
use clean tools: we should know what we mean and what we
do not, and we should know what we mean and what we do
not, and we must forearm ourselves against the traps that
language sets us. Secondly, words are not (except in their
own little corner) facts or things: we need therefore to prise
them off the world, to hold them apart from and against it,
so that we can realize their inadequacies and arbitrariness,
and can relook at the world without blinkers. Thirdly, and
more hopefully, our common stock of words embodies all the
distinctions men have found worth marking, in the lifetimes
of many generations: these surely are likely to be more numerous, more sound, since they have stood up to the long
test of the survival of the fittest, and more subtle, at least in
all ordinary and reasonably practical matters, than any that
19
Hart, Essays in Jurisprudence 2.
129
you or I are likely to think up in our arm-chairs of an afternoon the most favoured alternative method.20
The mundane way people apply concepts in ordinary contexts already includes philosophically interesting distinctions, for it embodies the judgment of many generations.
Austin concludes that when we examine what we should
say when, what words we should use in what situations, we
are looking again not merely at words (or meanings, whatever they may be) but also at the realities we use the words
to talk about.21 When we succeed in linguistic analysis, we
do not just produce a better dictionary. We also produce a
better philosophy. The argument is that linguistic usage already embodies explanatory generalisations. Analysis just
recovers them from unthinking obscurity.
Austins example above gives us an indication of how the
method would work. We are to examine the applications of
the word law. We are interested in ordinary use. The hope
is that the relevant usage might be clear and uncontroversial. But it is not in every case. For example, we speak of
the laws of physics and the laws of probability. It is clear
that such usage of the word law lies outside our relevant
domain. Austins method explains that we are not interested in the word law in all its manifestations but in the
meaning of law in the sense of a tool that helps us understand this social institution. The contextual understanding
of words helps us locate the appropriate domain and avoid
the confusion caused by homonyms or unrelated senses of
the same word.22 Another advantage of the linguistic
method is that we have a way of explaining the derivation of
explanatory principles. When we presented the general am20 Austin, A Plea for Excuses in Austin, Philosophical Papers, edited by J. O.
Urmson and G. J. Warnock, third edition (Oxford: Clarendon Press, 1979) 181-2.
21 Austin, Philosophical Papers 182.
22 Such was the argument for linguistic analysis offered by Hart in Definition
and Theory in Jurisprudence in Essays in Jurisprudence 21. Hart says (p. 21) that
I wish to suggest that legal notions however fundamental can be elucidated by
methods properly adapted to their special character.
130
bition of legal theory we noticed that there is a gap between

the various particular data and the making of generalisations. With semantic analysis the gap disappears. Our data
collect usages of the relevant term that already imply the
generalisation. We do not derive it, we recover it. We are not
examining the beliefs of participants, for it is evident that
these beliefs clash, for example in the various debates between legal positivism, legal realism and natural law, or the
disagreement between the will and the interest theory of
rights. Underneath such disagreements we locate the already formed generalisations immanent in language. By
unearthing them, and not making them up, we stay within
the limits of a theoretical, descriptive jurisprudence. Semantic recovery vindicates descriptivism precisely because
it is direct and transparent.
Stavropoulos study of Harts use of the philosophy of
language has thrown much light on Harts methodological
views. Stavropoulos confirms that there is a semantic
programme of recovery at work in Harts work and shows
that Harts ambition was to capture not the actual beliefs of
speakers but the folk theory of their practices. Stavropoulos concludes as follows:
Harts method implies, first, that conceptual analysis is a
mode of inquiry that is distinct from and logically prior to
substantive theory; and, second, that conceptual analysis
aims at recovering some, perhaps idealized, common understandings, in the sense that it articulates but can never
transcend the understanding already implicit in ordinary use
and reflection. The second claim implies that the intuitions
elicited by conceptual analysis reflect the ordinary, conventional understandings of the target concept.23
If this analysis is correct, we can find an argument for semantic recovery in Harts work. According to this argument,
the explanatory generalisations we need, the explications
that jurisprudence produces, are in some way already im23 Nicos Stavropoulos, Harts Semantics in Coleman (ed.), Harts Postscript,
59-71.
131
plicit in current linguistic usage. Our philosophical explanation just unearths them.
Are there shared semantic criteria about law? Ronald
Dworkin has argued for many years against this view. For
Dworkin, the idea of doctrinal law is an interpretive concept
and subject to interpretation. If there are any semantic criteria, these may concern the sociological concept of law at
most. Any such criteria fail to resolve the disputes that legal positivists, legal realists and natural lawyers have about
the doctrinal sense of law.24 But the general argument for
semantic criteria has also been attacked in its entirety, as a
flawed epistemological programme. According to the critics,
no philosophical conclusions can be based on linguistic
convention at all. For Quine, we do not claim to make clear
and explicit what the users of the unclear expression had
unconsciously in mind all along and we do not expose hidden meanings, as the words analysis and explication
would suggest.25 Instead, Quine says, we fix on the particular functions of the unclear expression that make it worth
troubling about, and then devise a substitute, clear and
couched in terms to our liking, that fills those functions.26
Quines argument was meant to apply to logic, mathematics
and natural science, but it works equally well on social theory. There is no reason to believe that linguistic conventions resolve our questions about the nature and character
of any social institution. Explication is more creative than
24 Ronald Dworkin, Laws Empire (Cambridge, Mass., Harvard University
Press, 1986) 6-44, Ronald Dworkin, Justice in Robes (Cambridge, Mass., Harvard
University Press, 2006) 223-240.
25 Quine, Word and Object 258.
26 Quine, Word and Ojbect 258-9. Quine notes that the notion that analysis
must consist somehow in the uncovering of hidden meanings underlies also the recent tendency of some of the Oxford philosophers to take as their business an examination of the subtle irregularities of ordinary language (p. 259). But he deplores this narrowness and writes: It is ironical that those philosophers most
influenced by Wittgenstein are largely the ones who most deplore the explications
just now enumerated. In steadfast laymanship they deplore them as departures
from ordinary usage, failing to appreciate that it is precisely by showing how to circumvent the problematic parts of ordinary usage that we show the problems to be
purely verbal (p. 261).
132
this picture allows, says Quine. We look for substitutes that

are clear and to our liking.
Let us assume that there are regularities in ordinary usage, such that we can easily and without controversy formulate general rules and principles for the correct use of
propositions of law. This is a highly unrealistic assumption,
which ignores the long traditional divisions between legal
scholars and legal practitioners (and their divisions into legal positivists, legal realists and natural lawyers). Nevertheless, let us assume that these disputes have now died out
and that the agreed principles on which competent speakers ground their propositions of law are easily and
uncontroversially demonstrated by semantic analysis. What
would be the autthority of such principles? The semantic
argument assumes that agreement would carry with it
some deep philosophical significance. But is this not a surprising leap of faith? It suggests a curious foundationalism:
the truths of philosophy are determined by the fact of agreement in usage. But it may well be the case that the users
are all similarly mistaken. Consider the potential agreement
of a linguistic community over a word equivalent to miracle. Under the sway of religious fervour, the word miracle
may indeed mean for the users of this language the actual
performance of actions that have defied the laws of physics.
But this fact does not change anything in the way physics
should be understood. Its philosophers and physicists
should continue to challenge the plausibility of miracles, on
the basis of the experimental methods at their disposal.
This modestly suggests that language is always secondary
to reality. J. L. Austin admits this much when he says that
certainly, ordinary language has no claim to be the last
word, if there is such a thing and that in principle, it can
everywhere be supplemented and improved upon and superseded.27 But then why should the philosophy of society
or the philosophy of the natural world recover shared
meanings?
27
Austin, Philosophical Papers 185.
133
This applies not just for agreed criteria but also to the
case of model examples, that allow for some degree of flexibility over interpretation. In Words and Things Ernst
Gellner made precisely this point with particular clarity.28
Gellner noticed that the argument from paradigm cases
was also question-begging. When we identify the standard
or paradigm case of the use of a word we have only done
just that. We do not assert that this use has any wider significance. If we did, we would be begging the question in
the following way:
The Argument from Paradigm Cases does not even say that a
word is always rightly used, but merely that it is rightly used
in the Paradigm Case of its employment: and surely we
should be prepared to grant this. Indeed, it is a contradiction
to deny it. Words mean what a given language, its rules and
custom say they mean, neither more nor less.29
For Gellner the fallacy is to treat a de facto rule of language as a valid philosophical conclusion.30 Gellner writes
that the fact that there are standard cases for the application of the term such as miracle in a given society in no
way proves that such terms have a legitimate use.31 There
is no doubt that ordinary speakers use these terms. But
this does not mean that they have an empirical application,
or that they help explain any area of reality adequately.
Philosophical questions are, therefore, about the valid use
28 Ernst Gellner, Words and Things: an Examination of, and an attack on, Linguistic Philosophy, revised ed. (London: Routledge and Kegan Paul, 1979). The
work was first published as Words and things: a Critical Account of Linguistic Philosophy and a Study of Ideology (London: Gollancz, 1959). All references are to the revised second edition.
29 Gellner, Words and Things, 55-56. See also J. W. N. Watkins, Farewell to the
Paradigm-Case Argument 18 Analysis (1957) 25.
30 Gellner, Words and Things, 59-61. For a similar argument suggesting caution in the use of linguistic and other intuitions in philosophical arguments see
Jaako Hintikka, The Emperors New Intuitions 96 Journal of Philosophy (1999)
127.
31 Gellner, Words and Things, 56.
134
of terms, not about how in fact, a word is used.32 Answering the second question entails nothing about the first.33
All such epistemological objections, however, need not be
employed or pursued in great detail. The reason is that, as I
already mentioned above, the fundamental assumption on
which this account of semantic jurisprudence may be built
is manifestly false. As a matter of fact, there is no linguistic
convergence about law. The speakers of our language do
not speak of law with one voice.
As Dworkin and many others have shown, our beliefs
and meanings remain too diverse for a single theory.
Some theories dominate one domain whereas others dominate elsewhere. Hence, the most likely conclusion of a
semantic theory should be that there cannot be a semantic general theory of law at all. Perhaps there can only be
textbooks of the various areas of legal doctrine. The case of
Riggs v Palmer, made famous by Dworkin is a good example
of the persistence of disagreement.34 Here the propositions
of the majority and the minority reflected conflicting and inconsistent beliefs about law, so much so that they seemed
to be backed by conflicting meanings for law. For Dworkin,
this shows that law is in fact an interpretive concept. Whatever it is, it is not the result of linguistic convergence. The
same could also be said of the term right. The currency of
the will and interest theories shows that convergence is
also lacking. Semantic jurisprudence does not have an effective response to the incidence of such deep disagreements. All it can say is that if the use is inconsistent, there
cannot be a general theory. If the facts resist, we should
Gellner, Words and Things 60.

It may be that what moves the argument is a hidden conventionalism.
Stavropoulos calls it communitarian semantics; ibid. 86. He later notes that the
attempt to distil metaphysical wisdom out of ordinary use, makes no sense without
the assumption that ordinary use is founded on shared, common ground that defines or individuates the concepts that figure in use; ibidem, 88. Hart himself explicitly distanced himself form such implausible metaphysics in the Postscript;
CL, 247.
34 Riggs v. Palmer, 115 N.Y. 506, 22 N. E. 188 (1889). See LE, 15.
32
33
135
perhaps fragment the domain until we locate relevant

convergences.
J. L. Austin was well aware of such conflicts and disagreements. He also did not have much to say about them.
He said that in cases where language lacks coherence or we
have conflicting schemes we can still make some progress.
In the first case if the usage is loose, we can understand
the temptation that leads to it, and the distinctions that it
blurs.35 If, on the other hand, our usages disagree, then
you use X where I use Y, or more probably (and more intriguingly) your conceptual system is different from mine,
though very likely it is at least equally serviceable: in short,
we can find why we disagree you choose to classify in one
way, I in another.36 These types of ambiguities or disagreements should not daunt us. Austin tells us that all that is
happening is entirely explicable.37 Yet the result must be
that disagreements or confusions of this kind rule out a
single concept and a unified general theory. Whatever the
merits and advantages of semantic recovery elsewhere, it
seems unable to sustain the idea of a general theory of law.
III. INDUCTIVE INFERENCE
Perhaps the appropriate domain of jurisprudence is not
language. Perhaps we are to work with the whole range of
relevant actions, beliefs, dispositions and intentions about
law in order to propose and defend the relevant explanations of jurisprudence. For this view, we fix the domain
more loosely. We propose relevant generalisations to cover
the facts. We do so not on the basis of linguistic convergence, but on a more general observation of all the relevant
facts.
Such an approach avoids the problems about the validity
of language. Here we appeal directly to the facts of experi35
36
37
Austin, Philosophical Papers 184.

Idem.
Idem.
136
ence. Our disagreement in beliefs and theories may thus

turn out to be illusory. Our descriptive generalisation will
thus be based on factual, not linguistic adequacy. This is a
more directly empirical argument. It is an argument for
what philosophers of science call enumerative induction.
This is a non-deductive inference that follows the enumeration of relevant facts. In making this inference we infer from
the fact that a certain hypothesis explains the evidence, to
the truth of that hypothesis. Enumerative induction argues
from an observed correlation either to a generalization of
that correlation or to correlation in the next instance. So,
we may have an inductive inference as follows:
From premises of the form Many many As are known to be
B, and There are no known cases of As that are not B, and
C is A, the corresponding conclusion can be inferred of the
form C is B.
Induction does not establish necessary inferences in the

mode of a deductive inference. Induction establishes only
likelihood and probability, not certainty.
As the actual argument of The Concept of Law unfolds, it
becomes evident that an empirical reading of descriptivism
along these lines is also plausible. Hart does not pursue J.
L. Austins linguistic exercises with any particular tenacity.38 When he rejects the command theory of law and outlines a new theory based on the internal view of rules he
tells us that the command theory was inadequate because
it failed to fit the facts.39 As is well known, Harts view is
that legal relations are not the same as the gunman situation, whereby someone threatens another with violence in
38 Hart says so in the Postscript, CL 246: Thus, my doctrine that developed
municipal legal systems contain a rule of recognition specifying the criteria for the
identification of the laws which courts have to apply may be mistaken, but I nowhere base this doctrine on the mistaken idea that it is part of the meaning of the
word law that there should be such a rule of recognition in all legal systems, or on
the even more mistaken idea that if the criteria for the identification of the grounds
of law were not uncontroversially fixed, law would mean different things to different people.
39 CL, 80.
137
case of non-compliance, but are situations where one is

under an obligation. The distinction between being obliged
and being under an obligation is a distinction made in ordinary language. But Harts point is not that the notion of obligation fits usage, whereas the notion of command does
not. The point is rather that being under an obligation fits
better with the whole set of facts of the law, facts that run
deeper than linguistic meaning. The fact that law as an institution can be understood better from an internal rather
than an external point of view is not simply an observation
about what we say, but a result of the much broader observation of conduct, attitude and belief. Hart writes as follows: Most of the obscurities and distortions surrounding
legal and political concepts arise from the fact that these
essentially involve reference to what we have called the internal point of view: the view of those who do not merely record and predict behaviour conforming to rules, but use the
rules as standards for the appraisal of their own and others behaviour.40 The point is that the internal point of view
is required for the analysis of the basic concepts of obligation and duty.41 The argument against the command theory of law is not, therefore, an argument from the use of the
word law or any other word or set of words. It is that its
own explanatory generalisation does not correspond to important features of the existing legal system. Harts criticism seems concerns the correspondence of a theory with
the facts of the case, not the practices of language. Hart
tells us on many occasions that the command theory of law
failed because, legal systems do not fit its description,42 it
did not fit the facts43 whereas his own theory of law does.44
Earlier he had stated that the command theory was misCL, 98.
CL 98. See also Hart, Commands and Authoritative Legal Reasons in H. L.
A. Hart, Essays on Bentham: Studies in Jurisprudence and Political Theory (Oxford:
Clarendon Press, 1982) 243-268.
42 CL, 48.
43 CL, 80.
44 CL, 81.
40
41
138
leading, since there is little in any actual system which corresponds to it.45
Harts argument can be seen to follow in the footsteps of
an empiricist tradition of social philosophy originating in
Hume and Mill. Hume, for example, warned us that the
only solid foundation we can give to the study of human
nature and the science of man must be laid on experience
and observation.46 Pursuing the same line of thought, Mill
sought to apply the methods of physical science to society,
by generalising the methods successfully followed in the
former inquiries, and adapting them to the latter so as to
remove this blot on the face of science.47 For this school of
thought a social theorist is concerned with the facts of the
social world, leaving speculation about the right and the
good behind. Induction, not deduction, is the preferred
method of descriptivism. Nevertheless, there are important
problems with this reading.
The first problem concerns the criteria with which we
judge the success of inductive inference. The empirical argument is roughly as follows. From the fact that all observed relevant instances A of the relevant domain (e.g. legal rules in our legal order) are B (e.g. are viewed from an
internal point of view by the relevant officials) we may infer
that all As are Bs (all legal rules are generally viewed from
an internal point of view by the relevant officials). Under
what conditions is one permitted to make the inference?
The inference is not a deduction and its truth is not
demonstrated.
Gilbert Harman has shown that induction is theory rich.
It works as an inference to the best explanation, on the basis of a number of independent criteria.48 So an inductive
CL, 27.
David Hume, A Treatise of Human Nature, second edition by L. A. Selby-Bigge
and P. H. Nidditch (Oxford: Clarendon Press, 1978; first published 1739) xvi.
47 J. S. Mill, A System of Logic Ratiocinative and Inductive, eighth edition (London: Longmans, Green & Co., 1900; first published 1843) 546.
48 Gilbert Harman, The Inference to the Best Explanation 74 Philosophical Review (1965) 88. See also Gilbert H. Harman, Enumerative Induction and Best Ex45
46
139
argument is not exactly a report of facts, it is a theory

about these facts. Gilbert Harman has called the theories
that inform inductive inferences intermediate lemmas.49 The
role of lemmas is to confirm the truth of the inference. One
of Harmans examples involves us inferring that a mans
hand hurts by seeing how he jerks it away from a hot stove
which he has accidentally touched. The inference is from
behaviour to pain and to the reflex reaction. The lemma involved here is the proposition that the experience of pain
causes the sudden removal of the hand. In the theory of
law the lemma must be something a great deal more complex. The requirement of the internal point of view, for example, is hypothesis that grounds Harts conclusion about
the role of the rule of recognition. Hence, the inductive inference in this case is not just the result of observation but
part of a theory, which gives content and substance to the
hypothesis. If we described the inference merely as the result of enumeration, we would have missed the role of the
intermediate presuppositions. Harmans conclusion is that
we should speak more accurately of an inference to the best
explanation that exposes the role of intermediate lemmas.
Showing that induction is theory-rich affects descriptivism in an important way. There is no problem fixing the domain or proposing a hypothetical explanation. But we face
some uncertainty about testing, the third stage of confirming a descriptive theory. Harts theory offers the rule of recognition and the internal point of view as a general explication of the concept of law. But it is evident that it is not
confirmed by all cases. For example, when a legal realist or
a follower of natural law sit as judges, they will provide instances that contradict the theory of the rule of recognition.
It seems, for example, that the result of the majority in
Riggs contradicts it. Hart says that the internal point view
planation 65 The Journal of Philosophy (1968) 529. For further discussion see Peter Lipton, Inference to the Best Explanation, second edition (London: Routledge,
2004).
49 Harman, Inference to the Best Explanation 91.
140
is normally the case for officials, but there may be mistakes. We must take him to be saying that the hypothesis is
confirmed by a sufficient proportion of instances. Other
theories may have less success with the facts. Yet, we are
not given any evidence for this assertion. There is no comparison of competing inductions. Can the argument work
without this factual, i.e. empirical, testing?
The problem under the empiricist argument is that
descriptivism does not allow us to choose without such evidence. Harts theory is presented as the better or more normal description. And here is the problem with the inductive
argument. Hart and his followers, as Dworkin notices
offer no empirical argument of this kind. Instead they test
the theory on different grounds. They argue for its analytical clarity and coherence. But such considerations must be
secondary. We need a better description, not the description of something better. The latter commits us to the error
of wishful thinking. Harman, for example, states that in
cases of inductive inference the best explanation may be
preferred because it is a better, simpler, more plausible
hypothesis. But this assumes that it has the facts right and
has adequate predictive force. The argument from induction
suggests that: as long as the hypothesis that the next A
will be similar is a better hypothesis in the light of all the
evidence, the supposed induction is warranted.50 Here is
then the problem. Harts assumption concerning the internal point of view is not defended as the explanation that
matches the most or an adequate proportion of instances of
law.
Hart argues that rules have an internal point of view in
the sense that they are conceived and spoken of as imposing obligations when the general demand for conformity is
insistent and the social pressure brought to bear upon
them to those who deviate or threaten to deviate is great.51
Harts argument is that the external point of view, which
50
51
Harman, Inference to the Best Explanation 91.

CL, 86.
141
limits itself to observable regularities of behaviour, cannot

reproduce the way in which the rules function as rules in
the lives of those who normally are the majority of society.
... For them the violation of a rule is not merely a basis for
the prediction that a hostile reaction will follow but a reason for hostility.52 The internal point of view is an explanation of the normal case, not every case that involves legal
propositions. But how is the idea of the normal case vindicated? Hart says at the beginning of the book that he will
try to advance legal theory by providing an improved analysis of the distinctive structure of a municipal legal system
and a better understanding of the resemblances and differences and differences between law, coercion, and morality
as types of social phenomena.53 He is referring to a municipal legal system, most likely the legal system of the United
Kingdom in 1961.54 Perhaps the British of the time endorsed routinely the internal point of view and so did,
more importantly, the legal officials. Nevertheless, Hart
does not defend the theory on this ground. He says that his
account is what is normally the case, meaning generally
true. But he gives no evidence that this is the case.
What if we showed that it was the most frequently occurring in the United Kingdom? It would not be enough for the
kind of general theory we seek to establish. We need explanatory generalisations about law. Harts is offering an
idea of law that is state-based, institutional and positive.
This account of law follows British law in that it puts a high
premium on the consistency of its sources and the rationality of its own principles of reasoning. But not all prima facie
legal systems share the features of this paradigm case. For
example, some legal systems do not have the element of
state enforcement, international law being a well-known exCL, 90.
CL, 17.
54 Brian Simpson challenged the claim that Harts theory fits the facts of English law by drawing attention to the customary basis of the common law. See A. W.
B. Simpson, Common Law and Legal Theory in A. W. B. Simpson (ed.), Oxford Essays in Jurisprudence, Second series (Oxford: Clarendon Press, 1973) 77.
52
53
142
ample. Some systems are strongly customary. Other legal

systems are in force only as a matter of the external point
of view, in that they are less reliant on institutions and reasoning and more based on the likelihood of sanctions and
violence. They do not involve a principled commitment to
the rules and no criticism against transgressors, e.g. the
dictatorship. Other legal systems derive from faith as an
essential requirement, e.g. a theocracy.
How do we account for this diversity? Our concept aims
at generality. If we limit our domain to Britain, we have arbitrarily limited our domain and are offering a theory of
British law, not a theory of law. So the empirical-descriptive
argument is not immediately general. It is the general introduction to all the textbooks of British law, but not a general
theory of law. Hart asserts a set of properties that it considers normal for a legal system and then builds the theory
around them. But such observations, even if true, are not
enough to warrant a general theory of law, a theory that applies across jurisdictions. Although this descriptive theory
seeks to be confirmed by empirical observation it never tells
us exactly of what.55
Empirical descriptivism does not seem to have an effective response to the diversity of jurisdictions and the pluralism of theories. Such views cause indeterminacy that we do
not know how to manage. We can only report the various
tensions and inconsistencies that the views create. We can
propose various explanations in terms of various lemmas as
tentative accounts but such lemmas cannot be idealisations and must be confirmed by observation. Yet pluralism
eliminates them all, since none meets the test. If Harts theory meets the test in the British legal system, it does not
meet the test in international law, or Saudi Arabian law, or
German law. If it meets the test for some legal practitioners,
it does not meet it for many others (say those believing in
legal realism or natural law). It seems thus that empirical
55 For a similar argument see Gerald J. Postema, Jurisprudence as Practical
Philosophy, 4 Legal Theory (1998) 329, at 335-341.
143
descriptivism is thus forced to scepticism about law and jurisprudence. There may not be any essential truth about
law with universal value. Pluralism leads to fragmentation.
But we know that Hart was neither a pluralist nor a sceptic
about the concept of law or rights. He defended both a clear
general legal positivist theory of law and a clear and general
will theory of rights. The problem is that they cannot be the
result of descriptivism.
How did Hart get there? There is an answer, taking us
beyond descriptivism. In The Concept of Law Hart states
that the purpose of legal theory is not to provide a definition of law, in the sense of a rule by reference to which the
correctness of the use of the word can be tested; it is to advance legal theory by providing an improved analysis of the
distinctive structure of a municipal legal system and a
better understanding of the resemblances and differences
between law, coercion, and morality, as types of social phenomena.56 This suggests that the project is more constructive than we have so far assumed. Jurisprudence is improvement on and not a mirror image of legal practice. In
one of his later essays Hart said that in the course of jurisprudence we pick out and collect clusters of features frequently recurrent in the life of a legal system, to which it
was important to attend for some statable theoretical or
practical purpose.57 But that project, Harts own, is not a
descriptive project.
IV. CONCLUSION
Harts account of his own method and his execution of
that method is still puzzling. Description and generalisation
seem to pull towards entirely different directions. Harts
programmatic statements set out to vindicate descriptive
jurisprudence, a kind of jurisprudence that goes beyond
56
57
CL, 17 (emphasis added).

Hart, Legal Rights 162, at 164.
144
the morally laden constructions of the past but still

achieves general conclusions. He rejected the moralism of
Blackstone, Kant, Aristotle and, of course, the natural lawyers. But when he set out to perform the task of description, he left accuracy behind. If the arguments above are
correct, his approach was neither semantic nor one based
on induction. Generalisation cannot result from accurate
description for a number of reasons. First, there are distinct jurisdictions with different traditions and practices.
Second, there are different areas of law with different internal modes of reasoning and persuasion. Third, there are
different theories of law, such as legal realism, interpretivism and natural law, whose adherents practice it at every
level. Fourth, there are different tasks of the law depending
on whether is a judge, an official or an advocate. If we set
out to construct distinct sociologies for these different
things, i.e. the jurisdictions, the areas of law, the theories
and the roles, we would create a complex and largely useless tableau. Every theory of law must somehow deal with
this complexity. If we are to vindicate a general theory of
law, in the way Hart wished to defend it, in the face of the
fragmentation of the materials and the pluralism of legal
ideas, we need to go beyond the model of description. Harts
actual argument about law as a union of primary and secondary rules seems to me to go well beyond that model.
One of the greatest failures of Hart was that he did not see
the conflict between his programmatic statements and his
own practice.
145
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
H. L. A. HARTS MODERATE INDETERMINACY

THESIS RECONSIDERED:
IN BETWEEN SCYLLA AND CHARYBDIS?*
Imer B. FLORES**
Resumen:
En este artculo el autor, en el contexto del cincuenta aniversario de El
concepto del derecho de H. L. A. Hart, reconsidera la tesis de la indeterminacin moderada del derecho, la cual deriva de la textura abierta del
lenguaje. Para tal propsito, pretende: primero, analizar la tesis de la indeterminacin moderada del derecho, i.e. determinacin en los casos fciles e indeterminacin en los casos difciles, la cual recuerda la doctrina del trmino medio de Aristteles; segundo, criticar la tesis de la
indeterminacin moderada del derecho por fracasar en dar lugar al trmino medio virtuoso entre extremos viciosos, al insistir que el ejercicio
de la discrecin requerida constituye una legislacin intersticial; y, ter* Revised version of the papers presented in the Special Workshop H.L.A.
Harts The Concept of Law Reconsidered at the XXV IVR World Congress of Philosophy of Law and Social Philosophy Law, Science, Technology, in Frankfurt am
Main (Germany), August 18, 2011; and in Congreso Internacional de Filosofa del
Derecho, Coordinacin del Programa de Posgrado en Derecho, Facultad de Estudios Superiores Acatln, in Santa Cruz Acatln, Estado de Mxico (Mexico), November 14, 2011; and of the keynote address delivered in Primeras Jornadas
Internacionales de Filosofa del Derecho, Universidad Francisco Marroqun, Guatemala (Guatemala), November 4, 2011.
** Professor-Researcher, Instituto de Investigaciones Jurdicas (Legal Research Institute) and Facultad de Derecho (Law School), UNAM. E.mail:
imer@unam.mx. I am grateful to Tom Campbell, Pierluigi Chiassoni, Noam Gur,
Eerik Lagerspetz, Roger Shiner, and Ronaldo Macedo for their commentaries and
reactions in the Special Workshop. I am also thankful not only to Kenneth E.
Himma and Wilfrid J. Waluchow for their comments and observations but also to
Juan Vega for his public remarks to the keynote address and his suggestions. All
of them were extremely helpful on how to improve this version, especially Ken
Himma, but errors are still mine.
147
IMER B. FLORES
cero, reorganizar un argumento para una verdadera posicin intermedia,

la cual requiere de una forma de discrecin interpretativa dbil, en lugar
de una forma de discrecin legislativa fuerte.
Palabras clave:
Discrecin, indeterminacin, interpretacin, legislacin.
Abstract:
In this article the author, in the context of the fiftieth anniversary of H. L. A.
Harts The Concept of Law, reconsiders the moderate indeterminacy of law
thesis, which derives from the open texture of language. For that purpose,
he intends: first, to analyze Harts moderate indeterminacy thesis, i.e. determinacy in easy cases and indeterminacy in hard cases, which resembles Aristotles doctrine of the mean; second, to criticize his moderate
indeterminacy thesis as failing to embody the virtues of a center in between
the vices of the extremes, by insisting that the exercise of discretion required constitutes an interstitial legislation; and, third, to reorganize an
argument for a truly mean position, which requires a form of weak interpretative discretion, instead of a strong legislative discretion.
Keywords:
Discretion, Indeterminacy, Interpretation, Legislation.
148
HARTS MODERATE INDETERMINACY THESIS RECONSIDERED
But sailing your ship swiftly drive her past and avoid
her [i.e. Charybdis], and make for Skyllas rock instead, since it is far better to mourn six friends lost
out of your ship than the whole company.
Circes advice to Odysseus,
Odyssey, Book XII, 108-10.
in
HOMER,
The
Virtue, then, is a state of character concerned with

choice, lying in a mean, i.e. the mean relative to us,
this being determined by a rational principle, and by
that principle by which the man of practical wisdom
would determine it. Now it is a mean between two
vices, that which depends on excess and that which
depends on defect; and again it is a mean because
the vices respectively fall short of or exceed what is
right in both passions and actions, while virtue both
finds and chooses that which is intermediate.
ARISTOTLE, Nicomachean Ethics, Book II, Chapter
VI, 1106b, 36-1107a, 6.
SUMMARY: I. Introduction. II. The Moderate Indeterminacy

Thesis. III. Hart in Between... IV. Harts Scylla
and Charybdis. V. Conclusion.
I. INTRODUCTION
Reconsider H. L. A. Harts legacy in the golden anniversary
of The Concept of Law,1 in general, and his moderate indeterminacy of law thesis, in particular, are the principal
aims of this article. Actually, Hart is considered unarguably
among the jurists who contributed more to jurisprudence
in the second half of the Twentieth Century, by restoring legal philosophy to a central place in the study of both law
and (general) philosophy. Certainly, The Concept of Law was
1 H. L. A. Hart, The Concept of Law, Oxford: Oxford University Press, 1961
[hereinafter CL1].
149
IMER B. FLORES
quintessential for that purpose and has been highly influential ever since the original publication in 1961 and subsequently with the appearance of the 2nd edition with a
Postscript (edited by Penelope A. Bulloch and Joseph Raz)
in 1994.2
Personally, I consider as the core contributions of Hart:
(1) The concept of law as a (complex) model of rules i.e.
the union of primary and secondary rules thesis; (2) The
separation of law and morals thesis; and (3) The moderate
indeterminacy of law following the open texture of language thesis. In what follows, I will reconsider the third
thesis, but the first and second theses will be reconsidered
as well. Hence, in this paper, I am assuming a conceptual
methodology in which normative argument is relevant but
my analysis intends to remain mostly descriptive with three
main objectives: first, to analyze Harts moderate indeterminacy thesis, i.e. determinacy in easy cases and indeterminacy in hard cases, which, I argue, resembles Aristotles doctrine of the mean; second, to criticize his
moderate indeterminacy thesis as failing to embody the virtues of a center in between the vices of the extremes, by insisting that the exercise of discretion required constitutes
an interstitial legislation; and, third, to reorganize an argument for a truly mean position, which requires a form
of weak interpretative discretion, instead of a strong legislative discretion.3
2 H. L. A. Hart, The Concept of Law, 2nd. ed., Oxford: Oxford University Press,
1994 [hereinafter CL2].
3 Most of indeterminacy talk is related to language and meaning and can
be characterized as linguistic and semantic indeterminacy. However, at this
point, I would like to introduce a distinction between two additional kinds of indeterminacy relevant to law. On the one hand, there is epistemic indeterminacy related to uncertainty, i.e. our inability to know which the (correct) answer is to a legal dispute. On the other hand, there is systemic indeterminacy related to
incompleteness, i.e. our inability to reach a (single) answer in a legal dispute,
which may contingently derive from language but not necessarily reduced to it. In
short, there is epistemic indeterminacy when there is no way of knowing which the
(correct) answer is; and systemic indeterminacy when there is no way of reaching a
(single) answer. In that sense, law may be affected by linguistic and semantic
indeterminacy and appear to have systemic indeterminacy but not necessarily if
150
II. THE MODERATE INDETERMINACY THESIS

Regarding the problems of legal reasoning, in general,
and legal interpretation (and adjudication), in particular,4
H. L. A. Hart adopts, analogously to Hans Kelsen, a moderate version of the indeterminacy thesis, which is both
epistemic and systemic.5 Let me advance, that for the Austrian jurist, such indeterminacy derives mainly from the
hierarchical structure of the legal system,6 whereas for
the British legal philosopher, such indeterminacy derives
mostly from the open texture of language.7 Moreover, in
my opinion, Harts strategy additionally resembles Aristotles doctrine of the mean.8
This strategy is quite explicit throughout Harts work9
and becomes self-evident by bringing into attention both
there is a way of reaching a (single) answer transcending the linguistic and semantic indeterminacy. I am indebted to Ken Himma for asking me to introduce
this distinction and even for proposing some labels.
4 It is worth to mention that Hart prepared the entry Problems of the Philosophy of Law for Paul Edwards Encyclopedia of Philosophy, which was published
originally, in 1967, containing only two sets of problems: Problems of Definition
and Analysis; and, Problems of the Criticism of Law. But in the revised version
published in his brown book, in 1983, he included a third set, inserted in between the two original ones, namely: Problems of Legal Reasoning. Vid. H. L. A.
Hart, Problems of the Philosophy of Law, in Essays in Jurisprudence and Philosophy, Oxford: Oxford University Press, 1983, pp. 98-109.
5 Vid. Hans Kelsen, Introduction to the Problems of Legal Theory, trans. Bonnie
Litschewski Paulson and Stanley L. Paulson, Oxford: Oxford University Press,
2002, pp. 77-89 [hereinafter PTL1]; and, Pure Theory of Law, 2nd ed., trans. Max
Knight, Berkeley and Los Angeles: University of California Press, 1967, pp. 349-50
[hereinafter PTL2]; and, Hart, CL1, pp. 121-50; and CL2, pp. 124-54. Vid. also
Duncan Kennedy, A Left Phenomenological Critique of the Hart/Kelsen Theory of
Legal Interpretation, in Enrique Cceres et al., Problemas contemporneos de la
filosofa del derecho, Mxico: Instituto de Investigaciones Jurdicas, UNAM, 2005,
pp. 371-83; and a substantially revised version printed in: Legal Reasoning: Collected Essays, Aurora, Colorado: The Davies Group, 2008, pp. 153-73 [references
will be made to this version].
6 Kelsen, PTL1, p. 77.
7 Hart, CL1, p. 124; and CL2, p. 128.
8 Cfr. Aristotle, Nicomachean Ethics, in The Basic Works of Aristotle (Richard
McKeon ed.), New York: Random House, 1941, Book II, Chapter VI, 1106b,
36-1107a, 6, p. 959.
9 Vid. for example, CL1, pp. 191-2; and CL2, p. 196: But if men are not devils,
neither are they angels; and the fact that they are a mean between these two ex-
151
IMER B. FLORES
the title of chapter VII of his The Concept of Law: i.e. Formalism and Rule-Scepticism,10 and the subtitle of his 1977
Sibley Lecture American Jurisprudence through English
Eyes: i.e. The Nightmare and the Noble Dream,11 both of
which allow Hart to stand somewhere in a center between
extremes represented not only by formalism and anti-formalism, i.e. rule-skepticism; but also by realism, i.e. nightmare (or too bad to be remembered) and idealism, i.e.
noble dream (or too good to be true).
1. Hans Kelsens Frame
Before proceeding with Hart, let me call attention to the
fact that for Kelsen the indeterminacy derives from the
hierarchical structure of the legal system and is labeled
as (more or less) relative to the levels higher or lower
of the legal system and to the movement from one level to
the next. In that sense, the higher-level norm determines
the process for the creation of the lower-level norm and
possibly the content as well or at least to some extent. As
Kelsen acknowledges:12
This determination, however, is never complete. The higherlevel norm cannot be binding with respect to every detail of
the act putting it into practice. There must always remain a
range of discretion, sometimes wider, sometimes narrower,
so that the higher-level norm, in relation to the act of applying it (an act of norm creation or of pure implementation),
has simply the character of a frame to be filled in by way of
the act. Even a meticulously detailed command must leave a
number of determinations to those carrying it out. If official
tremes is something which makes a system of mutual forbearances both necessary
and possible.
10 Hart, CL1, p. 121; and CL2, p. 124.
11 H. L. A. Hart, American Jurisprudence through English Eyes: The Nightmare and the Noble Dream, which was printed first in 11 Georgia Law Review, 969
(1977); and, later, in Essays..., supra note 4, pp. 123-44 [references will be made to
this version].
12 Kelsen, PTL1, p. 78 (emphasis added).
152
A orders official B to arrest subject C, B must use his own

discretion to decide when, where, and how he will carry out
the warrant to arrest C; and these decisions depend upon
external circumstances that A has not foreseen and, for the
most part, cannot foresee.
Additionally, Kelsen distinguishes between two kinds of

indeterminacy: intended and unintended.13 According
to him, the former can be part of the intention of authority
issuing the higher-level norm, who decides to leave open for
later settlement (by authorities responsible of determining
the lower-level norms) not only the answer to the question
what is the prescribed act but also why it is so; and, the
latter can transcend the intention of the authority issuing
the higher-level norm due to: (1) the ambiguity (or vagueness) of a word or a phrase used in expressing the norm; (2)
the discrepancy, which can be total or partial, between the
linguistic expression of the norm and the will of the
norm-issuing authority; and (3) the contradictory existence
of at least two norms purporting to be simultaneously valid
and applicable to the same factual situation.14 In Kelsens
own voice:15
In all these cases of intended or unintended indeterminacy
of the lower level, various possibilities for applying the
higher-level norm suggest themselves. The legal act of applying the legal norm can be made to correspond to one or another of the several possible readings of the norm. Or it can
be made to correspond to the norm-issuers will, however
discovered, or to the expression he chooses. Or, in the case
of the two norms contradicting each other, the legal act can
be made to correspond to one or the other of them, or it can
be so fashioned that decisions are taken as if norms abrogated one another. In all these cases the norm to be applied
13 Ibidem, pp. 78-80. It is worth to mention that in the English translation of
the second edition of Reine Rechtslehre the word indeterminacy has been
changed into indefiniteness. Vid. Kelsen, PTL2, pp. 349-50.
14 Kelsen, PTL1, pp. 78-80.
15 Ibidem, p. 80 (emphasis added).
153
IMER B. FLORES
is simply a frame within which various possibilities for application are given, and very act that stays within this frame, in
some possible sense filling it in, is in conformity with the
norm.
Kelsen not only advocates that the norm to be interpreted

represents a frame encompassing the cognition of various
possibilities for application but also challenges the traditional jurisprudence for its formalist inclination to believe
that, invariably, when the statute is applied in the concrete
case, it can provide only one correct decision, and that the
correctness of this decision its correctness in terms of
the positive law is based on the statute itself.16 In his
words:17
From the standpoint of the positive law, however, there is no
criterion on the basis of which one of the possibilities given
within the frame of the norm to be applied could be favoured
over the other possibilities. In terms of the positive law, there
is simply no method of according to which only one of the
several readings of a norm could be distinguished as correct
assuming, of course, that several readings of the meaning
of the norm are possible in the context of all the other norms
of the statute or of the legal system. In spite of every effort,
traditional jurisprudence has not yet found an objectively
plausible way to settle the conflict between will and expression. Every method of interpretation developed thus far invariably leads merely to a possible result, never to a single
correct result.
In short, Kelsen is right that as a matter of empirical fact

within a frame there is not one but many possible applications of a norm and that the authority responsible for determining the lower-level norm is not invariably (or mechanically) in a position to reach the single one and even
less that its decision is necessarily the correct one, i.e. an
16
17
Ibidem, p. 81.
Idem (emphasis added).
154
epistemic indeterminacy. Nevertheless, he is apparently

wrong in assuming a systemic indeterminacy, i.e. that there
is never a single answer following not from the legal statute
itself but from the law and the legal system as such. In my
opinion, Kelsen fails to distinguish between the application
of one of the various cognitive possibilities and the justification of such application as the correct one required by the
law as a whole. In that sense, by pointing to the discretion
of the authority to determine the lower-level norm, i.e. to
choose contingently among the various possible applications one but not necessarily the correct one, he is falling
not only short of the aims of a truly pure (normative) theory
of law, which separates law both from fact and from morality,18 but also giving up totally a claim for legal certainty
by labeling it an illusion.19
At this point, let me bracket his double rejoinder that
such decision: (1) is or constitutes an act of norm creation (or discretion in a strong sense); and (2) transforms
the norm of morality, justice or so on into a norm of positive law.20 Actually, for that purpose, we turn now to Harts
account and to the question on whether he provides a
better or even successful answer to the problems at
hand.
2. H. L. A. Harts Core and Penumbra
As stated before, for Hart the systemic indeterminacy or
the so-called open texture of law derives from the open
texture of language. His analytical argument seems well
structured and runs as follows: If all language (considered
as a whole) is open textured (or has open texture) and law
is expressed in (terms of) language; thus, it logically follows
18
19
20
Vid. Stanley Paulson, Introduction, in Kelsen, PTL1, p. xxvi.

Kelsen, PTL1, pp. 83-4.
Kelsen, PTL1, pp. 82-3.
155
IMER B. FLORES
that law is open textured (or has open texture), and as such
is indeterminate (or has indeterminacy):
All language is open textured
Law is expressed in language
! Law is open textured
In my opinion, Harts inference is wrong: from the fact

that all language (considered as a whole) is open textured
(or has open texture) and law is expressed in (terms of) language; it does not follow that law is open textured (or has
open texture), and as such is indeterminate (or has indeterminacy). At most, what Hart is able to demonstrate is that
since language and law are so closely interrelated the
open texture of language is present in law, but more precisely in the language in which law is expressed. Moreover,
that neither does mean that law as a whole is necessarily
open textured (or has open texture) nor that law is not prepared to deal with the open texture of language by appealing to something else beyond language to declare its meaning. In terms of Roberto Mangabeira Unger the latter is
nothing but a form of false necessity.21
Let me clarify, I am neither denying that all language is
open textured (or has open texture) nor that law is expressed in (terms of) language, but I am skeptical of reducing both law to (a form of) language and legal rationality to
(a form of) linguistic rationality. Certainly, language is used
to express propositions, in general, and propositions about
law, i.e. legal propositions, in particular, but clearly law is
neither identical to such propositions nor to language.
From the fact that law is identified in words that are verbally extricated or verbally formulated22 explicitly and
even implicitly it neither does follow that law is (or can
21 Cfr. Roberto Mangabeira Unger, Social Theory: Its Situation and Its Task. A
Critical Introduction to Politics, a Work in Constructive Social Theory, Cambridge:
Cambridge University Press, 1987.
22 Hart, CL1, pp. 122-3; and CL2, pp. 125-6.
156
be) exhausted by nor is (or can be) reduced to language. In

that sense, being expressed in (terms of) language is a necessary but not a sufficient condition of law; and, thats why
language can neither be the unique, i.e. the one and only,
criteria in law nor the ultimate one.23
Anyway, we will proceed with the exam of the merits and
demerits of Harts account:
First, Hart akin to Kelsen arrives at the conclusion
that (some degree of) indeterminacy in law is inevitable,
since sometimes or most of the time it is necessary to
leave certain issues open for later settlement, and also relative. But unlike Kelsen, Hart bases it mainly in the open
texture of language and not in the hierarchical structure
of law.24
Second, Hart alike Kelsen suggests that the authoritative general language in which a rule is expressed may
guide only in an uncertain way much as an authoritative
example does25 and identifies two connected handicaps:
The first handicap is our relative ignorance of fact; the second is our relative indeterminacy of aim.26
Third, Hart analogous to Kelsen reaches the conclusion that some form of discretion (i.e. choice) is inevitable
but, unlike the Austrian jurist, he holds that it is due to the
open texture of language and for him also of law. For that
purpose, reintroduces the No vehicles in the park example.27 In Harts voice:28
Faced with the question whether the rule prohibiting the use
of vehicles in the park is applicable to some combination of
23 I am indebted with Ken Himma for calling me to make explicit some ideas
that remained implicit in the previous paragraphs.
24 Hart, CL1, p. 124 and 128; and CL2, pp. 128 and 131.
25 Hart, CL1, p. 124; and CL2, p. 127.
26 Hart, CL1, p. 125; and CL2, p. 128 (emphasis added).
27 Hart introduces the example in his 1957 Holmes Lecture Positivism and the
Separation of Law and Morals delivered at Harvard Law School, which was published first in 71 Harvard Law Review 593 (1958); and, later, in Essays, supra
note 4, pp. 49-87 [references will be made to this version.]
28 Hart, CL1, p. 124; and CL2, p. 127.
157
IMER B. FLORES
circumstances in which it appears indeterminate, all that

the person called upon to answer can do is to consider (as
does one who makes use of a precedent) whether the present
case resembles the plain case sufficiently in relevant respects. The discretion thus left to him by language may be
very wide; so that if he applies the rule, the conclusion, even
though it may not be arbitrary or irrational, is in effect a
choice.
Fourth, Hart like Kelsen considers that (in)determinacy is a matter of degree: law is determinate in some areas and indeterminate in others. For the Austrian jurist law
is indeterminate inside the frame and determinate outside
it: it is not-law at all; and for the British legal philosopher
law is determinate in the core and indeterminate in the penumbra.29 In Harts words, as originally introduced in the
Holmes Lecture in 1957:30
A legal rule forbids you to take a vehicle into the public park.
Plainly this forbids an automobile, but what about bicycles,
roller skates, toy automobiles? What about aeroplanes? Are
these, as we say, to be called vehicles for the purpose of the
rule or not? If we are to communicate with each other at all,
and if, as in the most elementary form of law, we are to express our intentions that a certain type of behaviour be regulated by rules, then the general words we use like vehicle in the case I consider must have some standard
instance in which no doubts are felt about its application.
There must be a core of settled meaning, but there will be, as
well, a penumbra of debatable cases in which words are neither obviously applicable nor obviously ruled out.
And, at the end of chapter VI The Foundations of a Legal

System of his masterpiece The Concept of Law in 1961:31
29 Vid. Kennedy, A Left Phenomenological Critique of the Hart/Kelsen Theory
of Legal Interpretation, supra note 4, p. 154.
30 Hart, Positivism and the Separation of Law and Morals, supra note 27, p.
63 (emphasis added).
31 Hart, CL1, pp. 119-20; and CL2, p. 123 (emphasis added).
158
All rules involve recognizing or classifying particular cases as

instances of general terms, and in the case of everything
which we are prepared to call a rule it is possible to distinguish clear central cases, where it certainly applies and others where there are reasons for both asserting and denying
that it applies. Nothing can eliminate this duality of a core of
certainty and a penumbra of doubt when we are engaged in
bringing particular situations under general rules. This imparts to all rules a fringe of vagueness or open texture...
Fifth, Hart similar to Kelsen concludes that there are

at least in the peripheral cases no correct decisions or right
answers and points to the authority granted to those exercising choice or discretion and their finality (not infallibility): there is no answer which is clearly right or wrong.
These can be settled only by a choice, made by someone to
whose choices in this matter authority is eventually accorded.32 On this point, keep in mind that Hart had already wisely stated not only [I]n an ordinary game the
score is what the scorer says it is is not the scoring rule: it
is a rule providing for the authority and finality of his application of the scoring rule in particular cases33 but also
The scorer may make honest mistakes...34
III. HART IN BETWEEN...
So far, both Hart and Kelsen agree in the claim that indeterminacy in law is inevitable and relative, but disagree
in the reason for grounding it: indeterminacy results for the
former from the open texture of language, and for the latter from the hierarchical structure of law. Additionally,
they seem to differ regarding its implications chiefly to legal
certainty. In my opinion, on the one hand, Kelsen, by calling it a mere illusion, throws the baby out with the bath
32
33
34
Hart, CL1, p. 146; and CL2, p. 150 (emphasis added).

Hart, CL1, p. 140; and CL2, p. 144.
Hart, CL1, p. 139; and CL2, pp. 142-3.
159
IMER B. FLORES
water (or even worse throws the baby out and keeps the
bath water):35 interpretation of law is uncertain, i.e. epistemic indeterminacy, and more or less incomplete, i.e. systemic indeterminacy. On the other hand, Hart, by standing
over the shoulders of Oliver Wendell Holmes prediction/prophecy theory,36 defends it as a matter of degree: interpretation of law is more or less uncertain, i.e. epistemic
indeterminacy, and more or less incomplete, i.e. systemic
indeterminacy, depending on the open texture of language.
At the end, it is this strategy, as already advanced, which
allows Hart to stand apparently somewhere in a center of
virtue between extremes of vice not only (1) in between formalism and anti-formalism, i.e. rule-skepticism, and (2) in
between realism, i.e. the nightmare, and idealism, i.e. the
noble dream.
1. Formalism and Anti-Formalism (i.e. Rule-Skepticism)
On one side, Hart appears to give his dues both to formalism agreeing that there are some central or paradigmatic cases falling within a core of certainty or settled
meaning, but disagreeing that all cases are clear and precise; and, to anti-formalism (i.e. rule-skepticism) arguing that
there are other peripheral cases falling within a penumbra
35 Hart did use this expression to refer to Robert Nozicks Anarchy, State and
Utopia, New York: The Free Press, 1974, vid. H. L. A. Hart, 1776-1976: Law in the
Perspective of Philosophy, which was published first in 51 New York Law Review
538 (1976); and, later, in Essays, supra note 4, pp. 145-158 [references will be
made to this version.] Ibidem, p. 152: Other theories perhaps Professor Nozicks
among them do worse: they throw out the baby and keep the bath-water.
36 Vid. Oliver Wendell Holmes Jr., The Path of the Law, Harvard Law Review,
Vol. 10, 1897, pp. 457-78; and reprinted in Vol. 110, 1997, pp. 991-1009 [reference will be made to this version]. Ibidem, p. 993: If you want to know the law and
nothing else, you must look at it as a bad man who cares only for the material consequences which such knowledge enables him to predict, not as a good one who
finds his reasons for conduct, whether inside the law or outside of it, in the vaguer
sanctions of conscience. Ibidem, p. 994: But if we take the view of our friend the
bad man we shall find that he does want to know what the courts are likely to do
in fact. I am much of his mind. The prophecies of what courts will do in fact, and
nothing more pretentious, are what I mean by the law.
160
of doubt or unsettled meaning, but assenting that not all

cases are unclear and imprecise. In that sense, he not only
distinguishes between clear and envisaged cases (such as
the automobile/motor-car, the bus, and the motor-cycle) and unclear and unenvisaged ones (such as the bicycle, the roller-skates, and the toy automobile/motor-car
electrically propelled);37 but also insists in the existence of
a need for a further exercise of choice in the application of
general rules to particular cases.38
Actually, it is the abandonment of this need which constitutes the vice attributable to formalism or conceptualism [which] consists in an attitude to verbally formulated
rules which both seeks to disguise and to minimize the
need for such choice, once the general rule has been laid
down.39 Analogously, it is the mistreatment of another
(equally important) need which constitutes the vice attributable to anti-formalism or realism i.e. rule-skepticism:
an attitude to verbally formulated rules which seeks to disclose and to maximize the need for such choice, to the extent that the general rule laid down does nothing at all, i.e.
does not provide any guideline as such.40 On the contrary,
in order [t]o escape this oscillation between extremes,
Hart proposes a virtuous compromise between two social
needs and suggests:41
In fact all systems, in different ways, compromise between
two social needs: the need for certain rules which can, over
great areas of conduct, safely be applied by private individuals to themselves without fresh official guidance or weighing up of social issues, and the need to leave open, for lat37 Vid. Hart, Positivism and the Separation of Law and Morals, supra note 27,
p. 63. Vid. also Hart, CL1, pp. 125-6; and CL2, p. 129.
38 Hart, CL1, p. 126; and CL2, p. 129.
39 Idem.
40 Vid. Hart, CL1, p. 133; and CL2, p. 136: Yet rule-scepticism, or the claim
that talk of rules is a myth, cloaking the truth that law consists simply of the decisions of courts and the prediction of them, can make a powerful appeal to a lawyers candour.
41 Hart, CL1, p. 127; and CL2, p. 130.
161
IMER B. FLORES
ter settlement by an informed, official choice, issues which

can only be properly appreciated and settled when they arise
in a concrete case.
It is clear that the open texture of language allows Hart to

respect the first competing social need while permits him to
respond to the second one by appealing to a later informed
exercise of official choice or discretion: In every legal system a large and important field is left open for the exercise of
discretion by courts and other officials in rendering initially
vague standards determinate, in resolving the uncertainties of
statutes, or in developing and qualifying rules only broadly
communicated by authoritative precedents.42
At the end, Hart considers that we face a false dilemma:43
Either rules are what they would be in the formalists heaven and they bind as fetters bind; or there are no rules, only
predictable decisions or patterns of behaviour.
In that sense, he suggests:44

Formalism and rule-scepticism are the Scylla and Charybdis
of juristic theory; they are great exaggerations, salutary
where they correct each other, and the truth lies between
them. Much indeed that cannot be attempted here needs to
be done to characterize in informative detail this middle
path, and to show the varied types of reasoning which courts
characteristically use in exercising the creative function left
to them by the open texture of law in statute or precedent.
2. Nightmare and Noble Dream

On the other, Hart seems to pay his dues both to the
nightmare agreeing that in some cases judges make the law
which they apply to litigants, but disagreeing that they
42
43
44
Hart, CL1, pp. 132-3; and CL2, p. 136.

Hart, CL1, p. 136; and CL2, p. 139.
Hart, CL1, p. 144; and CL2, p. 147.
162
never declare the existing law;45 and, to the noble dream arguing that in some cases judges do not make law and declare the existing law, but assenting that sometimes they
do make (new) law and do not declare the existing law
since there is no existing law to be declared. In that
sense, American Jurisprudence has oscillated between two
extremes with many intermediate stopping-places.46 Anyway, Hart acknowledges that Litigants in law cases consider themselves entitled to have from judges an application
of the existing law to their disputes, not to have new law
made for them and proceeds not only to delineate the image of the judge following Lord Radcliffe as an objective, impartial, erudite, and experienced declarer of the law
but also to distinguish it from the very different image of
the legislator: the maker of the law, i.e. the law-maker.47
In the nightmare view identified with the American Legal
Realism movement of 1920s and 1930s, but the characterization is equally applicable to the Critical Legal Studies
movement of late-1970s and mid-1980s, and other critical
theories since then48 the distinction between the judge
and the legislator is a mere illusion. The American Legal
Realists especially Jerome Frank and Karl N. Llewellyn
according to Hart were concerned to stress the legislative
opportunities of the courts49 and their main effect was to
convince many judges and lawyers, practical and academic,
of two things:50
45 Vid. Hart, American Jurisprudence through English Eyes..., supra note 11,
p. 127: [In] the Nightmare view... judges make the law which they apply to litigants and are not impartial, objective declarers of existing law. Ibidem, p. 128:
[A]s if adjudication were essentially a form of law-making, never a matter of declaring the existing law.
46 Ibidem, p. 125.
47 Ibidem, p. 126 [reference is omitted].
48 Vid. v. gr. Brian Bix, Jurisprudence: Theory and Context, 3rd. ed., London:
Sweet & Maxwell, 2003, pp. 177-187, and 217-236.
49 Hart, American Jurisprudence through English Eyes..., supra note 11, p.
131.
50 Ibidem, p. 132.
163
IMER B. FLORES
[F]irst, that they should always suspect, although not always

in the end reject, any claim that existing legal rules or precedents were constraints strong and complete enough to determine what a courts decision should be without other extra-legal considerations; secondly, that judges should not
seek to bootleg silently into the law their own conceptions of
the laws aims or justice or social policy or other extra-legal
elements required for decision, but should openly identify
and discuss them.
On the contrary, in the noble dream view represented

originally by Roscoe Pound, among others, and in a more
contemporary version by Ronald Dworkin the distinction
between the legislator and the judge, as well as their respective functions, i.e. law-making ius dare and law-declaring ius dicere, is still quite significant.51 In that sense,
Hart suggests that the noble dream:52
Like its antithesis the Nightmare, it has many variants, but
in all forms it represents the belief, perhaps the faith, that,
in spite even of whole periods of judicial aberrations and
mistakes, still an explanation and a justification can be provided for the common expectation of litigants that judges
should apply to their cases existing law and not make new
law for them even when the text of particular constitutional
provisions, statutes, or available precedents appears to offer
no determinate guide. And with this goes the belief in the
possibility of justifying many other things, such as the form
of lawyers arguments which, entertaining the same expectations, are addressed in courts to the judges as if he were
looking for, not creating the law; the fact that when courts
overrule some past decision, the later new decision is nor51 Vid. Imer B. Flores, The Quest for Legisprudence: Constitutionalism v. Legalism, en Luc J. Wintgens (ed.), The Theory and Practice of Legislation: Essays on
Legisprudence, Aldershot, Ashgate, 2005, pp. 46-7; Legisprudence: The Forms
and Limits of Legislation, 1 Problema. Anuario de Filosofa y Teora del Derecho 247
(2007), pp. 257-60; and Legisprudence: The Role and Rationality of Legislators
vis--vis Judges towards the Realization of Justice, 1:2 Mexican Law Review,
91 (2009), pp. 100-6.
52 Hart, American Jurisprudence through English Eyes..., supra note 11, pp.
132-3.
164
mally treated as stating what the law has always been, and
as a correcting mistake, and is given a retrospective operation; and finally, the fact that the language of a judges decision is not treated, as is the language of a statute, as the authoritative canonical text of a law-making verbal act.
Finally, Hart concludes:53

I have portrayed American jurisprudence as beset by two extremes, the Nightmare and the Noble Dream: the view that
judges always make and never find the law they impose on
litigants, and on the opposed view that they never make it.
Like any other nightmare and any other dream, these two
are, in my view, illusions, though they have much of value to
teach the jurist in his waking hours. The truth, perhaps unexciting, is that sometimes judges do one and sometimes the
other.
IV. HARTS SCYLLA AND CHARYBDIS

Let me start this section, by recalling Harts characterization of formalism and anti-formalism, i.e. rule-skepticism,
as the Scylla and Charybdis of juristic theory and his insinuation that the truth lies between them... [i.e. in the]
middle path which courts characteristically use in exercising the creative function left to them by the open texture
of law in statute or precedent.54 By the by, the depiction is
equally applicable to realism (i.e. the nightmare) and to idealism (i.e. the noble dream), and his suggestion that Like
any other nightmare and any other dream, these two are
illusions, though they have much of value to teach the jurist in his waking hours. The truth, perhaps unexciting, is
that sometimes judges do one and sometimes the other.55
It is worth to mention that Scylla and Charybdis are
mythical sea monsters associated with two rocks portrayed
53
54
55
Ibidem, p. 144.
Vid. supra note 44.
Vid. supra note 53.
165
IMER B. FLORES
by Homer in The Odyssey.56 The former was described as a

six-headed monster with teeth set in three rows living in a
cavern in the higher cliff and the latter as a monster that
sucks down the water creating a whirlpool living in a fig
tree dense with foliage in the lower cliff. They were regarded
as a sea hazard located close enough to each other that
they posed an inescapable threat to passing sailors avoiding Charybdis meant passing too close to Scylla and vice
versa. In that sense, Odysseus following Circes advice57
opted to pass by Scylla losing only a few sailors, rather
than risking the loss of his entire ship in the whirlpool: For
you, steersman, I have this order; so store it deeply in your
mind, as you control the steering oar of this hollow ship;
you must keep her clear from where the smoke and the
breakers are, and make hard for the sea rock lest, without
your knowing, she might drift that way, and you bring all of
us into disaster.58 In the words of Aristotle: For of the extremes one is more erroneous, one less so; therefore, since
to hit the mean is hard in the extreme, we must as a second best, as people say, take the least of evils.59
My claim is that the open texture of language allows
Hart to stand apparently somewhere in the center between
extremes arguing for a moderate indeterminacy, i.e. law is
sometimes determinate and sometimes indeterminate, and
against both a radical determinacy, i.e. law is always determinate or never indeterminate, and radical indeterminacy,
i.e. law is always indeterminate or never determinate. But
by characterizing the exercise of choice or discretion required to face the moderate indeterminacy as legislative,
Harts position as Odysseus crashes into one of the extremes, i.e. Scylla, the lesser evil, failing to achieve the mid56 Vid. Homer, The Odyssey, in The Iliad and the Odyssey of Homer, trans.
Richmond Lattimore, Chicago: The University of Chicago and Encyclopdia Britannica, 1952, Book XII, 73-259.
57 Ibidem, 108-10.
58 Ibidem, 217-21.
59 Aristotle, Nicomachean Ethics, supra note 8, Book II, Chapter IX, 1109a,
33-5, p. 963.
166
dle path, by suggesting that in those cases judges do legislate. The problem is whether there is a truly mean position,
i.e. a middle way.
In a few words, Hart affirms that a creative judicial activity60 is required to face the moderate indeterminacy and
certainly it implies discretion, but the problem is that he
equates creative to legislative and judicial discretion to
judicial legislation. In Harts voice:61
Laws require interpretation if they are to be applied to concrete cases, and once the myths which obscure the nature of
the judicial processes are dispelled by realistic study, it is
patent that the open texture of law leaves a vast field for a
creative activity which some call legislative. Neither in interpreting statutes nor precedents are judges confined to the
alternatives of blind, arbitrary choice, or mechanical deduction from rules with predetermined meaning. Very often their
choice is guided by an assumption that the purpose of the
rules which they are interpreting is a reasonable one, so that
the rules are not intended to work injustice or offend settled
moral principles At this point judges may again make a
choice which is neither arbitrary nor mechanical; and here
often display characteristic judicial virtues, the special appropriateness of which to legal decision explains why some
feel reluctant to call such judicial activity legislative.
I accept that judges do realize a creative judicial activity

not only by creating an individual norm to be applied to the
case at hand but also by creating at the same time a criteria or precedent of interpretation that may be applied to future cases.62 However, I reject that such creative judicial
activity amounts necessarily to a legislative one. By the
Hart, CL1, p. 131; and CL2, p. 134.

Hart, CL1, p. 200; and CL2, pp. 204-5.
62 Actually, legislatures, especially in the common law, delegate a limited authority to courts to the extent that it can be described as judge-made law through
interpretation but not necessarily as judicial legislation.
60
61
167
IMER B. FLORES
way, Hart mistakenly describes two types of creative or legislative activity:63

On the one hand, courts deciding a later case may reach an
opposite decision to that in a precedent by narrowing the rule
extracted from the precedent, and admitting some exception
to it not before considered, or, if considered, left open. This
process of distinguishing the earlier case involves finding
some legally relevant difference between it and the present
case, and the class of such differences can never be
exhaustively determined. On the other hand, in following an
earlier precedent the courts may discard a restriction found
in the rule as formulated from the earlier case, on the
ground that it is not required by any rule established by
statute or earlier precedent. To do this is to widen the rule.
I admit that both narrowing and widening the rule are

the product of a creative judicial activity resulting from
the interpretation of a pre-existing rule, but deny that it
amounts either to the legislative creation of a (new) rule or
to the quasi-legislative change of an existing rule. In both
cases, i.e. narrowing and widening, there is already an existing rule, whose scope is narrowed or widened through interpretation, but the rule is neither created out-of-the-blue
nor changed out-of-nothing-at-all through legislation or
quasi-legislation. Similarly, I argue that in cases where
there are legal gaps to be filled the judge does exercise an
interpretative creative judicial activity in order to declare
the existing law or more precisely to cover the gap with
pre-existing legal material, including principles and aims or
purposes. Whats more, when courts overrule a previous
decision, for example, Plessy v. Ferguson with Brown v.
Board of Education, it seems that they are not necessarily
making (new) law nor changing the existing law but recog-
63
Hart, CL1, p. 131; and CL2, p. 135 (emphasis added).
168
nizing a previous mistake in interpreting the law by amending or correcting it.64

Notwithstanding, Hart comeback in the Postscript regarding judicial discretion insists on judges having lawmaking powers and so judicial legislation:65
[I]n any legal system there will always be certain legally unregulated cases in which on some point no decision either
way is dictated by the law and the law is accordingly partly
indeterminate or incomplete. If in such cases the judge is to
reach a decision and is not, as Bentham once advocated, to
disclaim jurisdiction or to refer the points not regulated by
the existing law to the legislature to decide, he must exercise
his discretion and make law for the case instead of merely
applying already pre-existing settled law. So in such legally
unprovided-for or unregulated cases the judge both makes
new law and applies the established law which both confers
and constrains his law-making powers.
And, further, suggests that such legislative powers are

interstitial:66
It is important that the law-creating powers which I ascribe
to the judge to regulate cases left partly unregulated by the
law are different from those of the legislature: not only are
the judges powers subject to many constraints narrowing
his choice from which a legislature may be quite free, but
since the judges power are exercised only to dispose of par64 Vid. supra notes 32-4 and accompanying text. I am not denying that there is
a change in the state of affairs but affirming that it is corrective of a previous mistaken interpretation and as such neither legislative nor quasi-legislative, but interpretative. Consider not only Justice John Marshall Harlans dissenting opinion in
Plessy v. Ferguson but also his and Justice Oliver Wendell Holmes dissenting opinions in Lochner v. New York. In both cases, the dissenting opinions: first, denounced a honest mistake in the interpretation made by the majority, which even
though counted as law; and, later, became part of the prevailing interpretation
within the Supreme Court, which corrected the previous interpretation. Vid.
There Is No Caste Here and Room for Debate and for an Honest Difference of
Opinion, in Mark Tushnet (ed.), I Dissent. Great Opposing Opinions in Landmark
Supreme Court Cases, Boston, Beacon Press, 2008, pp. 69-80 and 81-92.
65 Hart, CL2, p. 272 (emphasis in the original).
66 Ibidem, p. 273 (emphasis in the original).
169
IMER B. FLORES
ticular instant cases he cannot use these to introduce

large-scale reforms or new codes. So his powers are interstitial as well as subject to many substantive constraints. None
the less there will be points where the existing law fails to
dictate any decision as the correct one, and to decide cases
where this is so the judge must exercise his law-making
powers.
The part of Harts rejoinder referring to constraints positions him really close to Kelsen and even to Kennedy.67
Moreover, Harts insistence on the nature of such law-making powers as interstitial, but legislative still,68 gets Hart
back over the shoulders of Holmes: I recognize without
hesitation that judges do and must legislate, but they can
do so only interstitially; they are confined from molar to
molecular motions.69 However, the fact of being interstitial
does not cancel it being legislative. As you can imagine, my
feeling is that Hart by appealing to a form of discretion
that equates creative to legislative misses an important
distinction and we seem to be in danger of missing too.
The distinction I have in mind is between interpretative
and inventive (or legislative) creative judicial activity,
which corresponds to a weak and strong forms of discre-
67 Actually, Kennedy in an attempt to separate himself and to some extent

the Critical Legal Studies movement from the American Legal Realist movement
has insisted that the indeterminacy he has in mind is not radical since adjudication involves both freedom and constraint coming in terms both with Harts and
Kelsens moderate indeterminacy. Vid., Kennedy, A Left Phenomenological Critique of the Hart/Kelsen Theory of Legal Interpretation, supra note 4, pp. 153-73;
Freedom and Constraint in Adjudication: A Critical Phenomenology, 36 Journal
of Legal Education 518 (1986), and reprinted in: Legal Reasoning..., supra note 4,
pp. 11-85 [references will be made to this version].
68 Vid. H. L. A. Hart, Introduction, in Essays, supra note 4, p. 6: [I]n any
modern legal system there must be many occasions where the settled law fails to
dictate a decision either way, so that if courts have to decide such cases they must
exercise a limited interstitial law-making power, or discretion.
69 Southern Pacific v. Jensen, 244 U.S. 205, 221 (1917) (Holmes, dissenting).
Vid. H. L. A. Hart, American Jurisprudence through English Eyes..., supra note
11, p. 128.
170
tion,70 i.e. between the weak discretion to interpret the (existing) law and even implicit principles and aims or purposes to be applied to the case at hand and the strong
discretion to invent (or legislate) the (new) law or even to
change the (existing) law.
After all, Hart acknowledged:71
[J]udges do not just push away their law books and start to
legislate without further guidance from the law. Very often,
in deciding such cases, they cite some general principle or
some general aim or purpose which some considerable relevant area of the existing law can be understood as exemplifying or advancing and which points towards a determinate
answer for the instant hard case.
Actually, reconsider Harts No vehicles in the park example. Imagine that someday a boy, called Freddie, who all
he wants to do is to learn how to ride his bicycle, comes
into a park with it and is prevented from riding it by the
park keeper, who points out to the prohibition. Suppose
that his mother Mrs. Mercury challenges the decision on
his behalf and reaches a point in which a judge with final
authority has to settle the dispute. It is clear that the word
vehicles is vague but bicycles are typically included in vehicles, but it is unclear whether the prohibition incorporates bicycles or not.
What is the judge expected to do? In other words: Is the
judge expected to invent (or legislate) a (new) law or to
change the (existing) law, acting as a legislator? Or is expected to interpret the law, appealing not only to principles
but also to aims and purposes? Alternately: Is the judge in
a position to exercise a strong legislative discretion to go either way? Or is expected to exercise a weak interpretative
70 I am aware that Dworkin introduced the distinction between weak and
strong forms of discretion in The Model of Rules, but I am distancing from his
use and trying to develop it in a way consistent with Hart and with a truly mean
position. Vid. Ronald Dworkin, Taking Rights Seriously, 2nd. ed., Cambridge, Massachusetts: Harvard University Press, 1978, pp. 31-9 and 68-71.
71 Hart, CL2, p. 274.
171
IMER B. FLORES
discretion to remain guided in the quest for the solution to

the case at hand? In my opinion, the judge is expected to
exercise a weak interpretative discretion appealing not only
to principles but also to aims and purposes. In that sense,
the prohibition can be said to be intended to protect the users of the park from certain forms of vehicles that might
cause a danger to them. The question is whether a boy riding a bicycle endangers the rest of the users of the park or
it is compatible with them. I believe that the judge will rule
that the law is that a bicycle per se does not jeopardize
them and, for that reason, the prohibition No vehicles in
the park does not apply to bicycles.72
V. CONCLUSION
Taking the distinction seriously implies that judges, instead of exercising a strong inventive (or legislative) discretion pushing their law books aside, instead do and if I
may add must exercise a weak interpretative discretion
by appealing not only to principles but also to aims and
purposes in the quest for further guidance. The problem for
Hart and his followers will still be that this mean position, i.e. creative interpretation, in between the extremes
represented by non-creative interpretation, i.e. deductive or
mechanical application, and creative legislation, does echoes his archenemy as Hart himself acknowledged:73
This indeed is the very nucleus if the constructive interpretation which is so prominent a feature of Dworkin theory of
adjudication. But though this procedure certainly defers, it
does not eliminate the moment of judicial law-making, since
72 Let me advance that the weak interpretative discretion proposed in this paper as a way of reconsidering the moderate indeterminacy of law thesis by appealing not only to principles but also to aims and purposes implies reconsidering the
two other thesis identified at the outset of this paper, namely: (1) the concept of law
as a (complex) model of rules; and (2) the separation of law and morals. Again, I am
indebted with Ken Himma for pushing me to clear my thoughts in order to make
explicit my claims and grounds for them, especially in this section.
73 Hart, CL2, pp. 274-5.
172
in any hard case different principles supporting competing

analogies may present themselves and a judge will often
have to choose between them, relying like a conscientious
legislator, on his sense of what it is best and not on any already established order of priorities prescribed for him by
law. Only if for all such cases there was always to be found
in the existing law some unique set of higher-order principles assigning relative weights or priorities to such competing lower-order principles, would the moment for judicial
law-making be not merely deferred but eliminated.
173
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
ARE LEGAL RULES CONTENT-INDEPENDENT

REASONS?
Noam GUR*
RESUMEN:
En este artculo se argumenta que la respuesta a la pregunta de si las
normas jurdicas son razones independientes de contenido depende de
tres distinciones que demarcan cmo las normas jurdicas son razones
independientes de contenido en algunos sentidos, pero no en otros. La
primera distincin es entre dos sentidos de independiente de contenido
que el autor refiere como fuerte y dbil. En el argumento se sostiene que,
si bien las normas jurdicas generan razones independientes de contenido en el sentido dbil, el hecho de que generen razones independientes
de contenido en el sentido fuerte depende, a su vez, de otras dos distinciones: primero, una distincin entre enunciados evaluativos y descriptivos de razones; y segundo, una distincin entre razones para la accin y
razones para adoptar ciertas actitudes. La independencia de contenido
fuerte contina el argumento es una nocin consistente slo en la
medida en que se formule con enunciados descriptivos de razones (a diferencia de enunciados evaluativos de razones) que se relacionan con acciones (en contraste con las actitudes). Finalmente, se enuncia una tesis
implcita que vincula los diferentes sentidos en que las normas jurdicas
son razones independientes de contenido, al mismo tiempo que explica
sus diferencias.
* Shaw Foundation Fellow in Law, Lincoln College, Oxford. I am grateful for

helpful comments and suggestions of the participants in this symposium, in a Legal Normativity workshop at the IVR Conference 2011, in a jurisprudence workshop at the SLS Annual Conference 2011, and in the Oxford Jurisprudence Discussion Group. I am also grateful to Fred Schauer for helpful comments on an
earlier draft of this paper.
175
NOAM GUR
Palabras clave:
Razones independientes de contenido, normas jurdicas, normatividad, H. L. A. Hart, Joseph Raz.
ABSTRACT:
I argue that the answer to the above question turns on three distinctions
which make it clear that legal rules are content-independent reasons in
some senses, but not in others. The first distinction is between two senses
of content-independence, which I refer to as weak and strong content-independence. I argue that, while legal rules do give rise to content-independent
reasons in the weak sense, whether they can be said to generate content-independent reasons in the strong sense depends on two further distinctions: first, a distinction between evaluative and descriptive statements
about reasons; second, a distinction between reasons for action and reasons for adopting certain attitudes. Strong content-independence, I argue,
is a sound notion only insofar as it figures in descriptive reason-statements
(as opposed to evaluative reason-statements) with regard to actions (as opposed to attitudes). Finally, I uncover an underlying explanation that links
the different senses in which legal rules are content-independent reasons,
and accounts for the differences between them.
Keywords:
Content-Independent Reasons, Legal Rules, Normativity, H. L.
A. Hart, Joseph Raz.
176
ARE LEGAL RULES CONTENT-INDEPENDENT REASONS?
SUMMARY: I. Introduction. II. Strong and Weak Content-Independence. III. Does Law Provide Weakly/Strongly
Content-Independent Reasons? IV. A Discrepancy
with Good Reason. V. Conclusion.
I. INTRODUCTION
Since introduced by H. L. A. Hart and espoused by Joseph
Raz, content-independent reasons have become part and
parcel of the conceptual vocabulary of legal theorists.1 Yet
the range of reactions provoked by this notion manifests a
serious dissonance in jurisprudential thought. On the one
hand, several theorists have acknowledged that the notion
successfully captures a key feature of laws normativity and
authoritative nature.2 On the other hand, as one writer put
it, few believe I can have a reason to perform an action
simply and only because the law to which I am subject says
I must3 a denial that appears intuitively plausible yet
seems to stand in direct opposition to the idea of content-independence. In this paper, I argue that the key to resolving this dissonance lies in three distinctions: (1) a dis1 Hart, H. L. A., Essays on Bentham, Oxford, Clarendon Press, 1982, pp. 18,
243-68; Raz, Joseph, The Morality of Freedom, Oxford, Clarendon Press, 1986, pp.
35-7. See further Green, Leslie, The Authority of the State, Oxford, Clarendon Press,
1990, pp. 41-62, 225-6; Duff, R. A., Inclusion and Exclusion, Current Legal Problems, vol. 51, 1998, pp. 241-66, at p. 247; Himma, Kenneth E., Hart and the Practical Difference Thesis, Legal Theory, vol. 6, 2000, pp. 1-43, at pp. 26-7; Shapiro,
Scott, Authority, in Coleman, Jules L. and Shapiro, Scott (eds.), The Oxford
Handbook of Jurisprudence and Philosophy of Law, Oxford, Oxford University
Press, 2002, p. 389; Christiano, Thomas, Authority, in Stanford Encyclopedia of
Philosophy, <http://plato.stanford.edu/entries/authority/>, accessed 1 March
2011 (referring mostly to content-independent duties); Schauer, Frederick, Authority and Authorities, Virginia Law Review, vol. 94, 2008, pp. 1931-61, at pp.
1935-6.
2 See, e.g., Green, supra note 1, pp. 41-62, 225-6; Schauer, supra note 1, pp.
1935-6; Christiano, supra note 1.
3 Markwick, Paul, Independent of Content, Legal Theory, vol. 9, 2003, pp.
43-61, at p. 57; Markwick, Paul, Law and Content-Independent Reasons, Oxford
Journal of Legal Studies, vol. 20, 2000, pp. 579-96, at p. 586 (though this statement
of Markwick is merely incidental to his main objections against content-independent reasons). See further Edmundson, William A., Three Anarchical Fallacies, Cambridge, Cambridge University Press, 1998, pp. 12-4.
177
NOAM GUR
tinction between two understandings of content-independence, which I refer to as weak and strong contentindependence; (2) a distinction between evaluative statements about reasons and descriptive statements about reasons; and (3) a distinction between reasons for action and
reasons for adopting certain attitudes. Drawing on these
three distinctions, my analysis leads to a nuanced answer
to the question of whether rules (and other requirements) of
law provide content-independent reasons: in some senses
they do, in others they do not.4 In the final part of the paper, I uncover an underlying rationale that connects and
explains the disparity between the senses in which law does
and does not provide content-independent reasons.
II. STRONG AND WEAK CONTENT-INDEPENDENCE
Let us begin with the relevant ambiguity in the notion of
content-independence. Pertinent to my analysis are two
possible renderings of content-independence: according to
the first rendering, to say that the reasons law generates
are content-independent is to say that their status as reasons does not depend on the question of what action a
given legal provision requires, but rather is attributable to
the fact that law requires the action. It will be noticed that
this rendering consists of both a negative element (i.e. does
not depend on the question of what action a given legal provision requires) and a positive element (i.e. is attributable
to the fact that law requires the action). This sense of content-independence, which will be referred to as the strong
4 Phrasing the question this way, i.e. asking whether rules provide (or generate, supply, give rise to, etc.) content-independent reasons, is analytically more
correct than the formulation used in the title, i.e. whether rules are content-independent reasons, since the latter formulation is pleonastic: to say that rules are
reasons is to say that they themselves (or the fact of their enactment) are reasons,
rather than that their content happens to coincide with what we anyway have reasons to do. Thus, if rules are reasons they are, eo ipso, content-independent reasons. In light of this, I will be mostly using terms such as provide/generate/give
rise to content-independent reasons, and even when using different terminology I
should be taken to mean the same.
178
sense, appears to correspond with

characterization of this notion, as follows:
Harts
basic
Content-independence of commands lies in the fact that a

commander may issue many different commands to the
same or to different people and the actions commanded may
have nothing in common, yet in the case of all of them the
commander intends his expressions of intention to be taken
as a reason for doing them. It is therefore intended to function as a reason independently of the nature or character of
the actions to be done.5
Joseph Raz, at one juncture, provides a similar explanation of content-independence:

A reason is content-independent if there is no direct connection between the reason and the action for which it is a reason. The reason is the apparently extraneous fact that
someone in authority has said so, and within certain limits
his saying so would be reason for any number of actions, including (in typical cases) for contradictory ones. A certain
authority may command me to leave the room or to stay in
it. Either way, its command will be a reason. This marks authoritative reasons as content-independent.6
It can be readily seen that these passages refer both to

what I called the negative element (Hart: independently of
the nature or character of the actions to be done; Raz: no
direct connection between the reason and the action for
which it is a reason) and to what I called the positive element (Hart: the commander intends his expressions of intention to be taken as a reason; Raz: [t]he reason is the
apparently extraneous fact that someone in authority has
said so7). Similar expressions of the strong sense of conHart, supra note 1, p. 254.
Raz, supra note 1, p. 35.
7 This is not the only possible understanding of Razs comment: it is conceivable that his words someone in authority are intended to mean someone who
meets the substantive prerequisites for being a legitimate authority, which im5
6
179
NOAM GUR
tent-independence can be found in statements of several

other legal theorists, such as Leslie Greens comment that
[t]he core idea [of content-independence] is that the fact
that some action is legally required must itself count in the
practical reasoning of citizens, independently of the nature
and merits of that action,8 and Scott Shapiros remark that
[o]ne who obeys a command treats the command as a content-independent reason, because he complies for the reason that he was commanded, not because he has reasons
to act on the content of that command.9
An alternative rendering of content-independence is
what I will refer to as the weak sense of content-independence. According to this rendering, the content-independence of, for instance, a law-given reason means that it does
not derive its status as a reason from the content of a given
legal provision, i.e. from the nature and merit of the action
law requires. However, by this understanding of content-independence, the reason is not attributable to the mere fact
that law requires the action. Rather, it depends on substantive factors (other than the content of the legal requirement), viz. attributes that a lawmaking institution may possess, such as the fact that legal requirements in the
relevant jurisdiction ensue from a fair and just decision-making procedure; or that the relevant lawmakers are
well placed (relative to private actors) to determine what
ought to be done in certain situations or domains of activity; or that legal authorities, due to their salience and enforcement measures, are apt to facilitate beneficial coordination between people; or other possible attributes of a
plies that the reason for action is not the mere fact that law requires it. This construction arguably finds further support in Razs qualified words no direct connection instead of simply no connection. See also my caveat in note 9.
8 Green, supra note 1, p. 225.
9 Shapiro, supra note 1, p. 389. I do not make the claim that (some or all of) the
above authors necessarily endorse, or would subscribe to, the strong version of
content-independence or to the idea that law is such a content-independent reason. I only cite the above statements as examples which suggest that this sense of
content-independence figures in the literature and that content-independence can
be understood in this way.
180
lawmaking institution which, when present, militate in favor of compliance with its requirements (these will be frequently referred to as substantive considerations for
law-compliance, law-following values, or the like). A clear expression of this sense of content-independence may be
found, for instance, in the following remark of Raz:10
Take a rule saying that members [of a chess club] are entitled to bring no more than three guests to social functions of
the club. The considerations which establish that it is binding do not turn on the desirability of members having a
small number of guests, nor on the desirability of members
having the option to bring guests, but on the desirability of
the affairs of the club being organised by the committee
which laid down the rule [which he attributes to evaluative
considerations such as a recognition that the harm that
would be occasioned by a disorderly attempt to overturn the
rules of the committee is too great]. It is, in other words, an
instance of what I call (following Hart) a content-independent
justification. It is content-independent in that it does not
bear primarily on the desirability of the acts for which the
rule is a reason. [T]he considerations which show why the
rule is binding, i.e. why it is a reason for not bringing more
than three guests, do not show that it is good not to bring
more than three guests. They show that it is good to have
power given to a committee, and therefore good to abide by
decisions of that committee. But that can justify a variety of
rules. They are, in this sense, content-independent.11
10
Once more, I do not wish to make an exegetical claim that attributes to Raz a
general endorsement of this sense of content-independence. I only quote this comment of his as an example of a statement that comports with that sense.
11 Raz, Joseph, Reasoning with Rules, Current Legal Problems, vol. 54, 2001,
pp. 1-18, at pp. 8-9. Other statements consistent with the weak sense are made,
for example, by Duff, supra note 1, p. 247 (referring to content-independent reasons as reasons which justify her [i.e. a ruler] in demanding obedience from the
people at allreasons which give her the right to issue commands or rules, and
to back them by different threats) and by Thomas Christiano in The Authority of
Democracy, The Journal of Political Philosophy, vol. 12, 2004, pp. 266-90 (p. 267:
[I]f a process is genuinely democratic justice demands that individuals comply
with the decision-making process. The demand is content independent: compliance is required regardless of the content of the democratic decision ; p. 287:
[O]nly by obeying the democratically made choices can citizens act justly. Demo-
181
NOAM GUR
While this explanation, too, contains both a negative and

a positive element, its positive element is different from that
which appeared in the previously quoted statements.12 Here
content-independent considerations are said to consist not
simply and only in the fact that an authorized ruler has required an action, but rather in the desirability of allowing
him to regulate the kind of affairs concerned. Now, I should
note that, of the two alternative renderings I set out, only
the latter is warranted by the label content-independent
reasons in itself. The only type of independence that the label refers to is content-independence, where content is naturally understood as the content of a requirement; and this
is exactly the type of independence I called weak. The
strong type of independence, by contrast, envisages reasons
that are independent of any substantive considerations or
value judgment, whether content-based or other. The
proper name for this idea would be not content-independence, but rather substance- or value-independence. Nevertheless, it would be wrong to rule out of consideration the
strong rendering of content-independence, if only for the
fact that it corresponds with the way content-independence
has been understood and used by many legal theorists.
Another point that bears clarification at this stage is that
neither the weak nor the strong sense of content-independence entail that when legal rules apply to us the only reasons we have are content-independent. Thus, it is not denied by any conception of content-independence that,
insofar as the content of a law coincides with moral standards, some of the reasons for compliance with the law consist in the moral worth of what it requires.13 The claim we
are considering, therefore, is that legal requirements provide content-independent reasons, not that content-inde-
cratic directives give content-independent reasons since one must accept a democratic decision as binding even when one disagrees with it).
12 Text accompanying notes 5-9.
13 Himma, supra note 1, p. 26.
182
pendent reasons are exhaustive of the reasons applying to

laws subjects.
So far we have seen that there exist two alternative
senses of content-independence: a weak and a strong one.
This means that our initial question whether law provides
content-independent reasons actually envelops two different questions: (1) whether law provides content-independent reasons in the weak sense; (2) whether law provides
content-independent reasons in the strong sense. I now
turn to these questions.
III. DOES LAW PROVIDE WEAKLY/STRONGLY CONTENTINDEPENDENT REASONS?
The first of the two questions just listed appears to be the
easier. For there is nothing problematic or perplexing in
saying that the array of reasons that apply to laws subjects
may include not only reasons grounded in the content of an
applicable rule, but also considerations associated with
some qualities of rules themselves, of the institutions that
make them, or of the procedure from which they ensue.14
Thus, for instance, a recognition that the relevant lawmaking institution is better placed than a private actor to decide on certain matters, the conduciveness of legal rules to
social order, stability and coordination, and their capacity
to simplify and expedite daily decision-making, are all factors that when applicable count in favor of following legal
rules. To use a concrete and familiar example: a rule requiring people to drive on the right presents them with a
compelling coordination-related reason for doing so, which
is independent of any value attached to driving on the right
per se and would equally apply if the rule were to require
14 A question may arise, however, as to whether some of these reasons are excluded by the rule I have discussed this question in Gur, Noam, Legal Directives
in the Realm of Practical Reason: A Challenge to the Pre-Emption Thesis, American Journal of Jurisprudence, vol. 52, 2007, pp. 159-228.
183
NOAM GUR
driving on the left.15 This, of course, is just one, clear example of a coordinative rule, and there are many other types of
legal rule that involve a coordinative dimension, contribute
to determinacy and certainty in social life, and have other
desirable aspects which are not specifically tied with the
particular content adopted values which compliance with
the rule tends to protect and promote. It seems, therefore,
that the answer to our first question, whether legal requirements give rise to content-independent reasons in the weak
sense, is that they can and frequently do.16
Do legal requirements constitute content-independent
reasons in the strong sense (i.e. reasons attributable, not to
their content, but to the mere fact of their being legal)? It is
tempting to respond with a simple no. For one cannot
help but wonder: how can the mere fact that an action is
required by a lawmaker, his mere say-so, constitute a reason to act in and of itself? Human beings cannot create reasons out of thin air simply by requiring that others act in
some way. An utterance of a lawmaker, the thought may
continue, is no more than an artifact of the human will
which cannot by itself turn a false moral proposition into a
true one or determine by way of stipulation what is wrong
and what is right.17 It can only have normative bearing
when and insofar as it appears in conjunction with some
non-artificial, evaluative factors which make it the case
that, and explain why, it is good to follow the law factors
of the type mentioned in the preceding paragraph.18
15
True, this reason depends on the fact that the context of regulation is such
that involves a coordination problem, but it remains independent of the particular
content chosen by the legislature (within a certain range of possibilities). So, although this coordination reason depends on the regulated context, it is (in a limited, but important sense) content-independent.
16 That is, they do so when and insofar as the lawmaking procedure or institution has the attributes that make it desirable to follow legal directives.
17 A similar argument is mentioned by Hart, supra note 1, p. 265 (though he
mentions it as part of a more moderate objection to his conception of authoritative
reasons).
18 Idem.
184
The problem with this argument is not so much what it

says, but what it ignores. It overlooks a pertinent distinction between two ways of using the term reason: namely, a
distinction between descriptive statements about reasons
and evaluative statements about reasons. By descriptive
statements about reasons, I mean statements describing
the reasons that figure in peoples practical deliberation; or,
more particularly, the kind of reasons that participants in a
certain normative practice, such as law, characteristically
deem themselves to be given by the requirements of that
practice. Thus, in our context, a descriptive statement referring to reasons of type X conveys that within the boundaries of a particular normative practice actors regularly operate with an assumption that the practices requirements
provide them with reasons of type X. In contrast, evaluative
statements about reasons, rather than describing the reasons that people take themselves to have, evaluate what
reasons actually apply to them (with or without their
knowledge of those reasons). Such statements ensue from a
mode of inquiry that refuses to stop at the identification of
the normative assumption prevalent among participants in
the relevant practice, but rather looks beyond that assumption and seeks an ultimate and substantive explanation of
what, if anything, makes it a correct assumption.19 In this
sense, evaluative statements about reasons are committed,
not to the viewpoint of the normative practice in question,
but rather to the viewpoint of the normative universe at
large.20
19
It might be argued that descriptive reason-statements are no more than parasitic on evaluative reason-statements, and that we should therefore concern ourselves solely with the latter. This argument is misguided not only because the
question of how law figures in subjects practical reasoning is an important question in its own right, but also because, as will become apparent in Section IV, the
term parasitic grossly oversimplifies what is in fact a complex relation between
the descriptive and evaluative dimensions.
20 A few caveats should be made here: first, these evaluative statements concern only reasons that apply to participants in a given legal practice qua participants. They do not refer, for instance, to reasons applicable in the state of nature
or to reasons that have nothing to do with law. Second, the evaluative perspective I
185
NOAM GUR
Against this background it becomes apparent that there

are two different ways in which strong content-independence can be spoken of: it can be used in a descriptive reason-statement, claiming that the legal practice is such that
its participants typically treat its requirements as strongly
content-independent reasons (and it bears noting here that
the descriptive language is the one in which Harts arguments about strong content-independence are usually
couched21). Or it can be used in an evaluative reason-statement, claiming that legal requirements are strongly content-independent reasons from the viewpoint of the normative universe at large, not merely in the assumption of the
legal practice. Now, the objection earlier posed against
strong content-independence clearly presupposes and takes
as its target the second, evaluative statement. And indeed
the idea conveyed by this statement that a word mouthed
by a human agent can, by itself, create a reason to do what
he says without there being any substantive factors that
make it desirable to follow him is clearly untenable. And
even when it comes to legal rules whose formation is associated with social facts other than the utterance or will of a
human legislator, such as social customs, it would be an
unwarranted leap (from a non-evaluative premise to an
evaluative conclusion) to say that the mere fact of their existence constitutes reasons to act in the absence of substantive factors that show why it is good or morally right to
comply with them. Evaluative statements about reasons
necessarily depend for their correctness on the underpinning force of such substantive factors. Thus, the only type
of content-independence that can correctly figure in an
refer to can be entertained by participants themselves, even if this is not likely to
occur in the usual course of their practical operation, but in more reflective modes
of deliberation or conversation. Third, despite its readiness to transcend the normative assumptions of the legal practice, this evaluative perspective is entirely different from the type of external viewpoint that Hart famously criticized, namely
that which overlooks the normative dimension of law and focuses solely on regularities of behavior.
21 Hart, supra note 1, pp. 18, 243-68.
186
evaluative statement is one that allows for such factors

i.e. the type of content-independence I entitled weak.
Before turning to consider the descriptive statement
about strong content-independence, I should make a couple of caveats in this regard: first, my arguments will not
purport to describe how all or even virtually all participants
in the legal practice treat its requirements, but rather will
only claim to identify salient attitudes that seem to characterize at least a significant proportion of participants in typical legal systems. Second, I will not adduce experimental
or survey-based empirical evidence with regard to these descriptive claims, but rather will confine myself to
phenomenological observations and anecdotal evidence
from ordinary human experience. This, I admit, limits the
probative force of these descriptive claims, which will therefore be offered here as merely provisional observations
rather than conclusive findings. With these caveats in
mind, let us consider the notion of strong content-independence at the descriptive level: do participants in the legal practice regularly treat its requirements as providing
them with content-independent reasons in the strong
sense, i.e. as providing them with reasons by virtue of being
legal? At first glance, the answer appears to be a simple
yes. Consider, for example, the way people intend and understand a sign that says Buckle up! Its the law or It is
against the law to smoke in these premises. The formulator of this sign intended to present its readers with a reason
to act in the specified manner and this is what readers of
the sing commonly understand. To be thus understood all
that the formulator said, and needed to say, is this: Its the
law or It is against the law. This way of perceiving law is,
of course, not limited to communication through signs.
Thus, for instance, if an official in charge of enforcing a
smoking ban in certain public spaces were to be asked
why? when fining a violator and demanding him to stop,
his response may well be: because you are breaking the
law; and if an ordinary citizen were to be asked, upon
187
NOAM GUR
stopping her car at a red traffic light, why she had stopped,
her (surprised) response would probably be simply to point
at the red light or to say that this is what the law requires.
Indeed, that participants in the legal practice, who share
its internal point of view, regularly treat laws in the way
just exemplified is a point famously made by Hart in The
Concept of Law.22 And although, for Hart, a society may
have a legal system even if its private citizens, as opposed
to public officials, do not share the internal point of view,
he nonetheless referred to a society in which they largely do
not as an extreme case, and maintained that in normal societies citizens will often assume it.23 For such citizens, as
Hart noted, the red light is a reason for stopping in conformity to rules which make stopping when the light is red
a standard of behaviour and an obligation,24 and when a
legal official punishes law violation, according to Hart, he
takes the rule as his guide and the breach of the rule as
his reason and justification for punishing.25 Of course,
these statements of Hart were not primarily concerned with
the idea of content-independence, but rather more generally sought to highlight the normative dimension of law and
counter the type of reductionism that analyzes the legal
practice in terms of behavioral regularities alone. Nevertheless, by recognizing that legal rules are regarded as reasons
while omitting reference to the merits of the required action
or to substantive law-following values, Hart lends implicit
support to what he elsewhere explicitly calls content-independence, at least as far as it figures in descriptive reason-statements.26
22 Hart, H. L. A., The Concept of Law, 2nd. ed., Oxford, Oxford University Press,
1994, pp. 11, 55-6, 89-90.
23 Ibidem, p. 116-7.
24 Ibidem, p. 90.
25 Ibidem, p. 11 (emphases omitted). See also related comments at ibidem, p. 55.
26 Hart, supra note 1, pp. 243-68. Hart acknowledges that normative acceptance of law may (though need not) be based on ones recognition of the moral value
of the system (ibidem, pp. 256, 265-6; Hart, supra note 22, pp. 203, 257). The distinction I will draw between actions and attitudes (in text accompanying notes
188
But even this descriptive statement about (strong) content-independence may be challenged. It may be challenged
on grounds that were usefully highlighted by one of Neil
MacCormicks arguments regarding Harts account.27
MacCormick acknowledges that participants in the legal
practice who share the internal point of view typically take
its rules to provide them with reasons for action. Conversely to Hart, however, he stresses that acceptance of legal rules from the internal perspective is not unreasoned,
but rather must be, or at least normally is, underpinned by
a judgment on the part of participants about the value and
justice of the system: e.g. their recognition that the system
beneficially contributes to certainty and predictability, that
the power exercised by lawmakers derives legitimacy from
the fact of their being democratically elected, and that laws
are enacted through a procedure which is by and large fair
and just.28 If, as MacCormick observes, such substantive
factors undergird the acceptance of rules by their subjects,
this may be thought to show that rules are not treated as
content-independent reasons in the strong sense, i.e. in the
sense that signifies independence of those substantive factors. Now, I will accept the kernel of MacCormicks claim,
but will suggest that its effect is merely to qualify, rather
than utterly to refute the notion of strong content-independence as a description of how law operates in peoples
30-37) can reconcile such morally based acceptance of law with Harts endorsement of strong content-independence: Hart can be taken to mean that even when
peoples law-abiding attitude is underpinned by a moral judgment, it tends to influence their actions independently of a case-by-case assessment of law-following
values, which means that law operates in their practical reasoning as a strongly
content-independent reason for action.
27 MacCormick, Neil, Legal Reasoning and Legal Theory, Oxford, Clarendon
Press, 1978, pp. 63-4, 139-40. See generally Smith, J. C., Legal Obligation, London,
Athlone, 1976, pp. 22-33; Dworkin, Ronald M., Taking Rights Seriously, London,
Duckworth, 1978, pp. 51-8; Finnis, John M., Natural Law and Natural Rights, Oxford, Clarendon Press, 1980, pp. 13-4; Finnis, John M., On Hart's Ways: Law as
Reason and as Fact, American Journal of Jurisprudence, vol. 52, 2007, pp. 25-53;
Perry, Stephen R., Hart's Methodological Positivism, in Coleman, Jules L. (ed.),
Hart's Postscript, Oxford, Oxford University Press, 2001, pp. 330-47.
28 MacCormick, ibidem, pp. 63-64, 139-40.
189
NOAM GUR
practical reasoning. What ought to be accepted as plausible

is that peoples normative adherence to legal rules is partly
underpinned by a recognition that the system helps secure
at least basic goods in social life and conforms to principles
of justice at least to a minimal degree that makes it acceptable.29 Consistently with MacCormick, it seems to me highly
unlikely that people could honestly come to believe that legal rules give them normative reasons without this belief
being influenced or validated by any notion that there are
some actual reasons to follow the law.30 For people, even
those in possession of only the most basic level of human
rationality and understanding, can hardly fail to comprehend what was earlier observed here as a straightforward
verity: namely, that reasons to act cannot be magically
brought into existence by the mere word of an agent without there being some substantive factors that make it desirable to follow his word.31
Nonetheless, it seems to me that the role played by
law-following values in subjects practical reasoning is not
entirely inconsistent with the idea of strong content-independence. Rather, it only shows that this idea ought to be
qualified. The reason for this is that, although law-following
values seem to play a part in leading subjects to adopt the
general attitude they normally have towards law i.e. their
law-abiding disposition, the attitude of treating law as a
29
Which is compatible with peoples being dissatisfied with certain aspects of
the system, its officials, or the political party in power.
30 Hart mentions MacCormicks objection without discussing it (Hart, supra
note 22, p. 243). He seems to retain his position that allegiance to the system need
not stem from a moral judgment, but may be based on many different considerations: calculations of long-term interest; disinterested interest in others; an unreflecting inherited or traditional attitude; or the mere wish to do as others do (ibidem, pp. 203, 257; Hart, supra note 1, pp. 256-7). For reasons explained in the
main text above, I think these remarks of Hart fail to meet MacCormicks challenge. A better response to MacCormick can be given by recourse to the actions/attitudes distinction see text accompanying notes 30-37.
31 This would, inter alia, involve attributing to people the assumption that
there were two independent worlds or sets of objective reasons, one legal and one
moral which even Hart refers to as an extravagant hypothesis (Hart, supra
note 1, p. 267).
190
reason for action the influence of that attitude on their

actions does not depend on whether law-following values
apply to the particular situation they face. This requires an
explanation: We sometimes use the phrase a law-abiding citizen. When employing this phrase, obviously we are not referring to an action, not even to a series of actions, but
rather we are characterizing an agent. We are pointing out,
more specifically, a certain attitude of the person in question: a law-abiding disposition, an attitude whose possessors generally treat laws requirements as reasons for them
to act accordingly. What does the attitudinal dimension implied here add and how precisely does it differ from simply
speaking of actions? The attitudinal dimension suggests (or
at least allows for the possibility) that the law-abiding disposition referred to is a well-entrenched and deep-seated
feature of its possessors which tends to endure through
time, rather than being just a momentary response to a
particular situation.32 It suggests that, once adopted, the
disposition exerts its influence on its possessors actions independently of a case-by-case assessment of the reasons
that led to its adoption, and that it may thus make its force
felt even when those reasons are absent. Once adopted, in
other words, it exerts its force independently of whether
substantive considerations in favor of following the law apply to the particular case at hand (and this, it bears clarification, is consistent with saying that the force of the disposition remains overridable by substantive contrary reasons,
rather than absolutely and exclusively determinative of
ones actions33).
32
I say tends because dispositions need not be permanent; people sometimes
abandon or change some of their dispositions. Doing so, however, may be difficult
and normally takes time. It is not done on a case-by-case basis.
33 To see this, it may be helpful to think of this law-abiding attitude in terms of
an inclination: on the one hand, it is not a simple, constantly malleable reflection of
substantive reasons for and against compliance as applicable in specific cases if
it were only that, there would be no point or distinct significance to calling it an inclination. On the other hand, being no more than an inclination, it stops short of
excluding relevant case-specific reasons and remains overridable by such reasons.
191
NOAM GUR
Before I turn to explicate how precisely the attitude/action distinction bears on our question, it may be useful to
shed more light on the distinction itself by reference to an
example. I said that the envisaged law-abiding disposition
makes its influence felt independently of whether the reasons for having it, i.e. law-following values, apply to a particular action in a particular situation. One implication of
this is that those who are thus disposed to comply with law
may be led to perform actions which they would not perform if they were guided solely by the balance of reasons for
action. Consider, as a case in point, the frequently invoked
hypothetical of a traffic light in the desert.34 It runs (with
certain variations) along the following lines: you are driving
on a lonely road in the desert. You come to an intersection
where there is a red traffic light. The law prohibits driving
through a red light, but with a perfectly flat and desolate
landscape around you, you can tell that there are no other
vehicles or pedestrians within miles in any direction. It is
evident that an act of non-compliance on your part will not
be seen or discovered by anyone else. So it will not result in
your being punished and will not stimulate other actors
into disobedient behavior. Also suppose that given how
astute and strong-willed you are, or given other features of
your personality or condition running a red light in this
special situation will not weaken your resolve to comply
with the law in other circumstances. As many philosophers
acknowledge, in this type of situation there is no real reason (whether safety, coordination, or other) for you to stop
and wait for the light to turn green.35 And even if some
34 See Raz, Joseph, The Authority of Law, Oxford, Clarendon Press, 1979, p. 16;
Smith, M. B. E., Is There a Prima Facie Obligation to Obey the Law, Yale Law
Journal, vol. 82, pp. 950-76, at p. 971; Regan, Donald H., Law's Halo, Social Philosophy and Policy, vol. 4, 1986, pp. 15-30, at pp. 18-9; Hurd, Heidi M., Challenging Authority, Yale Law Journal, vol. 100, 1991, pp. 1611-77, at p. 1614;
Edmundson, supra note 3, pp. 12-34.
35 See, e.g., Smith, ibidem, p. 971 and Regan, ibidem, pp. 18-9. See also Alexander, Larry, Law and Exclusionary Reasons, Philosophical Topics, vol. 18, 1990,
pp. 5-22, at p. 8 (noting that in situations where I predict no effect on others behavior, no detection of my disobedience, and no sanctions, my reasons against A
192
such reason for action could be said to apply to the case it

would be too remote and weak to rationalize an act that is,
in all other respects, sharply inconsistent with common
sense: stopping and waiting when there is absolutely no
traffic and not a single soul for miles around. If so, assuming that your action is guided purely by a correct assessment of the balance of reasons for action, you will violate
the law in these circumstances. The case may well be different, however, if you act with a general disposition to comply
with law. If such a disposition characterizes your attitude
towards law, and if it is sufficiently forceful and deepseated, it is likely to lead you to stop and wait for the light
to change after all.36
At this point an objection might be raised against the
premise that the supposed driver in the desert-traffic-light
[where A is a legally prohibited action] remain exactly as they were before the law
was enacted). Raz notes with regard to a similar example that [m]any will say that
there is not even the slightest reason to stop at the red light in such circumstances (ibidem, p. 16). Without explicitly committing himself to the above view,
he says that [t]his example seems sufficient to convince one that in this case or a
similar case the utterance of authority can be held to be legitimate without holding
them to constitute reasons for action (idem).
36 These remarks are consistent with saying (as I will in Section IV) that people
have good reason to adopt the envisaged law-abiding disposition (even if it would occasionally lead them to perform actions they have no reason to perform). Moreover, I
do not wish to deny that there is a certain sense in which an act of compliance in the
desert-traffic-light situation could be described as a rational act (though that sense
is not the one contemporary philosophers commonly use when describing an act as
rational): if an agent has acquired a disposition, such that it is rational to acquire,
and this disposition leads her to comply with law, inter alia, in the desert-traffic-light
situation, one might suggest that her act of compliance is rational in the sense that
it ensues from a disposition whose acquisition was rational (see analogous arguments in Gauthier, David, Afterthoughts, in MacLean, Douglas (ed.), The Security
Gamble, Totowa, NJ, Rowman & Allanheld, 1984, p. 159; Gauthier, David, Rethinking the Toxin Puzzle, in Coleman, Jules L. and Morris, Christopher W. (eds.), Rational Commitment and Social Justice, Cambridge, Cambridge University Press, 1998, p.
47; Pink, Thomas, The Psychology of Freedom, Cambridge, Cambridge University
Press, 1996, p. 93). But this sense of rational act does not mean that the act is supported by the balance of reasons for action. Normative reasons for action, as I understand them, are facts linked with qualities of the action itself or its consequences (including consequences it has qua an act of compliance or non-compliance). To
establish that one has a reason to perform a certain action, it must be shown how
performance of that specific action, given the specific circumstances in question,
would serve a specific value. The mere fact that an action ensues from a rationally
adopted disposition is not a reason for action.
193
NOAM GUR
situation has little or no reason to stop. The objection I

have in mind contends that the driver would be right to
reason as follows: A. It is good (for reasons such as the desiderata of social order, coordination, etc.) generally to obey
the law in a society like ours. B. The law requires people
not to drive through a red light. C. The conjunction of A
and B entails that I have a good reason to stop at the red
light in front of me. The problem with this three-step reasoning is that it tars with too broad a brush: to establish
that there are real reasons for action (i.e. the action of stopping) in the desert-traffic-light situation, it must be shown
that the considerations which are said to make it good
generally to obey the law apply in that situation, i.e. that
an act of obedience in that specific situation would genuinely serve those substantive values. This, however, sends
us back to the previous paragraph, where I pointed out the
lack of rational basis for an act of obedience in that situation. In fact, the objectors readiness to settle for it is good
generally to obey the law and to gloss over the absence of
substantive reasons for compliance in the example at hand
may be attributable to his having precisely the same tendency which I call a law-abiding disposition.37
We are now in a position to provide a sufficiently qualified response to the question of whether laws subjects treat
its requirements as strongly content-independent reasons.
If I am right to argue that law-following values help leading
subjects to adopt a certain attitude towards law such that
they regard the fact that law requires X as a reason for
them to perform X and that the influence of that attitude
on their actions does not depend on whether law-following
values apply to the particular situation they face, then the
answer is this: law typically operates in the practical reasoning of its subjects as a strongly content-independent
37 See Edmundson, supra note 3, p. 27. One can think of a more sophisticated
version of the above objection which appeals to considerations of fair play (standing alone or combined with consequentialist considerations). For an explanation of
why such an argument would fail in the circumstances discussed, see ibidem, pp.
28-31.
194
reason for action, though subjects adopt this attitude towards law (i.e. the attitude which implies that they treat it
as a strongly content-independent reason) due to reasons
which are not content-independent in the strong sense.38
Thus, the notion of strong content-independence may be
adequately used to describe the way things look from the
internal point of view insofar as reasons for action are concerned, but its explanatory force does not extend to reasons
for attitudes.
The following observations emerge from the analysis thus
far: (1) content-independent reasons can be understood in
either a strong sense (ascribing the reason to the mere fact
that law requires an action) or a weak sense (ascribing the
reason to substantive factors other than the nature and
merits of the action law requires). (2) While weakly content-independent reasons are an unproblematic notion,
strong content-independence is a more questionable concept whose plausibility turns, in part, on a further distinction: a distinction between evaluative statements about reasons (evaluating what reasons people have from a
normative perspective not limited to the internal assumptions of their practice) and descriptive statements about
reasons (describing the reasons participants in a certain
38 I do not wish to take sides here in the philosophical debate regarding
state-given and object-given reasons (to use Derek Parfits terminology): namely, I
will take no stance on whether the reasons I referred to in the main text above
should be classified as (state-given) reasons to have the relevant attitude or (object-given) reasons to try to bring it about that we have it or to cause ourselves to
have it. The wording I will be using mostly reason to adopt a disposition/attitude is, I think, reconcilable with either of these alternatives, and even when I
will use a less neutral formulation (such as reasons for attitudes), I will do so
merely for ease of reference. As for the above debate, see e.g. Parfit, Derek, Reasons and Rationality, in Egonsson, Dan et al. (eds.), Exploring Practical Philosophy,
Burlington, VT, Ashgate, 2001, pp. 17-39; Dunn, Robert, Akratic Attitudes and
Rationality, Australasian Journal of Philosophy, vol. 70, 1992, pp. 24-39; Olson,
Jonas, Buck-Passing and the Wrong Kind of Reasons, Philosophical Quarterly,
vol. 54, 2004, pp. 295-300; Hieronymi, Pamela, The Wrong Kind of Reason, Journal of Philosophy, vol. 102, 2005, pp. 437-57; Piller, Christian, Content-Related
and Attitude-Related Reasons for Preferences, Philosophy, vol. 81, 2006, pp.
155-81; Morauta, James, In Defence of State-Based Reasons to Intend, Pacific
Philosophical Quarterly, vol. 91, 2010, pp. 208-28.
195
NOAM GUR
normative practice characteristically take themselves to

have). (3) Strong content-independence is clearly untenable
when it appears in evaluative reason-statements. When figuring in descriptive reason-statements, however, it ought to
be assessed in light of a further distinction: a distinction
between reasons for action and reasons for attitudes. (4)
While it appears descriptively correct to say that law typically operates as a strongly content-independent reason for
action in the practical reasoning of its subjects, it ought to
be acknowledged that the attitude of treating law that way
is acquired by subjects on account of reasons which are not
strongly content-independent reasons.
IV. A DISCREPANCY WITH GOOD REASON
One particularly noticeable aspect of the foregoing analysis is the finding of a discrepancy between the type of content-independence evaluatively warranted and the type of
content-independence that corresponds with the way subjects actually reason about their actions. On the one hand,
when we evaluate what reasons subjects have in normative
terms unbounded by the internal assumptions of the legal
practice, the idea of strong content-independence appears
to lack even minimal plausibility; from this viewpoint, the
only acceptable type of content-independence is the weak
one. On the other hand, when we focus attention on the
practical assumptions of participants in the legal practice,
it appears that they typically treat laws requirements as
strongly content-independent reasons for action; they regard the fact that law requires an action as a reason for its
performance independently of whether substantive law-following values apply to their situation.
How are we to explain this discrepancy? Should we conclude that the characteristic approach of laws subjects is
the product of a sheer error on their part? Or is there a less
radical explanation of the relation between the reasons that
truly apply to subjects and the way they actually reason,
196
i.e. one that makes sense of the latter in terms of the former? It appears to me, prima facie, that such an explanation is likely to exist. For it is highly improbable that the socially prevalent perception of laws normative force, as
developed over millennia and across many cultures, is the
result of an accidental human error. Even if that perception
is in some sense mistaken, there are likely to be good reasons why it developed the way it did. It is plausible that the
shape and content it came to have are such that fulfill
some valuable functions in social life. That said, it should
be clarified that these intuitive assumptions are not stated
here as a constitutive element of my argument or a premise
on which I seek to rely. Their only role is to serve as a background to what will be a self-contained argument about the
good reasons or valuable functions that may justify
treating legal requirements as strongly content-independent
reasons. I now turn to this argument.
I will initially put forth a basic hypothesis about the justification for treating law as a provider of strongly content-independent reasons. This hypothesis will be cast in a
rather general and abstract form, but will then be substantiated and filled out with more concrete content. The hypothesis is this: the subjects of a reasonably just legal system should treat its requirements as providing them with
strongly content-independent reasons for action (rather
than reasons that depend on content or on whether law-following values apply to their situation) because by so doing
they would be more likely to conform to substantive reasons that apply to them in the areas that law regulates.
Now, this justificatory hypothesis may strike a familiar
chord with readers of contemporary jurisprudence; it resonates, of course, with Razs normal justification thesis, according to which the 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
197
NOAM GUR
them, rather than by trying to follow the reasons which apply to him directly.39 Notwithstanding their close affinity,
however, there are a number of pertinent respects in which
my justificatory hypothesis differs from Razs normal justification thesis: (1) the two do not seek to justify the same
thing. According to Raz, insofar as an alleged authority satisfies the condition stated in the normal justification thesis,
it gains the status of a legitimate authority, which means
that (to this extent) its directives become pre-emptive reasons, reasons for action that exclude some otherwise applicable reasons. This is significantly more than what my hypothesis seeks to establish: firstly, I do not suggest that
certain directives (or the fact that they were issued) are reasons, but only that it is desirable that, in our daily operation as subjects of the law, we treat them as reasons;40 secondly, my argument does not refer to pre-emptive reasons,
but rather to content-independent reasons. These are two
distinct concepts, for a reason can be independent of content without excluding any other reasons.41
(2) While the normal justification thesis states what Raz
conceives of as the (main) condition under which a government would acquire legitimate authority, it does not contend that governments, or even reasonably just governments, generally do satisfy that condition. In fact, in several
places Raz denies that they generally do as, for example,
when he says that political authorities are likely to have
more limited authority [in the legitimate sense] than the authority many, perhaps all of them, claim to have, and that
people generally believe they have.42 In contrast, my hyRaz, supra note 1, p. 53 (emphasis omitted). See also Green, supra note 1, p. 56.
See Alexander, supra note 35.
41 As noted, I have discussed Razs pre-emption thesis and exclusionary reasons conception elsewhere (Gur, supra note 14).
42 Raz, Joseph, The Problem of Authority, Minnesota Law Review, vol. 90,
2006, pp. 1003-44, at p. 1008. See also Raz, supra note 1, pp. 70-80, 99-104; Raz,
Joseph, Ethics in the Public Domain, rev'd paperback ed., Oxford, Clarendon Press,
1995, pp. 341, 347-50. Following his above-quoted comment Raz says: This still
requires explaining why people are so mistaken , and he adds in a footnote that
this is a point orally made to him by H. L. A. Hart (Raz, The Problem of Authority,
39
40
198
pothesis does not merely state the condition under which it

would be justified to regard law as providing strongly content-independent reasons, but rather also contends that,
given a reasonably just legal system, the condition is generally fulfilled.
(3) Closely related to the preceding point, while Raz provides a general and abstract statement of the condition for
acquiring legitimate authority, i.e. his normal justification
thesis, he does not extensively discuss the concrete factors
that enable authorities to meet that condition (insofar as
they do). When he pronounces on the matter, however, he
stresses two main factors:43 first, comparatively superior
knowledge and expertise that an authority may have regarding the substance it regulates;44 second, the authoritys
capacity to secure and facilitate social coordination between
its subjects.45 Now, bearing out my own hypothesis will require a close look at the underlying substantive factors that
may justify treating legal requirements as strongly content-independent reasons. This will yield a picture rather
different from the one Raz appears to envisage: it will be observed that while differential expertise and coordinative caibidem, pp. 1008-9 and fn. 11). I agree with this last remark. As earlier indicated, it
seems to me that although the socially prevalent attitude towards laws
normativity may be in some sense mistaken, it is not likely to have evolved through
arbitrary processes leading to senseless implications for our life in society. My
analysis in this Section is an attempt to explain the sense and value of the prevalent normative attitude.
43 He notes: The two basic arguments for authority depend on its ability,
through concentrating expertise on various issues, to overcome common ignorance and on its ability to help solve common difficulties in securing coordination
(Raz, Joseph, Practical Reason and Norms, 2nd. ed., Princeton, Princeton University Press, 1990, p. 195). Another factor that seems to play a significant role in
Razs conception of legitimate authorities is their ability to provide a framework
within which social life can proceed in the face of fundamental disagreements and
differences between members of society (Raz, supra note 1, p. 58). The arguments I
will make as to comparative expertise and coordinative capacities of authorities
also apply, mutatis mutandis, to their ability to bridge such ideological gaps.
44 Raz, supra note 43, pp. 63-4, 195; Raz, supra note 11, pp. 14-5.
45 Raz, supra note 43, pp. 64, 195; Raz, Joseph, Facing Up: A Reply, Southern
California Law Review, vol. 62, 1989, pp. 1153-253, at pp. 1191-3; Raz, supra note
11, pp. 14-5. See further on the coordinative role of authorities in Raz, supra note
1, p. 49.
199
NOAM GUR
pacities are relevant factors, the root of the justification lies

elsewhere, in factors that Raz touches upon, but whose implications he never acknowledges46 indeed, when closely
examined, these factors seem to suggest that the ordinary
scope of legitimate governmental authority is significantly
more general and less piecemeal than Raz believes it to be.
I now turn to verify my hypothesis. Consider, first, the
justificatory force of factors of the type mentioned in the
preceding paragraph: i.e. epistemic considerations for compliance with law such as the recognition that ones knowledge and expertise as to a legally regulated matter are inferior in comparison to the lawmakers, or non-epistemic
considerations such as the recognition that an act of compliance would contribute to social coordination. Do these
law-following values, in themselves, explain why treating legal requirements as strongly content-independent reasons
(i.e. reasons independent of law-following values) is requisite for making optimal use of law as a guide to right action? The answer seems to me to be negative. Nothing in
the mere facts that a legal authority enjoys relative expertise and possesses coordinative abilities prevents these
facts from being adequately considered and given due
weight in the practical reasoning of laws subjects. Thus,
for example, if a regulator has greater expertise than mine
on a certain matter, I can factor his comparative expertise
into my respective decisions by treating the case for doing
what he prescribes as weightier than it would have seemed
to me otherwise; and the degree of extra weight I thus accord to the prescribed conduct can, and seemingly should,
be sensitive to the degree of expertise the regulator has on
the relevant matter, in much the same way that the weight
I accord to an investment advice would be sensitive to
whether it was given to me by the consultant at my local
bank branch or by, say, Warren Buffett. This holds true
also in respect of non-epistemic factors such as consider46 See his reference to biases in Raz, supra note 1, p. 75. The deficiencies of his
position will be explained in note 48.
200
ations of social coordination: for, firstly, the relative importance of coordination may vary with the context and type of
activity concerned; secondly, not all laws are equally conducive to goals of coordination; and, thirdly, even laws that
are in general conducive to coordination goals may, on occasion, fail to serve and even contravene these very goals.
Surely, the correctness of practical decisions partly depends on these variables. When we focus attention on differential expertise or social coordination, therefore, it is
hard to see why laws subjects should not treat its normative force as conditional upon these factors, i.e. as content-independent in the weak sense.
There are, however, other factors that emphatically militate against the mode of practical reasoning associated with
weak content-independence and recommend treating law as
a provider of strongly content-independent reasons. The
most significant of these factors, I will argue, is certain human biases and fallibilities to which we are commonly susceptible in our typical decision-making environment as
laws subjects.47 More particularly, I will advance the following twofold argument: (1) due to certain stimuli present in
the decision-making environment in which laws subjects
typically operate, an assessment of law-following values
which they carry out in this environment is likely to be influenced by certain biases pulling towards non-compliance;48 (2) the influence of those biases can be effectively
47 Another practical difficulty with the mode of reasoning associated with weak
content-independence is that carrying out a complete assessment of reasons for
compliance in each and every situation would require an exceedingly large amount
of time and effort on the assessors part, making this method of reasoning an extremely inefficient one.
48 Raz notes that one of the factors capable of establishing the legitimacy of authority is its having a steadier will less likely to be tainted by bias, weakness or impetuosity, less likely to be diverted from right reason by temptations and pressures (Raz, supra note 1, p. 75). Yet his account fails to accommodate the
implications of biases. As will become clear in the following paragraphs, the relevance of biases is pervasive: first, the biases pertinent to our question are common
human biases to which most of us, not only the unwise or reckless, are susceptible
when operating in our typical decision-making environment as laws subjects; second, laws aptness as measure against those biases is primarily a function of basic
201
NOAM GUR
counteracted by adopting an attitude which involves treating law itself as a reason for action independently of ones
assessment of law-following values as applicable to the situation at hand:49 precisely because this perceived reason
operates independently of ones assessment of law-following
values i.e. operates as a strongly content-independent
reason it remains free from the biases which affect that
assessment and can effectively counterbalance those biases.50
What biases, then, are prone to influence an assessment
of law-following values made by subjects in their typical decision-making environment? Since I have comprehensively
discussed these biases elsewhere, herein I will confine myself to a relatively brief account focusing on three of them.51
Consider, first, the tendency known to psychologists as
self-enhancement bias or the better-than-average effect:52
structural features commonly present in legal systems, e.g. the generality and prospective character of legal enactments. All this suggests that, under reasonably
just and competent lawmaking institutions, as those found in familiar legal systems, the condition stated in the normal justification thesis is fulfilled widely and
generally, rather than in the narrow and piecemeal manner Raz envisages.
49
I do not suggest that law is a suitable measure against the relevant biases because lawmaking officials possess personal qualities that make them bias-immune
individuals (which is, of course, not the case). Rather, law can fulfill this role primarily because the settings and mode of decision-making in which lawmakers typically operate are significantly less susceptible to those biases than the settings and
mode of decision-making in which subjects typically operate.
50 Cannot the relevant biases be effectively counteracted by the legal systems
generating prudential reasons for compliance in the form of punishment? While
generating prudential reasons for compliance is an essential part of the solution,
the impracticality and unattractiveness of a system of governance that would depend for its efficacy solely or mainly on the use of coercive force and on the fear of
punishment explain the need for another measure against common biases.
51 Manuscript on file with author (forthcoming, Oxford University Press, 2013).
52 See, e.g., Baumhart, Raymond, An Honest Profit: What Businessmen Say
About Ethics in Business, New York, Holt, Rinehart and Winston, 1968, pp. 20-5;
Larwood, Laurie and Whittaker, William, Managerial Myopia: Self-Serving Biases
in Organizational Planning, Journal of Applied Psychology, vol. 62, 1977, pp.
194-8; Svenson, Ola, Are We All Less Risky and More Skillful Than Our Fellow
Drivers, Acta Psychologica, vol. 47, 1981, pp. 143-8; Brown, Jonathon D., Evaluations of Self and Others: Self-Enhancement Biases in Social Judgments, Social
Cognition, vol. 4, 1986; Kruger, Justin and Dunning, David, Unskilled and Unaware of it: How Difficulties in Recognizing One's Own Incompetence Lead to In-
202
when people evaluate their own performance and skills in

comparison with others, most of them rate themselves as
better than the average person, with a disproportionately
large percentage placing themselves towards the top end of
the comparative scale.53 Thus, for example, experimental
evidence suggests that the majority of drivers take themselves to be more skilful and less risky than the average
driver (which obviously signifies an inflation in self-appraisal).54 The relevance of this bias becomes clear once it is
brought to mind that, for an actor to correctly assess the
weight of reasons to follow the guidance of an authority,
rather than his own view on the merits, he must estimate
the relative reliability or unreliability of his own judgment,
knowledge, and skill as to the subject of regulation. To use
again driving as an example, a driver who acts on his own
assessment of the reasons for and against following the legal speed limit, overtaking restrictions, or other traffic
rules, should take into account the degree to which his
knowledge about traffic matters may be inferior relative to
the road authoritys, as well as the full extent of his shortcomings and actual limits of his ability as a driver. Self-enhancement bias, however, means that he is liable to
underestimate these factors.
Consider, next, the tendency frequently referred to in
psychological parlance as self-serving bias: when people
make judgments about practical matters that bear on their
self-interest, more often than not, they evaluate data and
perceive moral principles in a manner beneficial to them-
flated Self-Assessment, Journal of Personality and Social Psychology, vol. 77,

1999, pp. 1121-34. See also: Cross, K. Patricia, Not Can, But Will College Teaching Be Improved?, New Directions for Higher Education, vol. 1977, pp. 1-15.
53 See references in note 52. Note that the observation is not that people always
suffer from this bias. Rather, what has been observed is a general tendency; a pattern that characterizes most peoples perception of their performance and skill. A
similar caveat applies to the other biases discussed here.
54 Svenson, idem. See also Preston, Caroline E. and Harris, Stanley, Psychology of Drivers in Traffic Accidents, Journal of Applied Psychology, vol. 49, 1965,
pp. 284-8.
203
NOAM GUR
selves.55 Thus, even if the decision ensues from a process of

reasoning that is not wittingly egoistic, excessive weight is
nonetheless likely to be ascribed to those considerations
that coincide with the decision-makers needs and wants at
the expense of other relevant considerations. That such
tendency exists in people was long ago noticed by thinkers
such as John Locke,56 David Hume,57 and Albert Venn
Dicey,58 and has been borne out by a large body of empirical
studies in more recent times.59 Now, the pertinence of this
bias in the context of our inquiry transpires once it is recalled that law makes social life and human cooperation
possible by imposing on individuals restrictions and burdens which are often at variance with their nearest self-in55 Messick, David M. and Sentis, Keith P., Fairness and Preference, Journal of
Experimental Social Psychology, vol. 15, 1979, pp. 418-34; Messick, David M., Social Interdependence and Decision Making, in Wright, George (ed.), Behavioral Decision Making, New York, Plenum, 1985, pp. 94-100; Thompson, Leigh and
Loewenstein, George, Egocentric Interpretations of Fairness and Interpersonal
Conflict, Organizational Behavior and Human Decision Processes, vol. 51, 1992, pp.
176-97; Babcock, Linda and Loewenstein, George, Explaining Bargaining Impasse: The Role of Self-Serving Biases, The Journal of Economic Perspectives, vol.
11, 1997, pp. 109-26; Dawson, Erica, Gilovich, Thomas and Regan, Dennis T.,
Motivated Reasoning and Performance on the Wason Selection Task, Personality
and Psychology Bulletin, vol. 28, 2002, pp. 1379-87. See further: Pogarsky, Greg
and Babcock, Linda, Damage Caps, Motivated Anchoring, and Bargaining Impasse, Journal of Legal Studies, vol. 30, 2001, pp. 143-59.
56 Locke, John, Two Treatises of Government, 2nd. ed., Cambridge, Cambridge
University Press, 1967, bk. II, ch. IX, 324, noting that, in the state of nature,
though the law of Nature be plain and intelligible to all rational Creatures; yet Men
being biassed by their Interest are not apt to allow of it as a Law binding to them
in the application of it to their particular Cases.
57 Hume, David, A Treatise of Human Nature, Oxford, Clarendon Press, 1888,
bk. III, pt. II, VII: As violent passion hinders men from seeing distinctly the interest they have in an equitable behaviour towards others; so it hinders them from
seeing that equity itself, and gives them a remarkable partiality in their own favour.
58 Dicey, Albert V., Lectures on the Relation Between Law and Public Opinion in
England During the Nineteenth Century, 2nd. ed., London, Macmillan, 1914, p. 15:
[M]en come easily to believe that arrangements agreeable to themselves are beneficial to others. A mans interest gives a bias to his judgment far oftener than it corrupts his heart. He overestimates and keeps constantly before his mind the
strength of the arguments in favour of, and underestimates, or never considers at
all, the force of the arguments against .
59 See references in note 55.
204
terest and immediate preference as, for instance, when

the law prohibits manufacturers from emitting harmful
substances or levies a tax on peoples income. When subjects are faced with such restrictions or burdens, therefore,
their assessment of the reasons for and against compliance
is likely to be tainted by a self-serving bias, leading them to
perceive the facts and balance reasons in a manner that is
somewhat overly sensitive to their individual circumstances
and needs. If they were to act solely on such an assessment, they would thus tend to exempt themselves too often
from performing their part in arrangements requisite for
social life and welfare.
Another bias with special pertinence to our discussion
manifests itself when people make decisions involving
intertemporal choice. As David Hume once noted,60 when
we express a preference or form an intention as to a future
course of action e.g. to go jogging tomorrow morning, go
on a diet, or break a bad habit61 temptations pulling in
the opposite direction may become harder to resist as the
time for the intended action approaches, and we sometimes
find ourselves departing from what we earlier held to be,
and what may indeed be, optimal action. Modern experimental work has confirmed the existence of this effect and
explained it as part of a common bias that behavioral scientists call myopic or hyperbolic discounting: in simple words,
a tendency to overvalue immediate gains or imminent gratifications at the expense of long-term rewards.62 The bearing
60
Hume, supra note 57, p. 536.
61
The examples are mine.

See, e.g., Ainslie, George, Specious Reward: A Behavioral Theory of Impulsiveness and Impulse Control, Psychological Bulletin, vol. 82, 1975, pp. 463-96;
Solnick, Jay H. et al., An Experimental Analysis of Impulsivity and Impulse Control in Humans, Learning and Motivation, vol. 11, 1980, pp. 61-77; Ainslie, George
and Haendel, Varda, The Motives of the Will, in Gottheil, Edward et al. (eds.),
Etiologic Aspects of Alcohol and Drug Abuse, Springfield, IL, Charles C Thomas,
1983, pp. 119-40; Millar, Andrew and Navarick, Douglas J., Self-Control and
Choice in Humans: Effects of Video Game Playing as a Positive Reinforcer, Learning and Motivation, vol. 15, 1984, pp. 203-18; Herrnstein, Richard J., Rational
Choice Theory: Necessary but Not Sufficient, American Psychologist, vol. 45, 1990,
62
205
NOAM GUR
of this bias on our question hardly needs to be explained:

legal rules are typically designed and applied in a prospective manner with the aim of forming social arrangements
sustainable through time and in the face of impulsive tugs
in human decision-making. This aim, however, is not likely
to be duly reflected in an assessment of law-following values made by a subject in his daily setting of operation under the law. For, when confronted with the personal inconvenience involved in following a rule and with concrete and
immediate stimuli pulling in the opposite direction, he is
prone to be led by a myopic bias to give too little weight to
the long-term advantages of following the rule.
We are now in a position to consider how situational biases of the type described above bear on the question of
content-independence. It is primarily such biases, and their
frequent presence in subjects decision-making environment, that make it desirable for subjects to adopt the attitude of treating legal requirements as strongly, rather than
weakly, content-independent reasons for action.63 A subject
(call him subject 1) who operates in accordance with the
pp. 356-67; Winston, Gordon C. and Woodbury, Richard G., Myopic Discounting:
Empirical Evidence, in Kaish, Stanley and Gilad, Benjamin (eds.), Handbook of
Behavioral Economics, Greenwich, CT, JAI Press, 1991, vol. 2B, p. 325; Kirby, Kris
N. and Herrnstein, Richard J., Preference Reversals Due to Myopic Discounting of
Delayed Reward, Psychological Science, vol. 6, 1995, pp. 83-9; Kirby, Kris N., Bidding on the Future: Evidence Against Normative Discounting of Delayed Rewards,
Journal of Experimental Psychology: General, vol. 126, 1997, pp. 54-70.
63 It might be a limitation of my analysis in this Section that it focuses on the
decision-making environment of private actors, rather than judges and other
law-applying officials. Even if so, it is important not to overstate the extent of this
limitation. Although some of the situational biases discussed above do not
paradigmatically feature in the decision-making environment of law-applying officials, other situational biases and fallibilities do. See, e.g., Schauer, Frederick,
Rules and the Rule of Law, Harvard Journal of Law & Public Policy, vol. 14, 1991,
pp. 645-94, at pp. 679-81, Schauer, Frederick, Playing by the Rules, Oxford, Clarendon Press, 1991, pp. 149-55, 158-62, 229-33, where Schauer refers to the fallibility that characterizes some official decision-making environments in his explanation of the benefits of rule-based decision-making. He grounds the principal
justification for reliance on rules in decisional modesty and the need to allocate
decisional power between different decision-making environments according to
their relative reliability or fallibility. My arguments are consistent with his position.
206
weak notion of content-independence regards himself as

having a reason to comply with law only if and insofar as he
recognizes that an act of compliance is justified by substantive considerations applicable to his situation. As we have
seen, however, when situated in his daily environment of
activity under the law, his perception of such considerations is likely to be tainted by biases towards non-compliance. By contrast, a subject (call him subject 2) whose mode
of practical reasoning accords with the strong notion of content-independence, regards law as providing him with reasons for action independently of his situational assessments of substantive considerations for compliance.
Precisely because law-given reasons enjoy this type of independence in his practical reasoning, they can exert their influence in the face of situational biases and effectively
counteract them. Hence, subject 2 is, ceteris paribus, less
likely to commit practical errors than subject 1. This, then,
is a good reason, grounded in substantive considerations,
to adopt the normative attitude of subject 2, rather than
subject 1 it is a weakly content-independent reason to
adopt the attitude of treating legal requirements as strongly
content-independent reasons.64
64 I said that the reason is a weakly content-independent reason, but it is perhaps more accurate to say that its content-independence, in addition to being
weak, is no more than relative. For my argument proceeded, and depends for its
plausibility, on an assumption that the relevant lawmaking institutions are reasonably just and competent, which, in turn, suggests that their enactments are
likely to have (even if not always necessarily have) acceptable (even if not optimal)
content.
I take myself to have shown in my earlier arguments (see particularly text accompanying nn. 31-38) that the above statement is internally coherent: i.e. that
the former reasons being weakly content-independent does not turn the latter
reason into a weakly content-independent reason as well. There are crucial differences between an actor who treats law as a provider of weakly content-independent reasons simpliciter and an actor who (for weakly content-independent reasons)
adopts the attitude of treating law as a strongly content-independent reason for action. Two notable manifestations of the difference, which emerge from the foregoing discussion, are, first, in the modes of deliberation respectively used by these actors, and, second, the lack of extensional equivalence in the outcomes produced by
their respective attitudes. First, at the deliberative level, since the first actor does
not regard law itself as a reason for action, he approaches legal requirements by
examining whether, and to what extent, compliance in the particular case at hand
207
NOAM GUR
V. CONCLUSION
Typical reactions to the idea of content-independent reasons range from unhesitant approval to outright dismissal;
some think it illuminatingly captures a central property of
laws normativity and authoritative nature, whereas others
find it impossible to understand how mere artifacts of the
human will can turn something into a reason independently of its substance and value. In this paper, I have suggested that this disparity of reactions results from insufficient attention to the relevance of three distinctions. The
first distinction is between two different senses of content-independence, which I referred to as weak and strong
content-independence. Weakly content-independent reasons are reasons that do not depend on the nature and
merit of the action that law requires. Strongly content-independent reasons are reasons that do not depend on the nature and merit of the action law requires or on any other
substantive law-following values (e.g. values associated
with social coordination, order and stability, or fair play
considerations). The latter rendering implies that the mere
fact that law requires an action must itself count in the
practical reasoning of its subjects.
In light of this distinction, it became transparent that
what initially appeared to be a general objection against
content-independence is in fact an objection against only
one sense of content-independence, i.e. the strong sense.
The weak sense, on the other hand, does not seem to impliwould serve any desirable substantive goals (e.g. this requirement is meant to facilitate coordination; but there is no one around; so by complying I would be coordinating with no one ). In contrast, the second actor treats legal requirements
themselves as reasons for action, his inclination to comply with law is not contingent on situational assessments of the applicability of law-following values (though
may occasionally be overridden by sufficiently compelling and clear reasons to the
contrary). Second, the two attitudes are not extensionally equivalent in terms of
the outcomes they produce; there are cases in which they diverge. Take again the
deserted traffic light example discussed above; in this situation a correct assessment of the balance of reasons for action (including law-following values) as applicable to the case at hand recommends running the red light, but an actor who
treats legal requirements themselves as reasons may well stop.
208
cate any contentious or problematic claims. I therefore

turned to consider the strong sense of content-independent
reasons. At this point, a second distinction became relevant: namely, a distinction between evaluative statements
about reasons (evaluating what reasons people have from a
normative perspective unbounded by the internal assumptions of the legal practice) and descriptive statements about
reasons (describing the reasons that characteristically figure in the practical deliberation of participants in the legal
practice). From the former, evaluative perspective, strong
content-independence, with its suggestion that reasons can
come into existence through the mere say-so of a human
agent, appeared no more plausible than a belief in the
power of spells. From this perspective, the only type of content-independence that can sensibly be spoken of is the
weak type. However, when we descriptively examined the
way participants in the legal practice typically treat its requirements, strong content-independence appeared to have
some explanatory force after all, though it had to be qualified by a third distinction: namely, a distinction between
reasons for action and reasons for adopting certain attitudes. Thus, I have argued that although law-following values are part of what brings subjects to adopt and maintain
the general attitude they normally have towards law i.e.
their law-abiding disposition, the attitude of treating legal
requirements as reasons for action this attitude makes its
force felt in their action-specific deliberation independently
of whether law-following values apply to the particular situation they face. In other words, law seems to typically operate in the practical reasoning of its subjects as a strongly
content-independent reason for action, though subjects
adopt and maintain the attitude of treating law that way
due to reasons that are not content-independent in the
strong sense.
Finally, I have argued that despite the observed discrepancy between the evaluatively sound type of content-independence and the descriptively correct type of content-inde209
NOAM GUR
pendence, the two are nonetheless interconnected by a

deeper rationale. That is, even though strongly content-independent reasons do not exist in the evaluative sense,
there is an evaluative consideration which makes it desirable that, in our practical operation as subjects of the law,
we proceed on an assumption or operate with an attitude
which involves treating legal requirements as strongly content-independent reasons. Put in its most general form, the
consideration is that by thus treating the requirements of a
(reasonably just) legal system, we would be more likely to
conform to substantive reasons that apply to us in legally
regulated spheres of activity. The fundamental reason for
this, I have argued, lies not so much with expertise or specialized knowledge that our lawmakers might possess, as
with certain situational biases and fallibilities to which we
are commonly susceptible in our settings of operation under the law. I have demonstrated that due to certain stimuli
characteristically present in the decision-making environment of laws subjects, a bare assessment of law-following
values carried out in that environment is likely to be
tainted by biases leading one to discount or underestimate
the weight of applicable law-following values. This observation made it clear why it is desirable that, in our usual operation as subjects of the law, we treat its directives as
strongly content-independent reasons: only if we treat legal
directives in this way can their normative force feed into
our decisions without first being mediated by our situational judgment and watered down by the biases implicated
in that judgment. Only this way, therefore, can law
adequately counteract situational biases and fallibilities,
and optimally fulfill its function as a normative guide to
right action.
210
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
TOWARDS A COMPREHENSIVE POSITIVIST

THEORY OF LEGAL OBLIGATION*
Kenneth Einar HIMMA
Resumen:
El concepto de obligacin jurdica es completamente central para la
prctica jurdica. Sin embargo, el positivismo carece de una explicacin
integral de la obligacin jurdica, puesto que se concentra solamente en
el reconocimiento de obligaciones de segundo orden que recaen sobre los
oficiales, y no explican las obligaciones jurdicas de primer orden que recaen sobre los ciudadanos. Cuando las obligaciones jurdicas se relacionan conceptualmente con normas vlidas jurdicamente, el error de no
tener una explicacin integral de la obligacin jurdica pone en crisis la
teora positivista de la validez jurdica. En este ensayo se desarrolla la
explicacin hartiana de la obligacin social y se complementa su explicacin de las obligaciones jurdicas de segundo orden del oficial en su calidad de oficial con una explicacin de la obligacin de primer orden de los
ciudadanos. Esta ltima se constituye, argumenta el autor, por la presin social en la forma de la autorizacin que tiene el Estado de echar a
andar la maquinaria coercitiva si no se cumple con la obligacin.
* Earlier versions of this paper were presented at the University of Bristol

School of Law; the Ono School of Law Conference on Positivism, Democracy and
Constitutional Interpretation; the Seminar Problema of the Legal Research Institute of UNAM (National Autonomous University of Mexico); and the McMaster University Conference on the Nature of Law: Contemporary Perspectives. I am indebted to Julian Rivers, Patrick Capps, and Marc McGee for comments and
suggestions. I am also grateful to the following for their very helpful comments on
earlier drafts of this paper: Larry Alexander, Brian Bix, Patrick Capps, Jules Coleman, Jorge Fabra, Imer Flores, Mark Greenberg, Nina Guzman, Matthew Kramer,
Marc McGee, Julian Rivers, Scott Shapiro, Seana Shiffrin, Brian Tamanaha, and
Wil Waluchow.
211
KENNETH EINAR HIMMA
Palabras clave:
Obligacin jurdica, validez jurdica, coercin, positivismo jurdico, teora del derecho.
Abstract:
The concept of legal obligation is utterly central to legal practice. But positivism lacks a comprehensive account of legal obligation, focusing only on
the second-order recognition obligations of officials with no account of the
first-order legal obligations of citizen. As legal obligations are conceptually
related to legally valid norms, this failure calls into question positivisms
theory of legal validity. In this essay, I develop Harts account of social obligation and supplement his account of the second-order legal obligations of
official qua official with an account of the first-order obligations of citizens.
The latter is constituted, I argue, by social pressure in the form of the authorization of the states coercive machinery for non-compliance.
Keywords:
Legal Obligation, Legal Validity, Coercion, Legal Positivism, Legal Theory.
212
TOWARDS A COMPREHENSIVE POSITIVIST THEORY
SUMMARY: Introduction. I. The Centrality of Obligation Talk

to Legal Practice. II. The Concept of Obligation.
III. Harts Theory of Social Obligation. IV. A Comprehensive Theory of Legal Obligation. V. Supporting Considerations.
INTRODUCTION
Perhaps no concept is more central to legal practice than
that of legal obligation. Statutes, case law, and legal arguments are characteristically framed in terms of what some
person or class of persons is obligated to do. Such practices presuppose that legal norms at least those making
certain actions mandatory regulate behavior by creating
legal obligations. Law characteristically regulates behavior
by creating obligations.
Both officials and citizens are subjects of legal obligations. Citizens are obligated to honor their contracts and to
refrain from violence under most circumstances; these are
first-order obligations defined by primary norms. Judges
are obligated to decide cases under the relevant norms;
these are second-order obligations (usually) created by recognition norms.
Hart appears to have at least the beginnings of a comprehensive theory of legal obligation. As is well known, Hart
believes that legal obligation is a form of social obligation
and that social obligations arise when accepted norms are
thought sufficiently important to back with social pressure
to conform. The second-order legal obligations of officials
are explained by their taking the internal point of view towards the rule of recognition. Although he rejected Austins
sanction theory of obligation as not accurately expressing
either the sense in which civil law binds or the sense in
which officials are bound, he seemed to intimate that
first-order legal obligations of citizens are explained by the
availability of institutional coercive mechanisms for enforcing first-order legal norms against citizens. As Hart puts
213
KENNETH EINAR HIMMA
the point, the typical form of legal pressure may very well
be said to consist in such threats [of physical punishment
or unpleasant consequences] (CL 179, 180).
In this essay, I wish to develop what I take to be Harts
account of social obligation and supplement his account of
the second-order legal obligations of official in their capacities as official with an account of the first-order obligations
of citizens. The latter is constituted, I argue, by social pressure in the form of the authorization of the states coercive
machinery for non-compliance.
At the outset, it is important to understand that there is
a difference between the authorization of coercive enforcement mechanisms and the application of such mechanism
in a case of non-compliance. These are two distinct notions.
The idea that such mechanism are authorized for non-compliance simply means that officials have authority to use
these mechanisms as legally justified responses to noncompliance. The idea that such mechanism are applied
simply means that those coercive mechanisms have been
used against someone on the ground that he failed to comply. But it is important to note that this does not entail
even that the use of such mechanisms are legally justified
as one would expect if legal mistakes are possible. The
authorization of coercive enforcement of a legal norm provides a legal justification for the appropriate application of
the relevant mechanism for non-compliance with the norm.
One might object that the violation of a legal obligation
justifies the application of coercive mechanisms and thus
that a legal obligation cannot be constituted by coercive enforcement applications.1 This misunderstands the thesis of
the paper. The claim being defended here is that the authorization of such mechanisms for non-compliance is, in part,
what constitutes a legal norm as binding and hence legally
obligatory and thus provides the justification for application
in genuine cases of non-compliance. The obligation is constituted, in part, by the authorization of such mechanisms
1
I am indebted to Scott Shapiro for this line of argument.
214
and is not identical with the existence or application of

such mechanisms.
I. THE CENTRALITY OF OBLIGATION TALK TO LEGAL PRACTICE
The concept of obligation is everywhere in legal practice.
For example, a plaintiff in a contract dispute typically
claims the defendant is obligated to perform some act,
while the defendant argues that the defendants performance is excused by the plaintiffs own breach of obligation. Likewise, a prosecutor will argue that the defendant
breached some obligation or duty defined by the criminal
law, while the defense will argue that the defendant did not
breach such a duty or obligation. Finally, judges frequently
couch their decisions in terms of what some party is obligated to do. In, for example, Henningsen v. Bloomfield Motors, Inc., the court held that [i]n a society where the automobile is a common and necessary adjunct of daily life,
and where its use is so fraught with danger to the driver,
passengers and the public, the manufacturer is under a
special obligation in connection with the construction, promotion, and sale of his cars.2
As these obligations arise under law, they are thought to
be legal in source and character. This, of course, is not to
suggest that moral obligation is irrelevant to ordinary talk
about legal obligation; it is simply to assert ordinary legal
talk and practice presupposes the existence of legal obligations analytically distinct from moral obligations. Although
the content of law and the content of morality frequently
converge, they frequently diverge as well; in such cases,
however, the law defines a legal obligation if not a moral
one.
The law regulates behavior by a variety of means, including power-conferring norms like those governing the creation of binding contracts and wills, but characteristically
2
161 A.2d 69 (1960), at 85 (emphasis added).
215
KENNETH EINAR HIMMA
constrains the behavior of citizens by creating such obligations. The law does not generally traffic in weaker oughts
that encourage behavior without making it mandatory in
some sense. Legislative enactments that do not create obligations are not actionable and cannot support a claim for
damages or punitive measures.
This is the view that Hart takes. Hart observes, for example, that Austin correctly assumes that systems of law necessarily create some legal obligations:
[T]he theory of law as coercive orders, notwithstanding its errors, started from the perfectly correct appreciation of the
fact that where there is law, there human conduct is made
in some sense non-optional or obligatory. In choosing this
starting point the theory was well inspired, and in building
up a new account of law in terms of the interplay of primary
and secondary rules we too shall start from the same idea.3
Further, Hart asserts that it is a conceptual truth that

primary legal norms generally define legal obligations (some
confer legal liberties): Rules of the first type impose duties
[i.e., primary rules]; rules of the second type [i.e., secondary
rules] confer powers, public or private (CL 80-81). If Hart is
correct, then law regulates the behavior of citizens by creating obligations that are legal in source and character.
Law is a normative institution and its normativity is conceptually linked to its capacity to generate obligations. This
suggests an adequacy constraint on conceptual theories of
law. While conceptual theories of law are most conspicuously concerned with giving an analysis of the concept of
law, they must also be concerned to provide an account of
all normative concepts figuring prominently in legal practice including that of legal obligation.
3 H. L. A. Hart, The Concept of Law, Rev. ed. (Oxford: Oxford University Press,
1994), 82; emphasis added. Hereinafter CL.
216
II. THE CONCEPT OF OBLIGATION

If ordinary talk is any indication, there are different types of
obligation. We distinguish, for example, moral, social and legal obligations and speak as if these types of obligation are
conceptually distinct. Even so, many theorists believe they
are instances of the same general type. As Joseph Raz puts
it: normative terms like a right, a duty, ought are used
in the same sense in legal, moral, and other normative
statements.4 While moral, social, and legal obligation differ
in important ways, there are certain elements essential to
the notion of obligation and these elements are present in
moral, social, and legal obligations.
This is certainly true of various kinds of norm. For example, moral and legal norms are conceptually distinct; the
content of moral norms sometimes diverges from the content of legal norms, as is presumably true of the content of
the moral and legal norms governing promise-keeping. But
moral and legal norms are both kinds of norm; as such,
they instantiate properties that are necessary and sufficient
for being norms. Although legal and moral norms have
many different properties, both satisfy the application-conditions for the concept-term norm.
One would expect, as Raz believes, that the same would
be true of the various kinds of obligation. Legal and moral
obligations presumably have different properties, but both
satisfy the application-conditions for the concept-term obligation in the following sense: satisfaction of the application-conditions for obligation will be necessary (though
not sufficient) for something to count as either a legal obligation or a moral obligation. If so, then the set of application-conditions for obligation will be a subset of the set of
4 Joseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979), 158. See
also Richard Brandt, The Concepts of Obligation and Duty, Mind p. 380 ([I]t is
dubious whether there are sharply distinct moral and non-moral senses [of obligation]. It may be that obligation and duty preserve an identical core of meaning
throughout moral and non-moral uses. This, in fact, is the view of the matter best
supported by the evidence.)
217
KENNETH EINAR HIMMA
application conditions for moral obligation and legal obligation.

If this is correct, then we cannot understand the concept
of legal obligation without understanding the general notion
of obligation.5 In what follows, I will sketch what I take to be
the central elements of the general concept of obligation.
1. Obligations and Mandatory Prescriptions
It is tempting to think that this much is clear about obligations: obligations are conceptually related to norms. While the
existence of a norm prescribing act X might not be a sufficient condition for X to be obligatory in the relevant sense, it
is a necessary condition. There simply could not be an obligation unrelated (perhaps in the strong sense of being defined
by) to a norm.
Although plausible, the idea that the existence of a norm
prescribing X is a necessary condition for someone to be obligated to X is problematic for the following reason. It cannot
be applied to morality without assuming a substantive account of morality that is controversial namely, the idea that
morality is grounded in general norms. Moral particularists
deny this assumption, believing that the morality of any particular behavior is too context-dependent to be captured by
general norms even those that state, so to speak, their own
5 Hart was concerned with analyzing the concept of obligation though he focused on social obligation, apparently believing that all obligations are social in character: (1) [I]t is crucial for the understanding of the idea of obligation to see that in individual cases the statement that a person has an obligation under some rule and the
prediction that he is likely to suffer for disobedience may diverge; (2) It is clear that
obligation is not to be found in the gunman situation, though the simpler notion of being obliged to something may well be defined in the elements present there; (3) To
understand the general idea of obligation as a necessary preliminary to understanding
it in its legal form, we must turn to a different social situation which, unlike the gunman situation, includes the existence of social rule; for this situation contributes to
the meaning of the statement that a person has an obligation in two ways; (4) The
statement that someone has or is under an obligation does indeed imply the existence
of a rule; and (5) Rules are conceived and spoken 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 (CL 85-86; emphasis added).
218
exceptions; however, particularists are not skeptics about morality or about the idea that we have moral obligations. A
theory that purports simply to articulate the content of the
general concept-term obligation should not have controversial substantive implications about morality.
What we can say, however, is that obligations are associated with prescriptions, which include claims claims about
what someone (or some class of persons) ought to do in some
state of affairs and norms. Obligations arise only where
there are prescriptions that guide and enable the appraisal of
human acts. If I have an obligation to do A at t, then there is
some prescription that either expresses or implies that I
ought to do A at t. That is, it is a necessary condition for
someones being obligated to perform some act that there is a
prescription that expresses an obligation owed by that person
to perform that act.
Not every prescription expresses or implies an obligation.
Although all prescriptions purport to commend some behavior (or abstinence), not all prescriptions require them; there
are things I ought to do that I am not obligated to do. There
are, for example, prudential norms that, other things being
equal, express the idea that one ought to exercise regularly,
but those norms do not create or express obligations because
prudential norms are prescriptive but do not create requirements or obligations in any meaningful sense and therefore
could not be mandatory in the relevant sense. The only prescriptions that create or express obligations are mandatory
prescriptions i.e., prescriptions that require some act.
It therefore appears to be a necessary condition for P to be
obligated to do a that there is a mandatory prescription that
requires that P do a. If there is no mandatory prescription requiring a, then there is no obligation to perform a; the claim
that a is obligatory but not required by a mandatory prescription seems self-contradictory. Obligations are thus correlated with mandatory prescriptions.
219
KENNETH EINAR HIMMA
2. Obligations as Reasons
Obligations are commonly thought to correlate with reasons. On this view, the claim that X has an obligation to do
a implies that X has a reason to do a.6 If Y asks X for a justification for Xs doing a, X was obligated to do a, if true,
is always relevant in assessing whether doing a was justified from the standpoint of practical rationality.
The reason can be moral, but need not be. Some obligations are associated with moral reasons but not all obligations are. If, as many theorists believe, it is not true that
the status of a norm as law does not afford a prima facie
moral reason to obey it even in reasonably just states, then
it is reasonable to think that one does not have even a
prima facie moral reason to obey wicked laws that create legal obligations. There are clearly other kinds of reasons,
such as prudential although the number of different
types of basic reason (i.e., reasons that are irreducible to
other reasons) are limited.
Indeed, it is very difficult to think of any other basic reasons than prudential and moral reasons. Perhaps there are
aesthetic reasons as well. But if there are no other basic
reasons, then every other kind of reason, including legal
reasons, will ultimately be compound in character, ultimately constituted by some combination of members of the
set of basic reasons.
The reason might be conclusive, but it need not be. It
seems that, as an objective matter of practical rationality,
we have a conclusive reason for doing what we are morally
obligated to do all things considered. I have a reason not to
torture another innocent person no matter what else might
be true and hence a conclusive reason for not doing so. But
whatever prudential reason Nazis may have had to do mor6 Not everyone accepts this view. For example, Scott Shapiro believes that obligations merely purport create or be reasons. On his view, there can be obligations
that neither create nor are identical to reasons. The argument of this paper, however, depends on the denial of this view, which I cannot defend here.
220
ally wicked things, it was clearly not conclusive; taking into

account the relevant moral reasons, they had a conclusive
reason not to do these things.
If ordinary talk is any indication, obligations are reasons.7 Again, it is always a relevant consideration in justifying the performance some act a or, relatedly, in deliberating
whether to do a that one has an obligation of some kind to
do a. Because I had an obligation to do a might not be an
adequate answer to the question why did you do a?; it
might be false that I had an obligation to do a or it might be
true that I had such an obligation, but it was outweighed
by some a more important obligation. But if, as seems reasonable, only reasons can practically justify an act, then
obligations are reasons. Genuine obligations are necessarily
normative and hence are reasons for action.
3. Obligations as Exclusionary
Obligations are defined by valid mandatory prescriptions,
and mandatory prescriptions are fairly characterized as
exclusionary in this respect: As desires and prudential interests are generally irrelevant with respect to whether A
should perform an act required by a mandatory prescription. If A fails to do p and p is required by a mandatory prescription, it is not a justification, other things being equal,
7 It is worth noting here that ordinary talk about law (and the corresponding
legal practices) presuppose that one can have a reason independent of ones mental states, which is incompatible with reasons-internalism. According to the internalist, there are no reasons that are external to the agents mental states; an agent
has a reason for doing P if and only if the agent instantiates the appropriate mental
state usually a belief-desire pair. While the assumption that internalism is false
is controversial, I am concerned with giving a conceptual account that harmonizes
with our ordinary law talk and legal practices, which presupposes that there are
other kinds of reasons than simply the belief-desire pairs. Ordinary talk does not
imply the denial of such reasons but characterizes such reasons as subjective. Ordinary talk and legal practice seem to presuppose that moral and legal reasons are
objective in character. Internalism would entail something like an error theory of
law. That might ultimately be correct, but it takes an argument to establish that.
221
KENNETH EINAR HIMMA
that A did not want to do p or that doing p did not conduce

to As interests.8
This characteristic of obligation is also related to the concept of wrongness. An act is wrong if and only if it is not justified or excused (one justification would be, of course, that
the behavior is permissible). Mandatory prescriptions, as a
conceptual matter, exclude certain kinds of justifications for
non-performance, and it is the exclusion of these stories as
not constituting valid reasons for non-performance that
helps to explain why the relevant acts are properly thought
of as mandatory or required: an act that people are generally
free not to perform because it is trivially justified under a
prescription is not required by the prescription.
The claim here is that, as a conceptual matter, the reasons for doing p do not depend on its satisfying our own
particular prudential interests is entailed by the core of
what we mean when we say p is required by a valid mandatory prescription. The claim p is required by a mandatory
prescription N is inconsistent with the claim non-performance of p can be justified under N, as a general matter, by
purely prudential considerations in much the same way
that the claim that p is a bachelor is inconsistent with the
claim that p is married. Obligations that are defeasible by
reference to anyones prudential interests, no matter how
trivial, is as incoherent as the idea that some bachelors are
married.9
Although the term exclusionary is sometimes thought
to be synonymous with the Razian notion associated with
8 This, of course, draws heavily on Joseph Razs influential Practical Norms
and Reason (Princeton, 1990).
9 It is worth noting that the claim that obligations are exclusionary reasons
logically entails that there are no prudential obligations. A prudential obligation
would be a prudential reason indeed, a very strong one that excludes prudential reasons as a justification for non-performance of an obligation, an implication
that appears to be logically incoherent. A prudential obligation would require one
to do what one has most prudential reason to do and hence could not exclude prudential considerations as justifying non-performance. This strikes me as the correct result: we talk of the victim of a robber as being obliged, rather than obligated,
to comply at least in the absence of other factors such as responsibility to his or
her family.
222
the term pre-emptive reason, they are not synonymous as

defined above. A Razian pre-emptive reason has a certain
structure consisting of a first-order reason to do (or not do)
some act and a second-order reason not to act on ones assessment of the first-order reason. The idea that mandatory
prescriptions are exclusionary claims or presupposes nothing about the structure of the relevant reasons, and hence
does not assume that obligations give rise to second-order
reasons. The claim is merely that a mandatory prescription
is exclusionary in the limited sense of excluding certain stories as justifying or excusing non-performance.
4. The Special Normative Force of Obligations:
Obligations as Binding
The concepts of obligation and wrongness are related to the
concept of being (normatively) bound. Obligation-talk is frequently couched in terms of a relationship in which the subject of the obligation is bound to the norm. As Hart puts the
point, The figure of a bond binding the person obligated
is buried in the word obligation (CL 87). Obligations, according to ordinary intuitions, bind us.
In what sense? The term must (and, less frequently, the
term shall) is frequently used to express that we have an obligation and are hence bound to perform some act. We
may do what is permissible and should do what is good, but
we must do what is obligatory.
It might be tempting to explain the concept of bound in
terms of some sort of psychological or physical compulsion.
The idea here is that persons are bound by a rule creating an
obligation in the sense that they are psychologically or physically unfree to do other than what the rule requires. But not
every obligation, as a conceptual matter, is supported by
compulsion of this kind. Many persons do not feel psychologically compelled (i.e., psychologically unfree) to satisfy moral
obligations. Further, there are many obligations not supported by physical compulsion; we are not physically
compelled (i.e., unfree in some physical sense) not to lie. Here
223
KENNETH EINAR HIMMA
coercion and compulsion, it should be remembered, are two

different things: a gunman coerces me with the threat of death
but, other things being equal, cannot compel me to obey.
It might also be tempting to think that the exclusionary
character of obligations (or mandatory norms) is enough to
explain the binding quality of obligations but, as that concept has been explained in this paper, the exclusionary character of obligations, by itself, lacks the resources to explain
the binding quality of obligations. The claim that a mandatory norm is exclusionary says something about its content
or; that is, it expresses the idea, as we have seen, that the
content of the norm is such that it disqualifies certain stories as justifying non-performance. But the claim that a
mandatory norm binds us is a claim about its normative
force; this is the point of the metaphor of a bond that ties
us to rule (i.e., the normative force binds us to the rule).10
Simply knowing that the content of a norm excludes certain
considerations as justifying non-performance does not tell
us much, if anything, about the nature of this bond or the
special normative force that it has. Indeed, it doesnt even
tell us whether a norm that functions this way has any normative force because it tells us nothing about whether the
norm is valid or applicable. Invalid mandatory norms are
exclusionary in this limited sense, but they have no normative force and hence do not bind.
III. HARTS THEORY OF SOCIAL OBLIGATION
Legal obligation, as conceived by Hart and most positivists, belongs to a special class of obligations. Since positivism explains law as a set of social practices, the concept of
10 Of course, as Raz defines the term, exclusionary reasons are capable of binding. See Joseph Raz, Engaging Reason (Oxford: Oxford University Press, 2000). I
have not adopted the Razian account here because it is not obvious to me that it is
a conceptual truth that legal obligations are exclusionary reasons in this sense Although it seems clear that moral reasons are such reasons, it is not clear that very
wicked legal norms would generate a robustly exclusionary reason. I think most
theorists would concede no more than that law purports to create exclusionary
reasons in the stronger sense intended by Raz.
224
obligation applicable in legal practice must itself be explicable in terms of social practices. Legal obligation, then, is a
species of social obligation.11 A full explanation of the concept of legal obligation, then, requires an explanation of the
concept of social obligation, which must harmonize with
the explication of the general concept of obligation. Harts
account of social obligation is developed below.
1. Social Prescriptions
The first element is straightforward. Although not every
social norm gives rise to a social obligation (e.g., some create social powers), social obligations arise under general social prescriptions or social norms, which are created, as
Coleman puts it, by a convergence of attitude and behavior.
Persons in the group converge on taking the internal point
of view towards the norm, accepting it as a standard that
governs the behavior of people in the group, and generally
conform to its requirements. Thus, if people in the group
(1) self-consciously accept the norm (this need not be for
moral reasons); (2) generally conform to the norm; and
(3) take a critical reflective attitude toward the norm using it to evaluate the behavior of other members of the
group, then it is, on Harts view, a social norm governing
behavior in the group.
2. Acceptance and Exclusionary Norms
Taking the internal point of view towards a mandatory
norm, on Harts view, involves regarding oneself and others
in the relevant group as being obligated by the rule. Acceptance of such a norm involves some sort of durable commitment to subject ones own behavior to governance of the
rule and to evaluate the behaviors of other people according
to the rule. Someone who genuinely commits to subjecting
11 It is unlikely that social obligations create social reasons that are basic (or irreducible) in character. See note 7, above.
225
KENNETH EINAR HIMMA
her behavior to the rule will accept and participate in a host

of normative practices regarding the rule including practices that treat members of the group, including herself, as
obligated. Someone who accepts a rule defining an obligation will surely regard herself as being obligated by the rule.
This suggests that persons who accept a mandatory social norm will accept it as a reason for complying with and
treat it as being exclusionary in the sense described above.
A mandatory norm is exclusionary in character in the sense
that it excludes certain justifications for non-conformity,
but this does not imply that any particular person does
treat or should treat the rule as what Raz calls a pre-emptive reason in her deliberations. Insofar as the person who
accepts the mandatory social rule will treat it as a reason of
some kind. But someone who accepts a mandatory social
norm and conceives it as an exclusionary reason might
but need not treat the norm as a pre-emptive reason in
her deliberations about what to do. Moreover, if accepting a
rule gives one a reason for following it (for as long as one
accepts it), such a person has a reason for treating the
norm as exclusionary at least for as long as she accepts
the rule.
3. How Social Obligation Binds
While unilateral acceptance alone can explain a persons
adoption of a social norm as functioning as exclusionary,
unilateral acceptance, by itself, cannot explain the normative force of the obligations to which social norms give rise.
After all, unilateral acceptance can always be given and
withdrawn at will, and if that is all there is to the story, it is
hard to see how a durable social obligation could arise.
What explains the binding (and hence durable) quality of a
social obligation owed by a member of the social group is,
in part, the attitudes of other members of the social group
towards non-compliance.
226
Hart explains the binding character of social obligations

in terms of considerations ordinary persons are likely to regard as having normative significance. According to Hart,
[r]ules 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 (CL 85-86).12 Social pressure in
the form of a hostile reaction is something people with ordinary psychological characteristics tend to regard as having
normative force. Not everyone responds in the same way to
(or cares as much about) social disapproval, but it is an empirical fact that ordinary persons tend to dislike criticism
and hostility and are willing to take at least minimal steps to
avoid it.
A couple of points deserve attention. First, deviating behavior under the norm is generally regarded as a reason or
justification for the application of social pressure. The claim
is not just that, as a general matter, deviating behavior correlates with social pressure. Rather, it is that members who
accept the rule regard the rule as a reason for applying social pressure: For [those who take the internal point of
view towards a rule], the violation of a rule is not merely a
basis for the prediction that a hostile reaction will follow
but a reason for the hostility (CL 90). This will be true, as a
conceptual matter, for any form of social obligation, on
Harts view, including legal obligation.
Second, the claim is not that social pressure is sufficient
for social obligation; after all, the gunman exerts social
pressure on his victim. Rather, it is that a convergence of
attitude and behavior on a rule, together with the appropriate kind of social pressure, constitutes the norm as obligatory. Such pressure is likely supported by a belief that it is
warranted (though not necessarily morally warranted),
12 Such social pressure may take only the form of a general diffused hostile or
critical reaction (CL 86), but may also rise to the level for physical sanctions (CL 86);
in this latter case, the rules are properly regarded as a rudimentary or primitive
form of law (CL 86).
227
KENNETH EINAR HIMMA
which is related to two factors: (1) the acceptance of the social norm; and (2) the belief that the norm is important because necessary to the maintenance of social life or some
highly prized feature of it (CL 87).
Harts explanation of social obligation can be summed up
as follows:
Hartian Theory of Social Obligation (HTSO): X has a social
obligation to do p if and only if (1) members of the relevant
group converge in attitude and behavior on a norm N governing X that requires X to do p; and (2) N is supported by significant social pressure and (3) because N is thought important because necessary to the maintenance of social life or
some highly prized feature of it.
According to HTSO, it is the presence of the appropriate

social pressure in a context that includes the existence of a
practice along with certain beliefs about the importance of
the norm that explains the sense in which the obligation is,
as a conceptual matter, binding: social pressure appears as
a chain binding those who have obligations so that they are
not free to do what they want (CL 87). No matter how important a social norm N might be thought by relevant members of the group, it is incorrect to characterize it as defining an obligatory and hence binding requirement if not
supported, in some way, by the appropriate level of social
pressure. As Hart puts the view, such pressure is the primary characteristic of obligation (CL 87).
This implies neither that every person feels the force of
the social pressure that makes a social norm binding nor
that any person should feel this force. The claims here are
quite limited. They are purely descriptive because they
make no claims about what people should regard as reasons. Further, they make no claim about what any particular person in a social group might feel in response to social
pressure; as Hart points out, there is no contradiction in
saying of a hardened swindler that he had an obligation
to pay the rent but felt no pressure to pay (CL 88). The assumption is significantly weaker: as an empirical matter,
228
people tend to care about social pressure enough to modify

their behavior in many circumstances.
One might be tempted to interpret Harts remarks on social pressure and social obligation as making the weaker
claim that social pressure signals that people in the group
regard the norm as obligatory, rather than the stronger
claim that it contributes to constituting the norm as obligatory. I think this is mistaken for two reasons. First, Hart
clearly takes himself as giving an analysis of the concept of
social obligation: To understand the general idea of obligation as necessary preliminary to understanding it in its legal form, we must turn to a different social situation which,
unlike the gunman situation, includes the existence of social rules; for this situation contributes to the meaning of the
statement that a person has an obligation in two ways (CL
85). The elaboration of the idea that social pressure supports social obligation occurs two paragraphs later. Second,
Hart rejects Austins view largely on the strength of the
gunman example. It would be uncharitable in the extreme
to construe Hart as lacking a theory of social and legal obligation when he rejects Austin, in part, on his perceived failure to provide a satisfactory account! Finally, Hart himself
is clear in thinking that an analysis of the concept of legal
obligation is foundational to a conceptual theory of law: for
example, he writes, It will be recalled that the theory of law
as coercive orders notwithstanding its errors, started from
the perfectly correct appreciation of the fact that and it
should be clear that this is a metaphysical claim about
law where there is law, there human conduct is made in
some sense non-optional or obligatory (CL 82).
IV. A COMPREHENSIVE THEORY OF LEGAL OBLIGATION
1. Second-order Legal Obligation as Defined
by a Social Rule of Recognition
Ultimately, there are two conditions, on Harts view, necessary and sufficient for the existence of law and legal obli229
KENNETH EINAR HIMMA
gation. First, officials converge in taking the internal point

of view towards and conforming to a conventional rule of
recognition. Second, citizens generally comply with the
rules validated by the conventional rule of recognition.
First- and second-order mandatory norms in such a system
define legal obligations.
The idea that officials take the internal point of view towards the rule of recognition suggests that they accept and
treat it as an exclusionary reason in assessing their own
and other officials behavior. Like all forms of obligation, legal obligations are exclusionary in the sense that certain
stories are disqualified as excuses or justifications for
non-compliance; this is just true in virtue of what it means
for a behavior to be required by a mandatory norm. But
insofar as officials accept the rule as a standard governing
their behavior, they regard it as a reason and have a disposition to treat the rule as exclusionary in character.
It is important to recall here that Hart does not argue
that it is unilateral acceptance that binds an official to the
rule of recognition; that would be problematic because unilateral acceptance does not provide anything that necessarily has independent normative force given what we know
about the psychology of ordinary persons. Hart argues instead that it is the joint acceptance by officials together with
social pressure on each to conform to the rule of recognition
that together warrant characterizing the rule of recognition
as being obligatory.
Such pressure is likely to have normative force for officials because they can be presumed to care about what
other officials think. Voluntary membership in a social
group governed by norms signals that the member regards
at least some of the beliefs and actions of the other group
members as having significant motivational force. It is,
thus, reasonable to think that someone who seeks out
membership in a social group, at least if their motivations
are sincere and non-subversive, will regard such pressure
to conform as having significant motivational force.
230
This does not imply that the motivation for conforming to

a social norm must be explained in terms of a desire to
avoid the social pressure.13 I assume that most people want
to avoid the condemnation accompanying a murder conviction and hence regard the prospect as having motivational
force and this includes people who commit murders. But
the motives that explain why most people do not commit
murder make no reference at all to these prospects. What
explains why most people do not commit murder is, in part,
a subjective moral reaction to murder (any decent person
would be horrified at the thought of committing such an
act) and a lack of extreme anger and hostility.
The point of these sorts of social mechanisms in Harts
analysis, then, is not to explain why officials accept the rule
of recognition. Officials who take the internal point of view
towards the rule of recognition are presumably motivated to
conform to the rule by whatever desires brought them to officialdom to begin with. While officials would also presumably want to avoid the disapprobation of other persons in
the relevant groups, Hart is not committed to explaining
their behavior in terms of some necessary motivation to
avoid such social pressure. Social pressure explains how
the rule of recognition obligates, and not why officials accept this rule.
2. Second-order Obligation as Explanation
of First-order Citizen Obligation
Harts theory of second-order obligation will not explain
first-order legal obligation. Merely showing that officials can
obligate themselves through some mechanism does not
show that their acts qua officials can obligate citizens. The
claim that you and I have obligated ourselves to behave in a
13 As Hart puts this important point, [t]he fact that rules of obligation are generally supported by serious social pressure does not entail that to have an obligation under the rules is to experience feelings of compulsion or pressure (CL 88).
231
KENNETH EINAR HIMMA
particular way does not entail any claim about the obligations of other people.
Whether officials can obligate citizens depends, in part,
on to whom the officials owe their obligations. If the officials obligations under the rule of recognition are owed to
citizens, then it is reasonable to think that citizens are obligated by the norms valid under it. Given the logic of obligation, it is hard to make sense of the idea that a judge owes
an obligation to all citizens to incarcerate citizens who violate norm N if N does not obligate citizens. It would be odd
if the concept of legal obligation behaved this way.
But Harts practice theory implies only that the obligations owed by group-members are owed to other members.
Hart has nothing that would explain how obligations binding members of the group could be owed to anyone outside
it; there is nothing in the practice theory as it explains the
obligations of officials that entails that the obligation is
owed to citizens. All the theory claims is that officials owe
these obligations to one another and this says nothing
that would justify thinking official acts obligate citizens.
Of course, non-members might be obligated to follow
rules of groups to which they do not belong. Non-Muslims
are required to abide by certain conventions that Muslims
have accepted regarding behavior inside mosques, but this
is explained by other standards to which non-Muslims are
subject; non-Muslims have a duty to respect those conventions when in mosques. Since admission to mosques is conditioned on consent to abide by certain standards, one
shouldnt enter a mosque unless prepared to abide by the
relevant standards.
3. Coercive Enforcement and First-order Legal Obligation
Once law is explained in terms of a social rule of recognition accepted by officials in an efficacious legal system, citizen obligation in modern municipal legal systems seems
232
best explained in terms of the authorization of formal institutional mechanisms of coercive enforcement. The idea here
is not that coercive enforcement of a norm, by itself, constitutes the norm as obligatory; rather, it is that coercive enforcement of a social norm in a system that satisfies certain
properties including the institutionalization of the relevant set of norms constitutes it as legally obligatory. Coercive enforcement of a legal norm constitutes it as legally
obligatory upon citizens, in part, because (1) the norm belongs to an institutionalized system of norms (2) grounded
in recognition norms accepted and practiced by officials
and is (3) minimally efficacious in regulating citizen behavior.
Here it is important to emphasize the normative dimension of this practice. While officials of the legal system need
not regard a first-order law as a moral justification for enforcing the law against non-compliance, they regard it as a
legal reason or justification (i.e., a reason that is internal in
the sense that it is within the system of law) for such enforcement. Obligation is explained by a normative web of
practices that includes the legal authorization of formal enforcement mechanisms as a legal justification for applying
them to citizens for non-compliance.
Formal institutional enforcement should be distinguished
from sanctions. Enforcement sometimes involves punitive
intent, as it does in the case of a defendant who is being
prosecuted for murder under the criminal law. But it need
not involve such intent,14 as in the case of a judge ordering
damages for breach of contract.15 Such enforcement mech14 It is worth noting that Austin is careful to point this out: Considered as thus
abstracted from the command and the duty which it enforces, the evil to be incurred by disobedience is frequently styled a punishment. But, as punishments,
strictly so called, are only a class of sanctions, the term is too narrow to express the
meaning adequately (PJ 22).
15 As natural law theorist John Finnis puts it: Not all lawful coercion is by way
of sanction or punishment. Even the most developed legal systems rightly allow
the arrest of certain suspected offenders or potential offenders, and of persons and
things (e.g. ships) likely otherwise to escape due process of adjudication. Judgments may be executed, and some other classes of debts satisfied, by seizure, dis-
233
KENNETH EINAR HIMMA
anisms include sanctions but also include the courts

power of contempt, which backs every court order. Moreover, the courts authority over these mechanisms includes
the authority to refuse to enforce or recognize a defective
instrument of some kind, which might include a contract,
will, or even a statute. Refusal to enforce a defective contract is part of how courts coercively enforce the laws governing formation of a contract.
What constitutes a mandatory norm as legally obligatory
in modern municipal legal systems is that coercive enforcement is legally authorized. If the application of coercive
force for violations of a valid legal norm N is authorized by
some valid legal norm as a normative response to nonfeasance, then N is legally obligatory and its binding force is
constituted by the authorization of the relevant coercive
mechanisms. Of course, it is probably true that it is also a
necessary condition for the existence of a legal obligation is
that the application of the relevant coercive mechanisms
are reliably applied in cases where they are authorized. But
this is not part of what constitutes a norm as legally obligatory.
V. SUPPORTING CONSIDERATIONS
1. The Centrality of Coercive Enforcement
in Modern Judicial Practice
The availability of formal, institutional coercive enforcement mechanisms is a central feature of law in modern municipal legal systems. Most obviously, the criminal law is
characteristically backed with punishment. But such mechanisms also play a central role in civil law: the point of
bringing a civil lawsuit is to get a court order requiring the
defendant to do something. Sometimes the plaintiff seeks
damages; sometimes the plaintiff seeks specific perfortraint, forced sale. Finnis, Natural Law and Natural Rights (Oxford: Clarendon
Press, 1980), 261.
234
mance. However, any plaintiff who brings a civil suit in any

legal system remotely resembling this one is asking the
court not only for a judgment, but also a court order.
The court has authority to enforce its lawful orders by a
formal, institutional coercive mechanism known as the contempt sanction. It is this power that enables the judge to enforce her orders in civil cases where they cannot plausibly be
characterized as imposing direct or indirect sanctions. In systems like ours, every court order is backed by the legal authorization of the contempt sanction for non- compliance.
This suggests that coercion is central to legal systems resembling that of the U.S. Since the contempt sanction is
both coercive and universally available to courts to enforce
its orders in civil and criminal matters, it follows that every
criminal and civil law is ultimately backed with a coercive
mechanism (since the courts contempt sanction is coercive). The authority of the court to issue coercively enforced
orders is foundational to its ability to decide disputes in
systems like this one.
At the very least, this much seems reasonable: in cases
where (1) formal coercive mechanisms are generally authorized for non-compliance and (2) officials lack authority to
apply these mechanism in enforcing a particular judgment,
norm, or order with coercive mechanisms, it is implausible
to characterize the judgment, norm, or order as obligatory. Such norms are more fairly characterized as advisory because there is no sense in which the relevant behavior is made mandatory by mechanisms reasonably
presumed to have normative relevance given human beings
as we understand them.16
This is not to suggest that legal obligation cannot exist in
a legal system without formal, institutional coercive mechanisms, which would entail that such mechanisms are a
16 This should not be understood as implying that legal reasons are purely subjective belief-desire pairs. At most, it presupposes that objective reasons, if such
there are, are capable of functioning as subjective reasons in our practical deliberations given what we know about our psychological characteristics.
235
KENNETH EINAR HIMMA
conceptually necessary feature of law that is to say, that

law is necessarily coercive. For purposes of this paper, I am
agnostic with respect to whether there could be a system of
law in normative systems where only informal social pressure is available as a coercive mechanism. I tend to think
that law is necessarily coercive in this respect but nothing
in the argument here should be construed to presuppose
that view. Although the Hartian account of social obligation, as I have construed it, entails that social pressure is a
necessary condition for social obligation, the account of legal obligation here assumes only that some form of social
pressure is a necessary condition for legal obligation. The
specific view that the authorization of formal, institutional
coercive mechanism constitutes the binding force of obligation applies only to modern municipal legal systems like
that of the U.S.17
In any event, the authorization of such measures is a
more reliable indicator of a legal obligation than the language in which the relevant law is expressed. A statement
asserting that the defendant must or shall perform some
act is, despite its language, best characterized as advisory
if no coercive legal consequences are authorized for failure
to comply.18 Further, a statute asserting that people
should perform some act is, despite its language, best
characterized as obligatory if courts are authorized to incarcerate people who do not perform the act.
As a general matter, officials are quite careful to ensure
that the words of an authoritative statement of law adequately signal whether coercive enforcement mechanisms
are available, but this is explained by non-conceptual con17 Nor is it to suggest that the coercive mechanisms must be applied by an
agency that is part of a legal system. As a conceptual matter, the prison system
could be privatized, for example, without altering the status of an institutional system of norms as a legal system.
18 There are, of course, some laws that authorize sanctions but are chronically
unenforced. Whether or not these count as legally obligatory will be determined by
whether they count as valid under that systems recognition practices. In some legal systems, the chronic failure to enforce or apply a norm suffices invalidate the
law, a situation sometime described as repeal by desuetude.
236
siderations. Conscientious officials want to ensure that authoritative statements of law convey appropriate notice of
what is required. The terms must and shall, in contrast
to should and ought, signal that some behavior is
required and provide constructive notice to citizens that
courts have recourse to some coercive mechanisms though
such terms do not say anything about the nature or severity of such mechanisms.
Still, it is the availability or non-availability of coercive
mechanisms, and not the language in which a rule of law is
expressed, that ultimately determines whether that rule defines a legal obligation. When the language in which a legal
norm N is expressed and the availability of coercive enforcement mechanisms do not agree, it is the latter that determines whether N is fairly characterized as legally obligatory upon citizens.
2. Is Coercive Enforcement a Conceptually
Necessary Feature of Law?
Many theorists believe that coercive enforcement is a
conceptually necessary feature of law. Natural law theorists frequently acknowledge the central role coercion plays
in law. John Finnis, for example, observes that [l]aw needs
to be coercive (primarily by way of punitive sanctions, secondarily by way of preventive interventions and restraints).
Likewise, Ronald Dworkin believes the conceptual function
of law is to justify the states use of its police power and
hence that the law includes the moral principles that show
statutory and judicial law in their best moral light. Further, positivists, like Joseph Raz, also acknowledge the centrality of coercion in law: The three most general and important features of the law are that it is normative,
institutionalized, and coercive.19
19 Joseph Raz, The Concept of a Legal System, 2nd ed. (Oxford: Clarendon
Press, 1980), 3. Hereinafter CLS. Raz has changed his view on this issue. See Raz,
Practical Norms and Reason (Princeton: Princeton University Press, 1990).
237
KENNETH EINAR HIMMA
Intuitively, there is something to be said for this view. No

matter how closely it might resemble societies with legal
systems, a society of angels with rules promulgated under
a rule of recognition does not seem to have law if these
rules are not subject to coercive enforcement; such a society seems utopian and as having transcended law.20 Indeed, it is the absence of a centralized authority with coercive enforcement power that leads many scholars to believe
that international law, strictly speaking, really isnt law
at all.21
If this view is correct, then the theory that explains
first-order legal obligation in terms of coercive enforcement
has the advantage of explaining the essential role of coercion in law by linking it to another concept central to law
namely, the concept of legal obligation. The central role coercion plays in every conceptually possible legal system is
explained by its conceptual role in defining the first-order
obligations of citizens. Moreover, it would provide a link between the claim that it is a conceptual truth that first-order
legal norms are enforced by the states police power and the
claim that it is a conceptual truth that first-order legal
norms define citizen obligations.
In any event, the theory defended here neither assumes
nor implies that coercive enforcement is a necessary feature
of law. This theory purports to explain legal obligation only
in systems, like those most familiar to us, generally backed
by coercive enforcement. It does not purport to explain legal
obligation in systems where mandatory norms are backed
only by generalized social pressure of the sort that typically
backs social obligations. Of course, in such systems (which
otherwise satisfy the conceptual prerequisites for law), the
foregoing analysis suggests that what constitutes such
20 For example, Finnis asks, Would there be a need for legal authority and regulation in a world in which there was no recalcitrance and hence no need for sanctions? (NLNR 266).
21 Hart, for example, observes that one good reason for thinking that what we
call international law is not really law at all is that there is no centrally organized
effective system of sanctions (CL 4).
238
norms as legally obligatory, in part, is that they are backed

by generalized social pressure.
This is a virtue, I think, because the jury remains out on
the conceptual necessity of formal coercive enforcement
mechanisms in law. Though he sometimes characterizes
systems lacking formal enforcement as pre-legal, Hart
more frequently characterizes them as being rudimentary
or primitive systems of law (CL 84).22 Indeed, Hart generally speaks of such mechanisms as being common but not
necessary: the typical form of legal pressure [supporting legal obligation] may very well be said to consist in such
threats [of physical punishment or unpleasant consequences] (CL 179, 180; emphasis added).
3. The Binding Force of Obligation
This theory explains the bindingness of mandatory legal
norms in terms of considerations likely to be regarded by
subjects as normatively relevant. First, being subject to coercively enforcement is a clear sense in which that norm
can plausibly be characterized as being non-optional. Second, the authorization of coercive enforcement mechanisms
including the contempt power) is something that is normatively relevant to any rational citizen. This, again, is not to
claim that citizens are necessarily motivated to obey the law
by a fear of sanctions; rather, the point is merely that rational self-interested citizen are, as descriptive matter, likely
to care about avoiding the coercive enforcement power of
the state.
One might worry, however, that the sort of reason provided this theory of legal obligation is, as a conceptual matter, the wrong kind of reason. In particular, one might object that this account explains the normative legal
obligation in terms of prudential considerations and hence
reduces legal reason to first-order prudential reasons. This
22
I am indebted to Scott Shapiro for this point.
239
KENNETH EINAR HIMMA
is problematic insofar as one thinks (1) prudential reasons

are not the only basic reasons constituting a legal reason
and (2) legal reasons are pre-emptive reasons.
As to (1), it seems clear that legal reasons, on a positivist
view, being the product of a human artifact manufactured
by social processes (i.e., a legal system) would have to be a
compound reason reducible to basic reasons. And it is clear
that a positivist cannot hold that it is a conceptual truth
that a legal reason is partly reducible to a moral reason
without violating the Separability Thesis that there are no
necessary moral constraints on the content of law. As, we
saw above, there is a limited palate of basic reasons to
choose from: there seem to be no other kinds of basic reason other than prudential, moral, and possibly aesthetic
reasons. And it should be clear that legal reasons are not
constituted, even in part, by basic aesthetic reasons if such
there be. If legal reasons are compound, the only kind of
reason they could be reduced to are prudential reasons.
As to (2), the idea that legal reasons are pre-emptive reasons is contentious. While it is clear that mandatory legal
norms are, by the very nature, exclusionary in the sense
that they exclude certain justifications for non-performance, this does not, by itself, imply that the reasons created by such norms are pre-emptive in the Razian sense.
Given the fact that Razian account of authoritative reasons
is contentious, the objection simply begs the question
against the account offered here.
4. The Right Kind of Normativity
The idea that the authorization of coercive enforcement
constitutes a mandatory norm as legally obligatory harmonizes nicely with another important idea concerning legal
obligation namely that there is no prima facie moral reason to obey the law. Most theorists have come to reject not
only the idea that the law necessarily gives rise to moral obligations, but also the weaker idea that it is necessarily the
240
case that we have a moral reason to obey legal requirements; indeed, many theorists are even skeptical about the
idea that law in a legitimate state necessarily gives rise to a
moral obligation to obey. If this plausible view is correct,
then the fact that a mandatory legal norm creates a legal
obligation does not imply that it creates a moral obligation
to obey it or even that there is a prima facie moral reason
to obey it.
This harmonizes nicely with the theory of first-order legal
obligation defended here. The only reasons for action that
are necessarily provided by a legally obligatory norm, if the
theory here is correct, are prudential in character. Clearly,
first-order legal obligation would be prudentially normative
on the story offered here: it is not in the interests of a person, other things being equal, to be subject to the sorts of
coercive mechanisms that are used to enforce mandatory
legal norms. Equally clearly, first-order legal obligation is
not necessarily morally normative on this story: there is
nothing in the claim that the state has backed a norm with
coercive enforcement mechanisms that would imply that
there is even a prima facie moral reason to obey that norm.
This is exactly what we would expect if the prevailing
view that law does not necessarily give rise to prima facie
moral reasons to obey the law is correct. An analysis of legal obligation that implies we have even a prima facie moral
reason to satisfy our legal obligations would be inconsistent
with this view. The fact that, on the analysis offered here,
legal obligation is not necessarily morally normative is a
strong point in its favor.
Nevertheless, it is important to note that legal obligation
is, as a conceptual matter, normative on the analysis offered here. Insofar as people have a prima facie prudential
reason to avoid having a norm coercively enforced against
them, they have a prima facie prudential reason to obey any
mandatory legal norm. But this coheres nicely with the prevailing view that it is a conceptual truth that law is norma-
241
KENNETH EINAR HIMMA
tive; since mandatory legal norms are at least prudentially

normative, they are, a fortiori, normative.
Accordingly, law provides content-independent considerations that a practically rational subject will regard as relevant from the standpoint or prudential rationality or, as it
sometimes put, content-independent reasons for action.
These reasons need not be conclusive and might be outweighed for the actor by other considerations, but the authorization of coercive enforcement for a valid law always
seems to provide some reason for complying with the laws
requirements.23 Legal obligation is thus, on this analysis,
necessarily normative but not necessarily morally normative.
23
See footnote 14. Norms may become invalid through desuetude.
242
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
HARTS THE CONCEPT OF LAW AS A STUDY IN SOCIAL

PHILOSOPHY
Eerik LAGERSPETZ*
Resumen:
El concepto de derecho de H. L. A. Hart es un clsico moderno de la teora jurdica. Pero tambin es relevante para la filosofa moral y social. En
este artculo se argumenta que las crticas de Hart hacia Austin y Bentham, as como sus teoras sobre la naturaleza del derecho, la moral y la
coercin, retoman una tradicin de pensamiento iniciada por David
Hume. Adems, la filosofa social de Hart tiene implicaciones normativas
interesantes. Este artculo se concentrar en un solo ejemplo: el concepto de libertad. A pesar de que Hart no analiza el concepto de libertad en
su libro, el autor sostiene que su anlisis del derecho puede ayudarnos a
comprender esa nocin tan altamente discutida.
Palabras clave:
Teora jurdica, filosofa social, filosofa moral, libertad, H. L. A.
Hart.
Abstract:
H. L. A. Harts The Concept of Law is a modern classic of legal theory.
However, it is also relevant to moral and social philosophy. Here, it is argued that Harts criticism of Austin and Bentham, and his views on the nature of law, morality and coercion continue the tradition initiated by David
* Professor of Practical Philosophy. Department of Behavioural Sciences and
Philosophy. 20014-University of Turku Finland. eerlag@utu.fi.
243
EERIK LAGERSPETZ
Hume. Moreover, Harts social philosophy has interesting normative implications. I shall focus on one single example: the concept of liberty. Although
Hart does not analyze the concept of liberty in his book, I argue that his
analysis of law may help us to understand that highly contested notion.
Keywords:
Legal Theory, Social Philosophy, Moral Philosophy, Liberty,
H. L. A. Hart.
244
HARTS THE CONCEPT OF LAW
SUMMARY: I. Introduction. II. Two Theories of Law and Coercion. III. The Third Theory: Hume and Hart.
IV. Law and Liberty. V. Bibliography.
I. INTRODUCTION
H. L. A. Harts The Concept of Law (hereafter CL) is undeniably a masterpiece of the twentieth-century philosophy of
law. However, in the preface of the book, Hart defines the
scope of his work in wider terms:
My aim in this book has been to further the understanding
of law, coercion, and morality as different but related social
phenomena. Though it is primarily designed for the student
of jurisprudence, I hope it may also be of use to those whose
chief interests are in moral or political philosophy, or in sociology, rather than in law.1
The claim that Hart was an important social philosopher

hardly requires an argument. It is enough to remind that
for example, the seminal concept of fairness was first introduced to philosophical discussion by Hart, not by John
Rawls as some people might think. The topic of this paper
is the moral and political (or social) philosophy of Harts
best-known work. First, I try to place CL into the tradition
of Western socio-political thinking. Then, I briefly discuss
the possible normative implications of Harts social philosophy. I shall focus on one single example: the concept of liberty. Although Hart does not analyze the concept of liberty
in CL, I argue that his analysis of law may help us to
understand that highly contested notion.
II. TWO THEORIES OF LAW AND COERCION
In the first chapters of CL, Hart target is the imperative-coercive theory (IC-theory); the view formulated by
1 Hart, H.L.A., The Concept of Law, Oxford, Great Britain, Oxford University
Press, 1961, p, vii.
245
EERIK LAGERSPETZ
Thomas Hobbes2 and first refined and canonized by Jeremy

Bentham3 and then by John Austin.4 In spite of their differences, Hobbes, Bentham, and Austin all accepted the key
elements of the IC-view. The key elements are the following.
Law can be reduced to duties, and duties are basically commands, addressed to the subjects by their superiors. All duties are attached with sanctions which, ultimately, consist
in physical coercion. The coercive element in law is required, for people are generally unwilling to perform their
duties voluntarily. The reason for this is that, due to the
human nature, self-interest tends to be the most powerful
motive for action. Fear of coercion becomes, therefore, the
most important reason for people to obey law. It is in the
general interest that people perform their duties, and fear
provides a self-interested motive for most people to perform
them. Any society is divided into two groups: the coercers
and the coerced although the coercers may themselves be
subjected to coercion, too. The ultimate coercer is the sovereign, the legal entity who gives commands but does not
obey anyone and is, therefore, not subject to any (legal) duties. Thus, the IC-theory is not only about law, but more
widely about law, coercion, and morality as different but
related social phenomena.
Although Hobbes, Bentham and Austin are historically
the most important exponents of the IC-view, many of its
elements are also accepted by other legal theorists. For example the Scandinavian Realists such as Karl Olivecrona,5
and some modern positivists like Kelsen while rejecting
the imperative aspects of the IC-theory and its necessary
consequence, the sovereign still seem to subscribe the
2
Hobbes, Thomas, Leviathan, Oxford, Great Britain, Oxford University Press,
1996.
3 Bentham, Jeremy, Selected Writings on Utilitarianism. Hertfordshire, Great
Britain, Wordsworth Classics, 2001.
4 Austin, John, The Province of Jurisprudence Determined, Cambridge, Great
Britain, Cambridge University Press, 1995.
5 Olivecrona, Karl, Law as Fact, London, Great Britain, Oxford University
Press, 1939.
246
theses that law is essentially coercive, and that fear is the

central motive for obedience. Similarly, the contemporary
Neo-Hobbesian social theorists, most notably Nobel-laureate economist James Buchanan,6 accept the main elements
of the IC-theory. The Neo-Hobbesians interpret the motivational conflict between self-interest and duties in game-theoretic terms, as an instance of the Prisoners Dilemma
game. This is an important achievement, for it makes
Hobbes theory more acceptable: although individuals in
the Hobbesian world are guided only by their self-interest
and therefore do not voluntarily perform their duties a
legal system which compels them to perform their duties is
still mutually beneficial. The earlier, standard reading of
Hobbes was more Augustinian: people are simply evil and
have to be controlled. Nevertheless, even in the NeoHobbesian interpretation, fear of sanctions is the overriding
motive for each individual to follow law.
Of course, the IC-view has never been universally accepted. Legal and political theorists, from Plato and Aristotle via the classical tradition of Natural Law to sociologists such Durkheim and Parsons, have emphasized the
role of consensus, shared values and voluntary cooperation
as necessary for the continuous existence of societies. Sociologist Talcott Parsons put the consensualist thesis into a
nutshell:
A relatively established politically organized community is
clearly a moral community to some degree, its members
sharing common norms, values and culture which is to
stay that I start with a view that repudiates the idea that any
political system that rests entirely on self-interest, force, or a
combination of them, can be stable over a considerable period of time.7
6 Buchanan, James, The Limits of Liberty. Between Anarchy and Leviathan,

Chicago, USA, Chicago University Press, 1975.
7 Parsons, Talcott, The Modern Sociological Theory, New York, USA, Free Press,
1967, p. 265.
247
EERIK LAGERSPETZ
Neither the IC-theory nor its consensualist antithesis

seems to be fully compatible with our everyday experience.
The IC-theory is problematic at least for two reasons. First,
a modern constitutional system is not like a tyranny. Normally and typically, the officials perform their legally defined duties and ordinary citizens, more often than not, follow the law even when there are no police officers around.
The fear of sanction does not in most cases play an important role in our practical reasoning. Second, the theory
faces a problem of infinite regress. If fear of sanctions were
the most important motive for obeying rules, the existence
of an effective constitution would be inexplicable, for the legal Leviathan could be constrained only by establishing another Leviathan that could interpret and enforce the constraints, and so on. According to John Austin both
international law and constitutional law are only positive
morality because they are not backed by sanctions.8 However, he fails to explain why this morality motivates. As Alf
Ross once remarked, ultimately there must be norms for
the exercise of force which are not themselves upheld by
force but are followed in fearless allegiance.9 But if this
fearless allegiance is possible, why suppose that only
ultimate legal norms can have this character?
The consensualist antithesis does not fare any better. If
there were a wide consensus on values in our (or any) society, we should at least be able to spell out the content of
that consensus.10 (Hardin) A small community may remain
stable because it is a moral community in the Parsonian
sense. But how well does the notion fit into a modern
State? Which are the values we all share? If we try to articulate the values shared by the citizens of the modern
State, we end up to vague and abstract principles like human life is valuable or democracy is good. But an agreeAustin, John, op. cit., pp. 213-17.
Ross, Alf, On Law and Justice, London, Great Britain, Stevens, 1958, p. 57.
10 Hardin, Russell, David Hume: Moral & Political Theorist. Oxford, Great Britain, Oxford University Press, 2007, p.125.
8
9
248
ment on such principles is not sufficient to maintain a legal

order. Innumerable wars and civil strives have been fought
between religious groups or dynastic factions sharing the
same values, at least on that level of abstraction. The contrast between individual self-interest and shared values is
not crucial. (To put it crudely history shows that Christians
have mainly fought against Christians, Muslims against
their fellow Muslims, nationalists against the nationalists of
the neighbouring country and so on).
There is, however, a third possibility, classically formulated by David Hume.11 This view recognizes that we humans are social beings whose condition is that of a mutual
interdependency. Yet, it admits that our sociability is limited and particular. In one sense, this view is a compromise
between the extremes. Yet, like all good middle-road solutions, it does not just provide an eclectic mixture of the
competing views, but, as an Aufhebung of an intellectual
conflict, it tries to explain why the previous views are both
partly correct.
III. THE THIRD THEORY: HUME AND HART
In my interpretation, the socio-political theory which
forms the background of The Concept of Law is essentially
Humean, although direct references to Hume in CL are
scanty. The part of Harts book most directly influenced by
Hume is Chapter IX, Law and Morals, and particularly the
sub-chapter entitled The Minimum Content of Natural
Law.12 Harts argument in Chapter IX is often interpreted
as a mild concession to the contemporary Natural Law theories. I think that this is a misinterpretation. The purpose
of the chapter is not to find the alleged sensible core of
the modern theories of Natural Law, but rather to lay a basis for Harts own theory, as developed in the earlier chap11 Hume, David, Humes Moral and Political Philosophy, Darien, USA, Hafner
Publishing, 1970.
12 Hart, H. L. A., op.cit., pp. 189-195.
249
EERIK LAGERSPETZ
ters of the book. Although, due to Harts argumentative

strategy in CL, the discussion on the Minimum Content
appears in the last part of the work, it could equally well be
the first chapter. Chapter IX is focused on those properties
of human nature which, according to him, make rules indispensable elements of our social life. These properties are
(1) vulnerability, (2) approximate equality, (3) limited altruism, (4) limited resources, and (5) both limited understanding and strength of will.13 Because of these limitations, cooperation between human beings tends to be fragile. In
order to overcome the problems of cooperation and interaction, human beings need rules. Harts list follows closely
the list of the Conditions of Justice given by Hume.14 After
Hart, similar lists have been produced, with slight modifications, by political theorists such as J. R. Lucas,15 John
Rawls16 and Jeremy Waldron,17 and by moral philosophers
such as G. J. Warnock18 among others.
In his main works, David Hume tried to show that our
fundamental duties or obligations most notably those related to property and to promises are artificial rather
than natural. Our basic social duties are neither imprinted to our minds nor results of an original agreement.
They have developed gradually as responses to the practical
problems resulting from our limited nature. Their role is to
help human to overcome their natural limitations in their
mutual interactions.
In a sense, Hart begins where Hume finishes. Primary social obligations are like Humes conventions of justice
Idem.
Hume, David, A Treatise of Human Nature, in Hume, op. cit., pp. 57-8.
15 Lucas, J. R., The Principles of Politics, Oxford, Great Britain, Clarendon
Press, 1966, pp. 5-10.
16 Rawls, John, A Theory of Justice, Oxford, Great Britain, Oxford University
Press 1971, pp. 126-30.
17 Waldron, Jeremy, Law and Disagreement, Oxford, Great Britain, Oxford University Press 2001, p. 102.
18 Warnock, G. J., The Object of Morality London, Great Britain, Methuen & Co.,
1972, pp. 17-26.
13
14
250
public, shared standards of conduct existing in particular

communities.19 Their role is to maintain interaction among
vulnerable beings whose altruism, understanding and
strength of will are limited. However, when these communities become larger and internally more complex, the internal limitations of human nature tend to create new problems. First, when the rules of obligation (and the
interactions regulated by them) become more complex,
problems of interpretation arise. Second, the primary obligations are enforced only by the critical reactions of the
members of the community. If the reactions are not coordinated those transgressing the rules are either not punished or punished too heavily the rules tend to lose their
effectiveness. This becomes more likely when the community grows and becomes more anonymous. Third, a regime
of primary rules tends to be too static. When the external
circumstances change, the same social values could perhaps be realized in a better way by adopting new rules. But
somebody should initiate the needed change, and the fact
that the rules are going to change should be communicated
to the members of society.20 For Hart, uncertainty, inefficiency and the static nature of the regime of primary rules
are reasons for a community to adopt secondary rules,
rules about rules.21 These rules provide for operations
which lead not merely to physical movement or change, but
to creation or variation of duties or obligations.22 Of
course, even a Humean simple regime of primary rules has
some secondary rules related to promising and contracts.
The rules analyzed by Hart, those related to the interpretation, enforcement and change of all other rules form only
an important subclass, called here as the meta-rules.
The emergence of meta-rules is an artificiality of second
degree. If the emergence of rules described by Hume marks
19
20
21
22
Hart, H. L. A., op. cit., p. 89.

Ibidem, p. 90-1.
Ibidem, p. 91-6.
Ibidem, p. 79.
251
EERIK LAGERSPETZ
the transition from natural sympathy to moral judgments,

the emergence of meta-rules envisaged by Hart marks a
further bifurcation between morality and law. Under the
simple regime of primary rules, the positivist separation between law and (prevailing, accepted) morality is irrelevant.
Admittedly, a community governed only by shared rules
backed by social pressure may be described as a simple legal system. It is however equally plausible to say that such
a community has no law, but only shared moral rules. In
contrast, in a modern legal system the distinction between
law and morality is clear. Although the basic rules of human conduct are still directly accepted, irrespective of
their legal status, any modern legal system contains innumerable rules which are not generally known, or understood, or endorsed by the general population. They exist because they are enacted, interpreted, applied, and enforced
by the limited sub-population of officials. The meta-rules,
most notably the famous rule of recognition, exist because
they are accepted as public standards of conduct within
that sub-population. In a sense, the positivist thesis that
law is based on social facts is itself grounded on a contingent social fact. The separation between moral standards
and legal standards which is the hallmark of positivism is
applicable only in modern legal systems.23
How does this view about law, coercion, and morality
differ from the imperative-coercive view? First, the role of
law is not just to constrain action. For Hume, law has essentially a coordinative role. This aspect of Humes philosophy has been brilliantly explicated by Russell Hardin in his
recent book.24 It becomes clearer when we compare it with
the Neo-Hobbesian idea of law as a solution to Prisoners
Dilemmas (PDs). A Prisoners Dilemma situation is a
mixed-motive game so that although the players prefer uni23 For a more detailed argument, see Lagerspetz, Eerik, The Opposite Mirrors.
An Essay on the Conventionalist Theory of Institutions, Kluwer, Dordrecht, The
Netherlands,1995, ch. 7.
24 Hardin, Russell, op. cit., ch. 4.
252
versal cooperation to universal defection, each individual

player still prefers a situation in which he or she can be a
free rider while the others are dutifully doing their task. In
a sense, it is misleading to say that there is a dilemma of
cooperation in the standard Prisoners Dilemma setting.
There is no dilemma simply because if a situation really is a
Prisoners Dilemma setting, the players do not cooperate,
period. Any solution to the problem for example, the
Hobbesian sanctions or the Durkheimian solidarity transforms the situation into another game by adding new motivational elements not reflected in the initial description.
The real dilemma is this: why and how actors who are involved in an interaction which prima facie looks like a standard PD may nevertheless be able to avoid the trap? In
other words, why, in spite of all appearances, the situation
in question is not really a Prisoners Dilemma after all? The
Humean answer is that the fundamental interaction problems resulting from the deficiencies of our nature are typically coordination problems, not PDs. In coordination situations, the observance of rules is self-maintaining: each
player has a motive to follow a rule if the others do. Thus
we have a rational motive to drive on the right (or left) side,
use the standard time, the national currency, and the language generally spoken in the community we happen to reside. This does not mean that we are necessarily happy
with the prevailing practices. We might prefer to drive on
the other side of the road or to speak some other language.
However, when all or most of the others follow a practice,
we do not have a rational reason to deviate. Such rules
should be analyzed as conventions rather than as contracts
or commands. A very small tightly-knit community could
perhaps live without conventions. For a large anonymous
community, they are necessary.25 Hart adds a further
non-coercive aspect to this Humean picture. The problems
enumerated in CL, those of uncertainty, inefficiency and
the static nature of primary rules can be interpreted as co25
Hume, David, op. cit., p. 67.
253
EERIK LAGERSPETZ
ordination problems. This gives rise to the meta-rules, the

secondary rules regulating the recognition, interpretation
and enforcement of other rules. These rules are typically
followed by the officials in fearless allegiance. Hence, they
are conventions rather than sanctioned commands.
One may argue that Harts Humeanism actually goes too
far. In his essay Of the First Principles Government Hume
famously argues that because force is always on the side of
the governed, the rulers must, in order to govern, enjoy at
least some voluntary support:
It is, therefore, on opinion only that government is founded,
and this maxim extends to the most despotic and most military governments as well as to the most free and most popular. The soldan of Egypt or the emperor of Rome might drive
his harmless subjects like brute beasts against their sentiments and inclination. But he must at least have led his mameluks or praetorian bands like men, by their opinion.26
Hence, fear cannot be sole motive for obedience in any

society. This Praetorian Guard argument has been accepted
as self-evident by theorists as diverse as J. R. Lucas,27
Hannah Arendt,28 Jos Ortega y Gasset29 and H. L. A.
Hart.30 However, philosopher Gregory S. Kavka31 and sociologist Barry Barnes32 have carefully argued that the Praetorian Guard Argument is not valid. To simplify, suppose that
there is a society consisting of the ruler, A, and of three
subjects, B, C, and D. Suppose that B obeys As commands
26 Hume, David, Of the First Principles of Government, in Hume, David, op.
cit., p. 307.
27 Lucas, J. R., op. cit., p. 75-6.
28 Arendt, Hannah, On Violence, in Arendt, Hannah, The Crises of the Republic, New York, USA, Harcourt, 1972, p. 140.
29 Ortega y Gasset, Jos, La rebelin de las masas, Madrid, Spain, Tecnos,
2003, p. 272-5.
30 Hart, H. L. A., op. cit., pp. 23, 89, 196, 198.
31 Kavka, Gregory S., Hobbesian Moral and Political Philosophy, Princeton, USA,
Princeton University Press, 1986, pp. 262-6.
32 Barnes, Barry, The Nature of Power, Oxford, Great Britain, Polity Press,
1988, pp. 96-100.
254
only because he believes that C and D obey them voluntarily and are willing to punish him if he refuses to obey.
The same is true of C and D, respectively. All the subjects
of A obey him only because of fear. Nevertheless, the force
is on their side: by acting as a coalition, they could easily
overthrow A. Only the pluralistic ignorance (to use a sociological term) prevents this. No loyal Praetorian Guard is
needed to explain As power over his subjects.
Still, the Praetorian Guard Argument is a plausible empirical conjecture. It is unlikely that a large modern society
could, in the long run, be governed only by coercion. Harts
methodology in CL is to focus on the central cases rather
than to define concepts through necessary and sufficient
conditions.33 This kind of approach is, I think necessary in
social philosophy and, more generally, in the human sciences. (An example to illustrate this is the notion of culture. It is certainly indispensable in these disciplines, but
how to define it precisely without depriving its descriptive
usefulness?) This approach would allow Hart to accept a
purely coercive regime imagined by Kavka and others as a
limiting case, but still insist that in the cases relevant to
legal and social theory his central thesis holds:
Sanctions are therefore required not as the normal motive
for obedience, but as a guarantee that those who would voluntarily obey shall not sacrificed to those who would not. To
obey, without this, would be to risk going to the wall. Given
this standing danger, what reason demands is voluntary
co-operation in a coercive system.34
To clarify, while the presence of sanctions remains psychologically important, it need not be important because
people fear sanctions. Rather, the mutual awareness of the
presence of sanctions is important because it maintains
voluntary cooperation. The possibility to rely on sanctions
has two important effects. First, it diminishes uncertainty:
33
34
Hart, H. L. A, op. cit., pp. 3-5.

Ibidem, p. 193.
255
EERIK LAGERSPETZ
the law-abiding citizen knows that if others are, for some

reason or another, not willing to follow rules voluntarily,
they will do so because they are afraid of sanctions. This
may be essential in large and anonymous societies where
people cannot only rely on personal trust based on mutual
acquaintance. Second, sanctions may be important to obedience because most peoples sense of justice requires that
wrongdoers are punished, and one important reason for
obeying the law is that it is perceived as being at least minimally just. To move from the purely coercive regime imagined by Kavka and others to the opposite limiting case:
even a system in which all citizens are willing to obey the
law if they believe that the others obey may need sanctions
in order to eliminate the uncertainty.35
IV. LAW AND LIBERTY
The problem with the IC-theory is not only that it cannot
account meta-rules. In CL, Hart shows quite effectively
that, at most, only a subset of ordinary laws can be conceived in terms of duties backed by sanctions. The key passages are the following:
Legal rules defining the ways in which valid contracts or
wills or marriages are made do not require persons to act in
certain ways whether they wish to or not. Such laws do not
impose duties or obligations. Instead, they provide individuals with facilities for realizing their wishes, by conferring legal powers upon them to create, by certain specific procedures and subject to certain conditions, structures of rights
and duties within the coercive framework of law.
The power thus conferred on individuals to mould their legal relations with others by contracts, wills, marriages &c., is
one of the great contributions of law to social life; and it is a
feature of law obscured by representing all law as a matter of
orders backed by threats.36
35
36
Lagerspetz, op. cit., pp. 198-207.

Hart, H. L. A., op. cit., pp. 27-8.
256
The fundamental point in these passages (and elsewhere

in CL) is that law has a constitutive role. The legal system
does not just regulate, constrain and direct pre-existing
forms of human behaviour. Rather it, like other complex
practices, creates new types of action which did not exist
before the emergence of the practice. Hume already remarked how it is impossible for men so much as to murder
each others without statutes, and maxims, and an idea of
justice and honour.37 Again, Hart continues a theme
initiated by Hume:
If such rules of this distinctive kind did not exist we should
lack some of the most familiar concepts of social life, since
these logically presuppose the existence of such rules. Just
as there could be no crimes or offences and so no murders
or thefts if there were no criminal laws of the mandatory
kind which do resemble order backed by threats, so there
could be no buying, selling, gifts, wills, or marriages if there
were no power-conferring rules.38
The non-constraining aspects of law emphasized by

Hume and Hart are related to the motivational aspect discussed above: if laws have other functions besides the constraining function, fear of sanctions need not to be the
most important motive for following it. Harts criticism is, I
think, a decisive one. The IC-theory cannot capture the
power-conferring aspects of law. Hence, it is inadequate.
The criticism made by Hart has largely been accepted
and it has become a part of the received wisdom. Legal theorists that are nowadays willing to defend the central theses of the IC-theory are few and far between. However, the
influence of the IC-theory goes beyond the legal theory
strictly defined. Hobbes and Bentham both tried to articulate a more general social philosophy, and some of its aspects are still at the centre of discussion. Therefore, Harts
37 Hume, David, Enquiry Concerning the Principles of Morals, in Hume David, op. cit., p. 206.
38 Hart, H. L. A., op. cit., p. 32.
257
EERIK LAGERSPETZ
observation that the theory is unable to account for some

important aspects of law has further consequences which
are also relevant to moral and political philosophy. Here, I
shall mention only one of them. In my interpretation, Harts
analysis gives a reason to rethink our notion of liberty, although Hart himself does not raise the issue directly in CL.
For Hobbes and Bentham, the IC-analysis of law was
only a part of a more general theory about law, coercion,
and morality. The negative concept of liberty was a necessary correlate of their view of law. As Bentham wrote:
liberty then is neither more nor less than the absence of coercion. This is the genuine, original and proper sense of the
word liberty. This idea is an idea purely negative. It is not
anything produced by law.39
According to this view, liberty, at least politically relevant

liberty, consists simply of the silence of laws, of the absence of coercive constraints. Not only is law unable to produce liberty. Because law is essentially coercive, every law
is contrary to liberty.40 Liberty-rights imply duties, and, as
we have seen, duties are considered as coercively enforceable commands. One consequence of this view is that the
sum of liberties has to remain constant.
As against the coercion applicable by individual to individual, no liberty can be given to one man but in proportion as
it is taken from another. All coercive laws, therefore () and
in particular all laws creative of liberty, are as far as they go,
abrogative of liberty.41
39 Benthams unpublished manuscripts. Quoted after Long, Douglas G.,

Bentham on Liberty, Toronto, Canada, Universityof, Toronto Press, 1977, p. 74.
40 Bentham, Jeremy, Principles of the Civil Code. Part I Objects of the Civil
Law, in Bentham, Jeremy, Selected Writings on Utilitarianism, Hertsfordshire,
Great Britain, Wordsworth Classics, 2001, p. 314.
41 Bentham, Jeremy, Anarchical Fallacies, in Bentham, Jeremy, Selected
Writings on Utilitarianism, Hertsfordshire, Great Britain, Wordsworth Classics,
2001, p. 401.
258
Hobbes and Benthams social topology is very simple.

Individuals are like freely-moving bodies. Laws, like physical barriers, limit their movement. The proper task of law is
to prevent these bodies to collide by building walls around
them. The liberty of an individual is the space within the
surrounding walls. The requirement of maximal liberty
means that the cells should be as large as possible. The requirement of equal liberty means that all the cells should be
of equal size.
The idea that politically relevant liberty is simply the absence of coercive restrictions has been more persistent than
the correlative idea that law is simply a set of such restrictions. Thus in his famous inaugural lecture42 Sir Isaiah
Berlin adopted the Benthamite notion of liberty and declared it a part of the liberal credo. Numerous commentators have accepted the basic claims put forth by Berlin or
implied by his essay: the negative concept of liberty is the
one which corresponds with our untutored intuitions or
with ordinary usage of the word, and it alone is compatible
with liberalism. Thus, those who have criticized Berlins
negative liberty for example, C. B. Macpherson,43
Charles Taylor44 or Philip Pettit45 have usually portrayed
themselves as critics of the liberal view. In this context, I
cannot assess these criticisms, but I just want to point out
that neither Berlin nor his critics have paid attention to the
close relationship between Hobbess and Benthams notions
of liberty and their notion of law. As far as I know, Berlin
did not accept the IC-theory of law, nor have his critics
claimed that he, as a consequent liberal, should have ac42 Berlin, Isaiah, Two Concepts of Liberty, in Berlin, Isaiah Four Essays on
Liberty, Oxford, Great Britain, Oxford University Press, 1969, pp. 118-172.
43 MacPherson, C. B., Berlins Division of Liberty, in MacPherson, C. B., Democratic Theory. Essays in retrival, Oxford, Great Britain, Oxford University Press,
1975, pp. 95-117.
44 Taylor, Charles, Whats Wrong with Negative Liberty, in Taylor, Philosophy
and the Human Sciences. Philosophical Papers 2, Cambridge, Great Britain, Cambridge University Press, 1985, pp. 211-229.
45 Pettit, Philip, Freedom with Honour: A Republican Ideal, Social Research 64,
1997, pp. 52-76.
259
EERIK LAGERSPETZ
cepted it. There is nothing specifically liberal in the IC-view;

indeed, many liberals of different persuasions, from Hayek
to Dworkin, have considered it a deeply illiberal view. Because the negative notion of liberty follows from the IC-theory rather than the other way round, it is possible to reject
the latter and nevertheless stick to the former. However, if
the negative notion of liberty is actually a sort of by-product
of one once widely accepted but rather implausible and
nowadays rejected theory of law, the connection between
liberalism and the negative notion is far less obvious than
both Berlin and his critics tend to think. I shall argue that
the negative concept of liberty is actually insufficient even
for liberal purposes. Indeed, without the IC-theory there
seems to be very few reasons to accept the negative concept
of liberty.
To make the matter more clear, consider a person who is
put under guardianship. He is unable to make valid contracts, to make will, to vote, or marry. This does not mean
that he is prevented from doing all these things by coercive
restraints, or that he is subjected to any specific duties. He
may well draft and sign texts entitled as My Last Will and
Testament or A Contract of Sales without being subjected
to any coercive measures. He may raise his hand when
votes are counted, he is allowed to hand a credit card to a
bank clerk, or to say I do at the presence of a minister. He
may be free to perform the same sets of physical movements as his fellow citizens. As a physical body, he is able
to move as freely as other similar bodies. The difference is
that his movements do not have their normal legally defined meaning. They do not count. The courts do not enforce the will or the contracts signed by him. When he
raises his hand in the meeting, it is not counted. The bank
clerk refuses to give him any money. And so on. It seems
plausible to say that the ward is less free than the rest of
us. He is less free because he does not possess certain powers which can be conferred to him only by legal rules. To
conceptualize the wards legal inability to do certain things
260
as unfreedom is, I think, fully in accordance with the normal use of the word freedom. Berlin and his followers
stress that inabilities do not limit our freedom. This is
prima facie a plausible claim. For example, if I am free to
walk on public pathway, my freedom is not diminished if I
during my walk break my leg and become unable to continue. But a legal inability is an inability of a different kind.
It, unlike my injury, is an intended result of the action (or
inaction) of the power-holders and can be removed just by
amending the rules.
If this argument is correct, important aspects of our politically relevant liberty the liberty to make contracts,
marry etc. is, contra Bentham, produced by law. Laws
which confer us powers to do these and similar things need
not to be abrogative to anyones liberty. Liberty that is,
the kind of liberty which should interest even liberals is
not a purely zero-sum notion, and it cannot be defined
merely as the silence of laws. Certainly, the freedom to
make contracts, or otherwise establish mutually beneficial
relations with other people is paradigmatically a liberal issue. If this freedom cannot be defined in terms of negative
liberty, there seems to be no reason to claim that it is the
liberal notion.
I am interested in this argument mainly because it is so
simple. Unlike most criticisms of the negative concept of
liberty, it does not presuppose any particular theory of human nature, or morality, or society. It accepts a background supposition made by Hobbes and Bentham: the
concept of (politically relevant) liberty is tied to the concept
of law. There is, for example, no need to postulate the
higher and lower selves, a postulate accepted by the Idealists and criticized by Berlin. Nevertheless, the argument
has interesting further implications. As Hart says in CL, the
role of law in our lives is far more complicated than that
suggested by the IC-view. Bentham though that the only
connection between liberty and law is that law can make
liberties more secure, by protecting us from the intrusions
261
EERIK LAGERSPETZ
of others. This is one link between liberty and law. But our
example reveals another connection. On the one hand, law
can make us freer by enabling us to do things which cannot
be done outside the legal context. It can also extend the total sum of our liberties by giving us new legal abilities. On
the other hand, law can also diminish our liberty without
constraining us by depriving some of our legally defined
powers. The topology of the (politically relevant) freedom is
much more complex than Hobbes and Bentham thought. If
our legally defined powers are a part of our liberty, it is not
clear what maximal liberty might mean. This might be considered as an argument against the inclusion of legally defined abilities. But it may also be considered as a limitation
of our ability to compare the amounts of freedom across
various legal systems.
The purpose of this brief exposition was to show that, after fifty years, H. L. A. Harts The Concept of Law is still
worth of reading, not only by those who are interested in
the philosophy of law in the narrow sense, but also, to
quote Hart once again, by all those whose chief interests
are in moral or political philosophy, or in sociology, rather
than in law.
V. BIBLIOGRAPHY
ARENDT, Hannah, The Crises of the Republic, New York, USA,
Harcourt, 1972.
AUSTIN, John, The Province of Jurisprudence Determined,
Cambridge, Great Britain, Cambridge University
Press, 1995.
BARNES, Barry, The Nature of Power, Oxford, Great Britain,
Polity Press, 1988.
BENTHAM, Jeremy, Selected Writings on Utilitarianism. Hertfordshire, Great Britain, Wordsworth Classics, 2001.
, Principles of the Civil Code. Part I Objects of the Civil
Law, in Bentham, Jeremy, Selected Writings on Utilita262
rianism, Hertsfordshire, Great Britain, Wordsworth

Classics, 2001.
, Anarchical Fallacies, in BENTHAM, Jeremy, Selected
Writings on Utilitarianism, Hertsfordshire, Great Britain, Wordsworth Classics, 2001.
BERLIN, Isaiah, Two Concepts of Liberty, in BERLIN, Isaiah
Four Essays on Liberty. Oxford, Great Britain, Oxford
BUCHANAN, James, The Limits of Liberty. Between Anarchy
and Leviathan, Chicago, USA, Chicago University
Press, 1975.
HARDIN, Russell, David Hume: Moral & Political Theorist.
Oxford, Great Britain, Oxford University Press, 2007.
HART, H. L. A., The Concept of Law, Oxford, Great Britain,
HOBBES, Thomas, Leviathan, Oxford, Great Britain, Oxford
HUME, David, Humes Moral and Political Philosophy, Darien,
USA, Hafner Publishing Co., 1970.
KAVKA, Gregory S., Hobbesian Moral and Political Philosophy,
Princeton, USA, Princeton University Press, 1986.
LAGERSPETZ, Eerik, The Opposite Mirrors. An Essay on the
Conventionalist Theory of Institutions, Dordrecht, The
Netherlands, Kluwer, 1995.
LONG, Douglas G., Bentham on Liberty, Toronto, Canada,
University of Toronto Press, 1977.
LUCAS, J. R., The Principles of Politics, Oxford, Great Britain,
Clarendon Press, 1966
MACPHERSON, C. B., Berlins Division of Liberty, in
MACPHERSON, C. B., Democratic Theory. Essays in retrival, Oxford, Great Britain, Oxford University Press,
1975.
263
EERIK LAGERSPETZ
OLIVECRONA, Karl, Law as Fact, London, Great Britain,

ORTEGA Y GASSET, Jos, La rebelin de las masas, Madrid,
Spain, Tecnos, 2003.
PARSONS, Talcott, The Modern Sociological Theory, New York,
USA, Free Press, 1967.
PETTIT, Philip, Freedom with Honour: A Republican Ideal, Social Research 64, 1997.
RAWLS, John, A Theory of Justice, Oxford, Great Britain,
Oxford University Press 1971.
ROSS, Alf, On Law and Justice, London, Great Britain, Stevens, 1958.
TAYLOR Charles, Whats Wrong with Negative Liberty, in
TAYLOR, Philosophy and the Human Sciences. Philosophical Papers 2, Cambridge, Great Britain, Cambridge University Press, 1985.
WALDRON, Jeremy, Law and Disagreement, Oxford, Great
WARNOCK, G. J., The Object of Morality London, Great Britain, Methuen & Co., 1972.
264
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
FAREWELL TO THE RULE OF RECOGNITION?

Giorgio PINO*
Resumen:
En este artculo se argumenta que la regla de reconocimiento, tal como
fuera concebida por Hart, es o bien un concepto redundante y en consecuencia intil o un concepto limitado en su poder explicativo. En
cualquier caso, se trata de un concepto cuyo alcance es, frente a los sistemas jurdicos contemporneos, mucho ms estrecho de lo que Hart
pudo imaginar. De igual modo se argumenta que la regla de reconocimiento, en alguna de sus posibles (y plausibles) reformulaciones, puede
tener a pesar de todo un papel significativo y no redundante, pero solamente si se emplea en un sentido radicalmente distinto al que propuso
Hart o que se propone en buena parte de la literatura positivista posthartiana.
Palabras clave:
Regla de reconocimiento, sistema jurdico, validez jurdica,
aplicabilidad, H. L. A. Hart.
Abstract:
I will argue that the rule of recognition, as it has been conceived by Hart, is
either a redundant, and hence mostly useless, concept, or a concept with
limited explanatory potential in either case, at best a concept whose
scope is, in contemporary legal systems, much narrower than Hart envis* Associate Professor of Legal Philosophy, University of Palermo giorgio.pino
@unipa.it, www.unipa.it/gpino.
265
GIORGIO PINO
aged. I will also argue that the rule of recognition, in one of its possible
(and plausible) reformulations, can nevertheless play a significant, non-redundant role, but only if employed in a rather different way than the one
proposed by Hart, as well as by much of post-Hartian positivist literature.
Keywords:
Rule of Recognition, Legal System, Legal Validity, Applicability,
H. L. A. Hart.
266
SUMMARY: I. Trouble in Paradise. II. Harts Account of the Rule

of Recognition. III. What kind of Rule is the Rule of
Recognition. IV. The Rule of Recognition Revisited.
V. Bibliography.
I. TROUBLE IN PARADISE
The concept of a rule of recognition the ultimate rule
establishing the criteria of validity of the other rules of a
given legal system is notoriously one the most
long-standing and far-reaching contributions of H. L. A.
Hart to legal theory.1 It has attracted countless sympathetic
attempts of refinement and development, as much as harsh
criticism. At the same time, while Hart envisaged for the
rule of recognition a paramount clarifying role for legal theory and jurisprudence,2 his own exposition of the doctrine
of the rule of recognition is famously affected by some ambiguities and obscurities.
1 See M. Kramer, Where Law and Morality Meet, 104 (the idea of a rule of recognition is [o]ne of Harts greatest contributions to the philosophy of law); A.
Marmor, Philosophy of Law, 50 (No other idea is more closely associated with
Harts theory of law than the idea that legality is constituted by social rules of recognition).
Jules Coleman probably overstates the point though when he says that commitment to the rule of recognition, along with the separability thesis, is a defining feature of legal positivism (see Authority and Reason, 287, 316 fn 5). On the one
hand, if the rule of recognition is understood in a strictly Hartian sense, then as a
matter of fact it is not true that all legal positivists are committed to such a thesis
(one hardly needs to mention Austin, Kelsen, and Ross to this effect). On the other
hand, if the rule of recognition is understood in a very broad and very weak sense,
as a criterion whatsoever to delimitate the law, then not only legal positivism but
virtually every kind of legal theory uses or presupposes a rule of recognition of
some kind (some criterion to distinguish the law, as the subject matter of legal theory, from other phenomena such as religion, sheer violence, the game of chess,
etc.). Accordingly, Scott Shapiro rightly notes, for example, that in a sufficiently
weak sense also Law-as-Integrity amounts to a rule of recognition: see What Is
the Rule of Recognition (and Does It Exists)?, 267.
2 According to Hart, the existence of a socially practised rule of recognition
provides the foundations of a legal system (The Concept of Law, 100). Moreover,
Hart repeatedly claims, paraphrasing Austin, that the union of primary and secondary rules (among which there is the rule of recognition) is the key to the science
of jurisprudence: see The Concept of Law, 81; Positivism and the Separation of
Law and Morals, 59.
267
GIORGIO PINO
In this essay I will argue that the rule of recognition, as it

has been conceived by Hart and by much of post-Hartian
legal theory, is either a redundant, and hence mostly useless, concept, or a concept with limited explanatory potential in either case, at best a concept whose scope is, in
contemporary legal systems, much narrower than Hart envisaged. I will also argue that the rule of recognition, in one
of its possible (and plausible) reformulations, can nevertheless play a significant, non-redundant role, but only if employed in a rather different way than the one proposed by
Hart (as well as by much of post-Hartian positivist literature). So, at the end of the day the answer to the question
posed by the title of this essay will be a qualified No it is
not necessary to give up entirely the concept of rule of recognition, but only insofar as we are willing to mould the
concept of rule of recognition in a different way, assigning it
a slightly different job than the one that seems to appear
from The Concept of Law, as well as from much of
post-Hartian literature. (I believe nevertheless that such a
proposal turns out to be not only compatible, but also
deeply coherent with the overall positivist project
underwritten by Hart).
The argument proceeds as follows. I will first ( 2) summarize Harts main arguments on the rule of recognition,
also pointing at some possible developments to which
Harts original statement of the relevant concept is reasonably amenable. Then ( 3) I will move on to consider the
main interpretations, or reformulations, that the concept of
rule of recognition has undergone in post-Hartian legal theory, briefly testing them in light of their theoretical pros
and cons. Finally ( 4), I will try to offer an alternative reconstruction of the concept of rule of recognition, trying to
show that this concept can play an important, non redundant role, provided we accept that it is not the role
envisaged by Hart.
A few caveats are in place here. First, my argument will
be at first exegetical in character (I will try to be as faithful
268
as possible to Harts rendering of the rule of recognition),

but my main aim in this paper is not just to provide a
trustful interpretation of Harts thought; rather, what I aim
to is to carve out a useful and significant concept in its own
right I want to explore the possibility that the rule of recognition be rescued from the charge of redundancy (a
charge that I am by no means not alone in levelling).3 Second, this essay will be firmly grounded in the legal positivist
theoretical field, or so its author hopes. In other words, my
theoretical objections to the rule of recognition will not be
the upshot of a Dworkinian, interpretivist, theoretical
stance,4 which it is not my concern neither to endorse nor
to critically evaluate here. Suffice it to say that I am here
assuming the general soundness and defensibility of a positivist, Hartian theoretical stance, and I am exploring the
possibility of improving it further.
II. HARTS ACCOUNT OF THE RULE OF RECOGNITION
Briefly, Harts own account of the rule of recognition runs
as follows.5
Every modern, developed legal system includes a peculiar
secondary rule the rule of recognition whose function
is to provide authoritative criteria according to which primary rules of obligation can qualify as valid. The rule of
recognition can perform such a task in many possible ways
typically, by pointing to one or more features possession
of which makes of a rule a rule of law, a valid legal rule.
See, among others, the authors quotes infra, fn 59.
For Dworkins multifaceted criticism of the rule of recognition (that are part
and parcel of his broader rejection of legal positivism, both in Hartian guise and
otherwise), see The Model of Rules I and Laws Empire. On interpretivist theories
of law see N. Stavropulos, Interpretivist Theories of Law.
5 The main source is obviously The Concept of Law, 91-123. Hart provides
other insights on the rule of recognition (even if he does not always refer to it exactly in this way), at least in Positivism and the Separation of Law and Morals;
Lon L. Fuller: The Morality of Law; Kelsens Doctrine of the Unity of Law; Postscript. A good synthesis (but limited to The Concept of Law and the Postscript) is
in J. Dickson, Is the Rule of Recognition Really a Conventional Rule?.
3
4
269
GIORGIO PINO
The rule identified following the criteria stated by the rule

of recognition can then be said to be part of the legal system indeed, the idea itself of a legal system requires the
existence of a unifying rule of recognition.6 The criteria of
validity encapsulated in the rule of recognition can take a
variety of forms: reference to a certain text, or to a procedure (legislative enactment) are the most common hypotheses.7 If in the relevant legal system law-making power is itself regulated by (secondary) rules rules of change then
the rule of recognition of that system will necessarily include reference to the operation of such rules.8 When there
are multiple sources of the law, the rule of recognition normally provides also some criteria for their hierarchical ordering.9 As a (ultimate) criterion of validity, the rule of recognition is not itself valid; indeed, it is neither valid nor
invalid, since there is no further criterion that can be used
to assess its validity.10 Neither the rule of recognition is explicitly stated, at least for the most part:11 its existence
has to be inferred from the actual law-ascertaining practices of the officials; in other words, its existence is a matter of fact.12 A rule of recognition exists if there is a convergent practice among officials of treating as valid law what is
validated by the rule of recognition (the relevant practice
qualifying as an acceptance from the internal point of
view).13
Harts account of the rule of recognition has spurred a
huge and manifold debate. The main theoretical enquiries
prompted by Harts analysis of the rule of recognition have
largely focused i) on the existence conditions of the rule of
recognition; ii) on the possible content of the rule of recog6
7
8
9
10
11
12
13
The Concept of Law, 95, 116, 148.

Ibidem, 95, 100, 101
Ibidem, 96.
Ibidem, 95, 101, 106.
Ibidem, 109.
Ibidem, 101, 102.
Ibidem, 110; see also 108, 116.
Ibidem, 102, 116, 117.
270
nition, i.e. the problem of logical or conceptual constraints

on the criteria of validity that can possibly be enshrined in
the rule of recognition; iii) on the function of the rule of
recognition.
Legal theorists arguing problem i) are engaged in trying to
figure out what kind of state of affairs must be obtained in
order to bring a rule of recognition into existence, elaborating on Harts scattered remarks on the convergent behaviour of officials sharing an internal point of view: and much
of the discussion has revolved around the theoretical viability of understanding the relevant practice as a convention
an idea that Hart himself eventually endorsed, even if in
somewhat mild terms.14 Debate on ii) aims at establishing if
there are conceptual limits to what can be included by a
rule of recognition among the criteria of legality: those who
deny that there are such conceptual constraints, and especially that such constraints do not rule out morality from
acting as a criterion of legal validity, have come to be
known as inclusive or soft positivists, or incorporationists. Those who argue that the rule of recognition cannot contemplate reference to morality as a criterion of legal
validity, for conceptual reasons related to the concept of authority and to the function of guidance of conduct that is
said to be typically performed by the law, are said to adhere
to exclusive or hard positivism.15 A third strand, lastly,
denies that the rule of recognition should make reference to
contested moral principles, and this is asserted on ethical,
14 See Postscript, 255, 267. Conventionalist accounts of the rule of recognition
are elaborated by G. Postema, Coordination and Convention at the Foundations of
Law; E. Lagerspetz, The Opposite Mirrors; J. Coleman, The Practice of Principle, esp.
ch 7; A. Marmor, Positive Law and Objective Values, chs 1 and 2; S. Shapiro, Law,
Plans, and Practical Reason. For a critical assessment, see L. Green, The Authority
of the State, ch 4, and Positivism and Conventionalism; B. Celano, La regola di
riconoscimento una convenzione?; J. Dickson, Is the Rule of Recognition Really
a Conventional Rule?; A. Schiavello, Perch obbedire al diritto?
15 On the inclusivist side, see e.g. W. Waluchow, Inclusive Legal Positivism; J.
Coleman, The Practice of Principle; M. Kramer, Where Law and Morality Meet. On the
exclusivist side, see e.g. J. Raz, Legal Positivism and the Sources of Law and Authority, Law, and Morality; S. Shapiro, On Harts Way Out.
271
GIORGIO PINO
not conceptual grounds this has come to be known as

normative positivism.16
In the following of this essay, I will not touch upon matters under i) and ii). By contrast, in this essay I will deal
with iii) a comparatively less explored issue, that nevertheless has also attracted a fair amount of theoretical interest in its own right. So in the following I will try to make
sense of the issue of the function of the rule of recognition:
what is it there for, what kind of (theoretical, conceptual, or
otherwise) role does it perform in a legal system, and also in
a legal-theoretical account of a legal system?
Before turning to this, however, I will point to some possible refinements that Harts account of the rule of recognition seems amenable to, and that will be somewhat useful
also for the rest of the present discussion.
First, while Hart repeatedly asserts that the (secondary)
rule of recognition provides the test of validity for primary
rules of obligation, it is common ground in positivist,
Hartian legal theory that the rule of recognition validates all
the other rules of the system, primary and secondary alike.17
Second, Hart usually talks of one rule of recognition, but
occasionally also of rules of recognition;18 so how many
rules of recognition are to be found in a legal system? The
answer seems to be that while in each legal system there is
just one (ultimate) rule of recognition, this is bound to be
very complex and hierarchically structured, and it is also
possible that different kind of officials in the legal system
follow different or partly different rules of recognition due to
their different institutional roles.19 Lastly, it is also possible
that the officials of the legal system do not follow exactly the
same rule of recognition, not because of their different in16 See N. MacCormick, A Moralistic Case for A-Moralistic Law; J. Waldron,
Normative (or Ethical) Positivism.
17 See e.g. J. Raz, The Identity of Legal Systems, 91.
18 See The Concept of Law, 102, 104, where Hart refers to rules (plural) of recognition.
19 J. Raz, The Concept of a Legal System, 200; P.M.S. Hacker, Harts Philosophy
of Law, 24.
272
stitutional positions, but because they disagree in identifying the relevant rule of recognition. While radical disagreement between officials as to the rule of recognition
endangers the unity and the working itself of the legal
system, with consequences that can range from extreme
unpredictability in judicial decisions to revolution and civil
war, a degree of marginal disagreement, or disagreement at
lower levels in the chain of validity, can be easily accommodated in the conceptual framework of the rule of recognition; this is so because as Hart explicitly admits the
rule of recognition partakes of the open texture that affects
every rule and so it cannot be expected to provide a determinate answer to all cases.20 What degree of indeterminacy
is tolerable in the rule of recognition, or to put it differently,
what degree of overlap between the different rules of
recognition is needed, is a question that is not possible to
answer in abstract terms.
Third, does the rule of recognition compound all the criteria of validity for one legal system (regulates the validity of
all the kind of legal norms that belong to the system), or is
it an ultimate test of validity? Hart actually defines the rule
of recognition exactly as the ultimate criterion of validity,
meaning that the rule of recognition is not in turn validated
by other rules.21 Moreover, the rule of recognition needs not
specify the criteria of validity of all the norms of the system:
the criteria of validity for certain kind of norms (call them
N2) can be established not by the rule of recognition but by
some other norm N1, whose validity is established by the
rule of recognition (and so N2 is directly validated by N1,
and indirectly validated by the rule of recognition): so there
can be derivative criteria of validity, as it were, along with
the ultimate ones provided by the rule of recognition.22
20 The Concept of Law, 147-154. The issues discussed in this paragraph are
further explored by M. Kramer, Where Law and Morality Meet, 105-110.
21 The Concept of Law, 105-106.
22 J. Raz, The Identity of Legal Systems, 95; K. Greenawalt, The Rule of Recognition and the Constitution, 5-6; M. Kramer, Where Law and Morality Meet, 110; L.
273
GIORGIO PINO
Fourth, in The Concept of Law Hart seems to be committed to the view that the rule of recognition instantiates pedigree only criteria of validity, meaning that validity is to be
assessed only by reference to formal, factual criteria.23 In
this way, the rule of recognition can establish the validity of
legal sources only, and not the validity of legal norms as
well a norm being the meaning, the content of a legal
source: norms are what the sources say, or are taken to say
by means of interpretation. A possible explanation for this
limitation on the criteria of validity in the rule of recognition is that Harts original account was probably influenced
by his semi-formalist theory of legal interpretation,24 according to which sources of law are capable of carrying a
core of settled, undisputable meaning, along with an area of
penumbra where there is some uncertainty and the interpreter is called upon to use some degree of interpretive discretion. Be that as it may, the source-based account of validity has notoriously been the main ground for Dworkins
original attack25 on Hartian positivism, for its alleged incapability of explaining the use of legal principles in adjudication. Responding to Dworkin, in the Postscript to the second edition of The Concept of Law, Hart eventually admitted
that reference to source-based only criteria of validity was
just a matter of emphasis: conceptually, his model is intended to accommodate also material, substantive criteria
of validity as opposed to only formal ones.26
Alexander and F. Schauer, Rules of Recognition, Constitutional Controversies,
and the Dizzying Dependence of Law on Acceptance, 177.
23 This emerges from the repeated references Hart makes to enactment and
similar concepts as the main test of validity; see The Concept of Law, 94, 95 (fact of
their having been enacted), 96, 100, 101, 148, 209 (valid by the formal tests).
Probably the only counter-example is at 204 (in some systems, as in the United
States, the ultimate criteria of validity explicitly incorporate principles of justice or
substantive moral values).
24 This is suggested by E. Diciotti, Regola di riconoscimento e concezione
retorica del diritto, 13.
25 See R. Dworkin, The Model of Rules I.
26 See Postscript, 250, 264-266. Hart anticipated this point also in Lon L.
Fuller: The Morality of Law, 361.
274
III. WHAT KIND OF RULE IS THE RULE OF RECOGNITION?

I assume that the picture drawn in last section (Harts
own account of the rule of recognition, plus a few minor
amendments, or clarifications) compounds the standard
view on the rule of recognition in contemporary positivist
literature.27
Given this background, I now turn to the central problem
for this essay, that of the nature and function of the rule of
recognition. Here there is considerably less agreement between jurisprudents, and on close inspection Harts account of the rule of recognition reveals many puzzling features. (As I have made it clear earlier, I am here broadly
endorsing the Hartian project, and so I am assuming that
something like the rule of recognition exists, or better that
something like the rule of recognition is a good conceptual
tool in order to attain a satisfying understanding of law and
legal systems. So, the puzzles I have in mind are internal,
so to say, to the Hartian account of rule of recognition. It is
of course possible to adopt an external stance towards
Harts jurisprudence, and hence to deny that anything like
the rule of recognition can help us in the way of a sensible
understanding of the law; I am not exploring here this line
of argument).
First and foremost, it is not entirely clear what kind of
rule the rule of recognition is meant to be. Hart famously
distinguishes between primary and secondary rules, and he
plainly says that the rule of recognition is a secondary rule.
The problem is that 1) the distinction itself between primary
and secondary rules is far from clear, and moreover 2) the
rule of recognition in Harts characterization does not seem
to square well with either kind of rules.
1) As far as the distinction between primary and secondary rules is concerned, many critics have remarked that
27 To be sure, exclusive legal positivist are likely to deny the last point made in
the previous section, i.e. the possibility of substantive, as opposed to formal and
factual, criteria of legal validity. But I dont think they would deny that Hart ended
with admitting such a possibility.
275
GIORGIO PINO
Hart uses the distinction in a number of different, and

sometimes incompatible, ways.28 For instance, Hart on the
one hand says that primary rules are duty-imposing rules,
rules of obligation, while secondary rules are power-conferring rules, they provide facilities (which would make the relevant distinction deontic in character);29 on the other hand
Hart also says that primary rules are rules aimed at guiding conduct, while secondary rules are rules about (other)
rules (which would make the relevant distinction one of
content).30 Such two distinctions are not equivalent, because a rule can of course impose a duty to apply other
rules: as such, it would be a primary rule on one interpretation (because it is duty-imposing), and a secondary rule
on another (because it is a rule about other rules).
2) Hart explicitly defines the rule of recognition as a secondary rule. As such, it should be regarded as a power-conferring
rule. At the same time, however, Hart says that the rule of recognition states the criteria according to which validity of other
rules is to be assessed. As such, the rule of recognition looks
like a test, or a list of features possession of which grants validity. Now, the obvious problem is that a list or a test do not
confer any power whatsoever. It is like a definition, and a definition is hardly a norm, let alone a power- conferring norm.
Against this problematic background, lets take a quick
look at the more recurrent ways of conceiving of the rule of
recognition in post-Hartian legal theory. (To be sure, for
each of the following instances it is not always clear if it is
best understood as an interpretation of Harts thought, or
rather as a reformulation of the concept itself of the rule of
recognition, leaving aside the question if Hart would have
endorsed such reformulation.) This will enable us to grasp
at least what the rule of recognition plausibly cannot be.
28 L. Cohen, Critical Notice; N. Bobbio, Norme primarie e norme secondarie;
J. Raz, The Functions of Law; N. MacCormick, H. L. A. Hart, 106; W. Waluchow, Inclusive Legal Positivism, 75; L. Green, The Concept of Law Revisited, 1699.
29 The Concept of Law, 81, 96, 97.
30 Ibidem, 81, 94.
276
(a) The rule of recognition as a duty-imposing rule.31 According to this interpretation, against Harts many explicit
statements that the rule of recognition is a secondary rule
(and so either a power-conferring rule, or a rule about other
rules, or both), the rule of recognition is in fact a primary,
duty-imposing rule: it imposes upon officials a duty to apply (valid) law; as a consequence, the rule of recognition imposes a duty whose content is the exercise of a power (judicial power). This interpretation is substantiated in a
number of ways. For instance, it is said that since Hart acknowledges only two kinds of rules (duty-imposing and
power-conferring), and since the rule of recognition is not
power-conferring, then it must be a duty-imposing rule.32
Another argument is that Hart explicitly provides an analysis of social rules as sources of obligations (the so-called
practice theory of rules)33 while he does not provide an
equally accurate analysis of the acceptance/existence of
power-conferring rules; and since the rule of recognition exists only as a social rule, then it must be a duty-imposing
rule.34 Moreover, the duty-imposing character of the rule of
recognition is regarded as the key (or at lest the first step)
towards explaining laws normativity.35
31 P. M. S. Hacker, Harts Philosophy of Law, 24; N. MacCormick, H. L. A. Hart,

ch 9; L. Green, The Authority of the State, 118; J. Coleman, The Practice of Principle,
84 ff; S. Shapiro, What Is the Rule of Recognition (and Does It Exists)?, 240, and
Legality, 85.
32 See S. Shapiro, What Is the Rule of Recognition (and Does It Exists)?, 239.
Such an argument a contrario is somewhat weak, because Hart does not intend the
dichotomy duty-imposing/power-conferring rule to be exhaustive (The Concept of
Law, 32; Lon L. Fuller: The Morality of Law, 358). But it is true nonetheless that it
would appear strange if Hart would have left just the rule of recognition in the terra
incognita of the unexplored kind of rules different from the duty-imposing and
power-conferring ones.
33 This label is due to J. Raz, Practical Reason and Norms, 49-58. Hart later endorsed this definition: see Postscript, 254-259.
34 J. Raz, The Identity of Legal Systems, 92-93; S. Perry, Where Have All the
Powers Gone? Hartian Rules of Recognition, Noncognitivism, and the Constitutional and Jurisprudential Foundations of Law, 305-308.
35 For a detailed analysis of this, see J. Coleman, The Practice of Principle, chs.
6-10.
277
GIORGIO PINO
This approach is tempting, and indeed in the next section

I will defend a version of it. But as it stands it suffers from
some weaknesses. For instance, it does not explain why
sometimes judges have a duty to apply non-valid law, or
conversely even a duty not to apply valid law (I will return
on this point shortly). Moreover, under this approach, if
taken at face value, the rule of recognition imposes a duty
to apply valid law, but does not point to any direction to ascertain what counts as valid law. So either the legal system
provides elsewhere some independent criteria of validity,36
or the rule of recognition must be supplemented with
in-built, as it were, criteria of validity. To this last possibility I now turn.37
(b) The rule of recognition as a duty-imposing rule and a list
of criteria of validity.38 Under this reading, the rule of recognition both includes a list of features possession of which
renders a norm a valid legal norm, and states a duty for law
applying officials to apply valid rules in complying with
their institutional roles.
While this interpretation escapes some of the difficulties
affecting interpretation (a), it is still the case that sometimes judges have a (legal) duty to apply non-valid law, as
well as they can have a (legal) duty not to apply valid law.39
This is so, because judges may quite often have a legal duty
to apply norms that are not valid in any plausible meaning:
This seems to be MacCormicks position: see H. L. A. Hart, 115.
A third possibility is utterly rule-skeptic, or legal-realistic: valid law is whatever judges feel they are compelled to apply. On this reading, the rule of recognition would not properly exist, or either it would be indistinguishable from rules of
adjudication (once judges are in place, whatever they do is law).
38 J. Raz, The Identity of Legal Systems, 93; M. Atienza, J. Ruiz Manero, A Theory of Legal Sentences, ch 5; S. Perry, Where Have All the Powers Gone? Hartian
Rules of Recognition, Noncognitivism, and the Constitutional and Jurisprudential
Foundations of Law, 305-306.
39 W. Waluchow, Inclusive Legal Positivism, 65-66, 77 (not all the judges must
(legally) apply is law validated by a rule of recognition; and [] not all laws validate
by a rule of recognition must be applied by judges); for some examples, see J. J.
Moreso, Legal Indeterminacy and Constitutional Interpretation, 108 ff; G. Pino,
Norme e gerarchie normative, 279-288, 294-299, and La aplicabilidad de la
norms jurdicas.
36
37
278
foreign laws, moral norms or considerations,40 invalid laws

(since a lower court may well have the legal duty to apply
unconstitutional norms whose invalidity has not been ruled
by a Constitutional Court)41 or even repealed statutes.
Sometimes, law-applying institutions do not have exactly
an explicitly stated legal duty to this effect, but still their
resorting to extra-legal or extra-systemic standards is considered legally appropriate and acceptable.42 As a consequence, it remains to be explained how is it possible that
the same rule states criteria of validity, imposes a duty to
apply valid law, and authorizes departure from valid law at
the same time.
Moreover, conflating validity and duty to apply (= bindingness) takes us right to a Kelsenian conception of validity
as binding force, a conception whose compatibility with a
coherent positivist approach is questionable to say the least
(since it is hard to reconcile the positivist thesis that law is
a fact, with the claim that this fact has binding force),43
and that Hart for one surely did not share.44
(c) The rule of recognition as a closure rule.45 One possible
additional function of the rule of recognition, often asserted
40 I am emphatically not claiming that a moral norm cannot at the same time be
also a legal norm: rather, I am here referring to a different phenomenon, when the
law does not refer to a specific moral principle, but rather explicitly defers to the
judge the use of a moral assessment of some aspects of the case at hand. Here I
have in mind what in many Continental, civil law systems is known as giudizio di
equit.
41 For a subtle but interesting discussion of this problem, see the exchange between K.E. Himma, Making Sense of Constitutional Disagreement: Legal Positivism, the Bill of Rights, and the Conventional Rule of Recognition in the United
States, and M. Kramer, Where Law and Morality Meet, 115-140; see also W.
Waluchow, Four Concepts of Validity.
42 This is the case of the recent use of foreign authorities by the US Supreme
Court; see Roper v. Simmons 125 S. Ct. 1183 (2005); Lawrence v. Texas, 123 S. Ct.
2472, 2483 (2003).
43 For a classic discussion of this, see A. Ross, Validity and the Conflict between Legal Positivism and Natural Law.
44 See J. Raz, The Purity of the Pure Theory, 311 (validity for him [viz., for
Hart] indicates just membership in a system established in a certain way. It has little to do with binding normative force).
45 J. Raz, The Identity of Legal Systems, 93; N. MacCormick, H. L. A. Hart, 21, 110.
279
GIORGIO PINO
along with other functions such as the one of imposing on

officials a duty to apply valid law discussed in (a) and (b)
above, is that of a closure rule, meaning that it requires officials to apply only the law it validates. This is somewhat
justified also by some statements by Hart, who links the
rule of recognition to the idea of a legal system46, i.e. the
idea of a closed, discrete set of norms.
This interpretation has the defect of turning a contingent
feature of legal systems into a necessary truth, by encapsulating it into the definition of rule of recognition. Indeed it is
entirely contingent, I submit, that a legal system is closed
in the relevant sense, i.e. in the sense that its law-applying
institutions are forbidden to rely on every non-legal or extra-systemic consideration in discharging their functions. A
legal system that happens to allow such a possibility does
not shed its quality of being a system. Moreover, as a matter of observable fact, many contemporary legal systems
routinely allow judges to resort also to extra-legal or extra-systemic considerations.47
So this cannot be considered as a necessary, conceptual
defining feature of the rule of recognition.
(d) The rule of recognition as a power-conferring rule.48
This interpretation seems to be, superficially at least, the
closest to Harts own wording. Indeed, Hart qualifies the
rule of recognition as a secondary rule, and qualifies secondary rules as (inter alia) power-conferring rules the conclusion that the rule of recognition is a power-conferring
rule seems then straightforward.
The important point then is to understand what exactly
is the content of the legal power conferred by the rule of
recognition. Recall that, under any plausible definition, a
legal power is the power to change someones rights, duties,
and generally someones legal status (either by creating, or
The Concept of Law, 95.
See supra, fns 39-42 and accompanying text.
48 L. L. Fuller, The Morality of Law, 137; R. Dworkin, Laws Empire, 34 (the rule
of recognition assigns to particular people or groups the authority to make law).
46
47
280
by modifying, or by erasing some legal norm). So the exercise of a legal power cannot consist only in an activity of
double-checking a list of criteria in order to ascertain, in a
purely speculative way, as it were, the validity of a putatively legal norm. The exercise of a legal power must necessarily yield some legal effect. So, if we want to qualify the
rule of recognition as a power-conferring rule, we are bound
to point at the legal effects that are engendered by the use
of the rule of recognition.
In short, then, exactly what kind of power would be
granted by the rule of recognition? Is it a power to identify
valid law?49 It is hard to understand that such a thing can
be the object of a legal power, unless one interprets it as a
power to constitutively declare what the law is (and so, the
law is what courts say it is). This idea has found independent support by some legal theorists,50 and it also bears at
least a grain of truth, but still raises several problems. For
one thing, it conflicts with the widespread intuition that
courts are law-applying, and not law-creating institution.
Moreover, Hart distinguishes between finality and infallibility of judicial decisions, arguing that, unless one is playing
the game of the scorers discretion, an authoritative ruling,
even a definitive one, might still be legally wrong: it can be
final, but it does not change the rules it purports to apply.51
In short, if the law is constitutively determined erga omnes
by judicial decisions (and this would be so because the rule
of recognition allegedly confers the legal power to declare
what the law is), this would be tantamount to saying that
the rule of recognition is a rule that authorizes the making
of new law, and therefore it would rather be a rule of
49 Sometimes Hart seems to express himself in this way. See The Concept of
Law, 95 (the rule of recognition is a rule for conclusive identification of the primary rules of obligation), and 97 (on the relation, and the partial overlap, between
rule of recognition and rules of adjudication).
50 I am referring here to Kelsens doctrine of the constitutive character of judicial decisions. See General Theory of Law and State, 135.
51 The Concept of Law, 141-147. See also J. Raz, Practical Reason and Norms,
137-141 (on systems of absolute discretion).
281
GIORGIO PINO
change. On the other hand, if such constitutive power of

the courts is limited to the case at hand, then it seems that
we are more properly dealing with a rule of adjudication
rather than with a rule of recognition.
On this reading, then, the rule of recognition is either a
rule of change directed to the courts (because it empowers
the courts to declare the law erga omnes), and so valid law
is, a posteriori, what the courts say; or it is a rule of adjudication properly understood, because it empowers the courts
to state what the law is with final authority for the case at
hand.
A fortiori, if the relevant power is bestowed on law-making
institutions properly understood, as opposed to law-applying institutions, then the rule of recognition becomes indistinguishable from the rules of change.
(e) The rule of recognition as a duty-imposing and powerconferring rule.52 Under this reading, the rule of recognition
is both duty-imposing and power-conferring. It is so because the rule of recognition imposes on officials obligations to recognize certain sources of law as binding, and it
bestows powers on them to engage in authoritative acts of
law-identification that can fulfil those obligations.53
While I have no (or few) quarrel with the duty-imposing
side of this interpretation of the rule of recognition, assigning it also a power-conferring role is problematic. Indeed,
also in the light of the discussion sub (d), it is not difficult
to see that this interpretation includes in the rule of recognition functions that, on Harts account, are typically performed by rules of adjudication. Hence, on this reading one
of the two conceptual devices elaborated by Hart, the rule
of recognition and the rules of adjudication, becomes
redundant.
52
53
M. Kramer, Where Law and Morality Meet, 104-105.

Ibidem, 104.
282
(f) The rule of recognition as a list of criteria of validity.54

Hart repeatedly refers to the rule of recognition as a test, a
list of criteria complying with which a norm can be said to
be valid;55 as such, the rule of recognition would not
amount to a mandatory norm, it would not impose nor require any conduct whatsoever it is rather a conceptual
rule, a definition: it provides the definition of valid law according to a certain legal system.
This is surely an elegant interpretation, and one that is
strictly positivistic at that, since it does not endow neither
the rule of recognition nor any valid law with the quasipositivistic quality of bindingness; that there is a valid law,
that there is a socially practised rule of recognition, are not
per se reasons to do what the law requires or to play along
the legal game with those who for whatever reason already practice it.56 Still, the interpretation of the rule of recognition as a systemic test of validity suffers from some fatal flaws. In short, where it is the case, as it is in the
normality of developed legal systems, that law-making powers are regulated by (secondary) rules of change, the definition of the various conditions of validity for the norms of
the system turns out to be incorporated in those very rules
of change: a norm is valid if the relevant procedure has
been duly followed by the duly competent institution (and
the same goes also for the making of valid private
arrangements, such as contracts, wills, and so on).
This point seems to surface at times also in Harts own
account of the rule of recognition. For one thing, Hart explicitly says that there is a very close connection between
54 L. Cohen, Critical Notice, 408; E. Bulygin, Sobre la regla de reconocimiento; D. Lyons, Ethics and the Rule of Law, 53; R. Caracciolo, El sistema
jurdico. Problemas actuales, 44-54; B. Leiter, Legal Realism and Legal Positivism
Reconsidered, 66-68; A. Marmor, Positive Law and Objective Values, 1, 22, 32-33;
E. Diciotti, Regola di riconoscimento e concezione retorica del diritto; R. Guastini,
The Basic Norm Revisited.
55 The Concept of Law, 103, 109, 110.
56 A. Marmor, Positive Law and Objective Values, 22 (the existence of a social
practice, in itself, does not provide anyone with an obligation to engage in the practice. The rules of recognition only define what the practice is).
283
GIORGIO PINO
rule of recognition and rules of change, since if in the relevant legal system law-making power is itself regulated by
rules of change, then the rule of recognition of that system
will necessarily include reference to the operation of such
rules.57 But quite interestingly, at the various junctures
where Hart discusses of law-making powers, powers regulated by rules of change, he consistently relates the exercise
of those powers to the concept of validity (i.e. the purported
subject matter of the rule of recognition).58 According to
Hart, in the end, a valid rule is nothing more and nothing
less than a rule produced in accordance with the relevant
rules of change.
As a consequence, a rule of recognition conceived as a
list of criteria of validity does not add anything to what can
already be achieved using (even only in a cognitive, speculative way, as it were) the relevant rules of change of the
system.59 The rule of recognition becomes redundant it is
a needless reduplication60 of the rules of change.
At best, such a rule of recognition can provide the jurist
with something like a shortcut formulation, a synecdoche of
whatever criteria of validity are already in place through the
operation of the rules of change.
(g) The rule of recognition as a validating rule.61 On this
reading, the function of the rule of recognition is to answer
the question what is the reason for the validity of the highest rule of change of the system?.
57 The Concept of Law, 96. See also J. Raz, The Identity of Legal Systems, 95
(all the laws conferring legislative powers [] determine criteria of validity).
58 The Concept of Law, 31, 68-70, 72, 106, 148.
59 See N. Bobbio, Norme secondarie; N. MacCormick, H. L. A. Hart, 114-115; J.
Waldron, Who Needs Rules of Recognition?; S. Perry, Where Have All the Powers
Gone? Hartian Rules of Recognition, Noncognitivism, and the Constitutional and
Jurisprudential Foundations of Law, 307-308; A. Marmor, Philosophy of Law, 49
fn 20 (conceding that a criteria-of-validity definition of the rule of recognition conflates it into a rule of change).
60 Indeed, so Hart epitomized Kelsens Grundnorm: see The Concept of Law,
293.
61 J. Finnis, On Harts Ways: Law as Reason and as Fact, 44.
284
The positive contribution of this way of conceiving of the

rule of recognition is that it allows to qualify the constitution as valid law, which seems coherent with the common
parlance of jurists; moreover, Hart himself sometimes talks
as if the rule of recognition lies somewhat behind the constitution. But in the end this interpretation is not convincing. If validation is taken to mean that the rule of recognition imposes a duty to apply valid law62 (in a somewhat
Kelsenian way)63, then this way of conceiving it is not distinguishable from stance (a) above. And if validation is
taken to mean providing a criterion of validity, then it is
the same as (f). All the more, this position adds either to (a)
or to (f) (or to both, as the case may be) a superfluous additional level in fact, again a needless reduplication. Indeed, if judges already use the constitution to identify valid
law, or if judges already assume the constitution as binding
(in short, if judges already accept the constitution, either as
a conceptual rule or as a duty-imposing rule), then why on
Earth should one suppose the putative existence and acceptance of another rule that directs to the constitution?64
Plainly, it is the constitution that does the job, provided it
is accepted and used by the officials in the relevant way.
More precisely, (part of) the rule of recognition coincides
with those parts of the constitution that are effectively accepted and used by officials in order to identify the law that
is susceptible of application. But this does not make the
constitution per se the rule of recognition:65 it is so only
62 The idea of an implicit rule imposing officials a duty to apply the law seems to
be endorsed for instance by S. Shapiro, Legality, 86 (even thought not explicitly
mentioned in the text, it is part of federal constitutional law that judges are under
such a duty [viz., a duty to apply the Constitution] because officials accept this
mandate from the internal point of view).
63 J. Waldron calls this the Grundnorm function: see Who Needs Rules of Recognition?, 346-348.
64 See S. Munzer, Legal Validity, 66 and ch III generally; R. Guastini,
Conoscenza senza accettazione.
65 Further arguments against the identification between the rule of recognition
and the constitution are expounded by Raz, On the Authority and Interpretation of
Constitutions: Some Preliminaries, 333-334.
285
GIORGIO PINO
insofar as it is accepted and used by the relevant officials to

ascertain applicable law.
This point is explained in a very convincing way by Matthew Kramer, who highlights the difference between the
foundational and the epiphenomenal levels of existence of
the rule of recognition.66 The foundational level is the fact
that some law-identifying criteria is actually used and accepted in the relevant way. The epiphenomenal level instead obtains when the (foundational) rule of recognition is
also explicitly and authoritatively formulated in some text,
constitutional or otherwise. So a constitutional provision
becomes (part of) the rule of recognition only if it reflects
the actual practice of the officials i.e., only if the relevant
officials actually use it in their law-making and law-ascertaining activities. Moreover, while conceptually distinct, the
foundational level (i.e. the actual practice of recognition)
and the epiphenomenal level (i.e. the codification of that
practice) are bound to interact deeply. On the one hand the
epiphenomenal level will reflect, albeit not necessarily perfectly, the foundational level: would the codified criteria be
too remote from the prevailing official practice of recognition, they probably become dead letter soon. On the other
hand, codified criteria of recognition will in turn constrain
to some extent the subsequent practices of recognition:
official law-making and law-ascertaining practice will easily
rely on them as the default option.
An example may help clarify this last point. In many civil
law systems there is some explicit provision, sometimes
even at the constitutional level, requiring judges to ground
their decisions only on statutes passed by the Parliament.67
Still, as a matter of observable fact, it is often the case that
in such legal systems judges regularly resort, in their
law-ascertaining activities, to precedents and decisions
from other courts, even in the face of an explicit prohibition
See M. Kramer, Where Law and Morality Meet, 110-114.
For instance, the Italian constitution includes such a clause: see art. 101 (I
giudici sono soggetti soltanto alla legge).
66
67
286
to this effect.68 There are a number of ways to describe this

situation. One is to distinguish between validity and efficacy, so that in civil law systems only statutory law enjoys
validity (and provides grounds for valid judicial decisions),
whereas precedents are only efficacious or applied law
(American legal realists used to distinguish law in books
and law in action along similar lines). Another possible description is to say that while statutes are mandatory
sources, precedents are only permissive sources, sources
which courts may rely upon.69 Still another one is to say
that the constitutional clause restricting valid sources to
statute only is not practiced as the rule of recognition in
that legal system at best, it has to be supplemented with
an implicit rule of recognition according to which also precedents are grounds for judicial decisions. In such cases,
then, there is no complete coincidence between the foundational and the epiphenomenal levels of the rule of recognition.
To conclude this point. To assume that the rule of recognition validates the constitution is indeed a needless reduplication. The constitution is not validated by the rule of
recognition, rather it is the rule of recognition (or more accurately, the rule of recognition consists in the constitutional clauses that bear upon law-making activities), even if
only at the epiphenomenal level. The constitution represents the rule of recognition to the extent that it is actually
used in the officials law-making and law-ascertaining practices realistically, the written constitution is only part of
the rule of recognition, since the criteria it provides are easily supplemented with other, non-stated criteria as well
with criteria provided by other sources.
68 A rich comparative analysis is in N. MacCormick, R. Summers (eds.), Interpreting Precedents: A Comparative Study; and also J. Bell, Comparing Precedent.
69 Hart himself seems to embrace such a distinction: see The Concept of Law,
295; see also at 153-154, referring to the possibility that courts manipulate the
rule of recognition at its fringes.
287
GIORGIO PINO
The consequence, plainly, is that the constitution is not

itself valid law indeed, it is neither valid nor invalid.70
IV. THE RULE OF RECOGNITION REVISITED
Lets take stock. The many different interpretations that
the concept of a rule of recognition has undergone seem to
oscillate between two different defects. On the one hand,
the duty-based approaches to the rule of recognition ((a),
(b), (c), (e) above) cannot easily explain the fact that the
courts are not always under a legal duty to apply valid law
because they can be allowed either to depart from valid
law, or to apply non-valid law; so, it is inaccurate71 to postulate that the rule of recognition both states the criteria of
validity and imposes a duty on judges to apply all and only
the norms that comply with such criteria. On the other
hand, the power- or definitional-based approaches to the
rule of recognition ((d), (f), (g) above) cannot help but conflating the rule of recognition with the rules of change, or
even with the rules of adjudication. The first kind of defect
makes the rule of recognition an inadequate concept on the
explanatory level: there are a few interesting things under
its sway that are left quite unexplained. The second kind of
defect makes the rule of recognition a redundant concept: it
has no real job to do.
Is there a way to rescue the rule of recognition from these
defects? I believe there is, but only insofar as one is prepared to leave aside substantive portions of Harts original
account of the rule of recognition. Under that original account, as we have seen, the rule of recognition is deeply related to indeed, it is constitutive of legal validity. On re70 This conclusion is indeed endorsed by many positivists: see R. Guastini, Sur
la validit de la constitution du point de vue du positivisme juridique and On Legal
Order: Some Criticism of the Received View; L. Ferrajoli, Principia Iuris, vol. 1,
892-893.
71 I mean, it is inaccurate as a theoretical explanation of what goes on in adjudication, while it is not necessarily inaccurate as an interpretation of Harts
thought.
288
flection, though, it appears quite clear that such a job is

entirely performed by a different kind of rules. Validity is
not determined by the rule of recognition, but by the rules
that regulate the various law-making activities (legislation
and so on); and these, even on Harts account, are rules of
change.
So what is the role, if any, of the rule of recognition, once
it is severed from legal validity? Here is a possibility. The
rule of recognition is an array of normative considerations
that generally direct law-applying officials in the identification of the law that they are bound to apply in their institutional competence. A few clarifications are needed here.
First, the directives that compound the rule of recognition need not have the form of an obligation: they can also
be permissions72 (so, the rule of recognition can still qualify
as duty-imposing, but only in quite a broad sense).
Second, the rule of recognition does not necessarily instruct the law-applying institutions to apply only valid law.
Indeed, while valid law is normally apt for judicial application,73 the rule of recognition can also direct judges towards
the application of different kind of standards, including
moral considerations, foreign law, invalid law, and so on.74
In short, the rule of recognition is not a criteria of legal validity, but rather a criteria of legal applicability;75 it is a tool
of legal reasoning rather than a tool of legal dynamics.
Third, in any minimally complex legal system the rule of
recognition as here understood will necessarily include a hierarchical ordering of sources of law, as well as a (defeasible)
On permissive sources, see supra, fn 69 and accompanying text.

And in this sense what Hart says about the relation between the rule of recognition and the rules of change is perfectly coherent and non-redundant at all:
see The Concept of Law, 96.
74 See supra, fns 39-42 and accompanying text.
75 On the concept of applicability, see W. Waluchow, Inclusive Legal Positivism,
ch 3 (institutional force); P. Navarro, J. J. Moreso, Applicability and Effectiveness
of Legal Norms; P. Navarro, C. Orunesu, J. L. Rodriguez, G. Sucar, Applicability of
Legal Norms; G. Pino, Diritti e interpretazione, ch 2, and La aplicabilidad de la
normas jurdicas.
72
73
289
GIORGIO PINO
ordering of interpretive methodologies:76 the rule of recognition will include the preference for some interpretive methodology over others, or possibly even a ban on certain interpretive methodologies.
Fourth, the rule of recognition, as here understood, is the
key concept structuring the job of law-applying institutions
it directs them in accomplishing their institutional duties,
it compounds the reasons judges and officials generally
may have to apply the law. As such, it is normally the object of an ethical-political acceptance. Normally, officials
adopt a certain rule of recognition because that rule of recognition reflects their fundamental, and sometimes even
unstated as the case may be, ideals of political legitimacy.
The rule of recognition is a concept compounding the officials legal ideology.77 Officials follow it, they are committed
to it normally on moral grounds, because that rule of recognition (as here understood) is coherent, or is as coherent as
it can be, with their fundamental assumptions of political
legitimacy.78 The repeated normalcy qualification introduced above is intended to embrace the not entirely unrealistic case of officials who act on purely self-interested, prudential grounds, as well as that of the conformist ones.
76 Many commentators have suggested that the criteria of validity provided by
the rule of recognition should be supplemented with criteria of interpretation: see
for instance N. Bobbio, Norme primarie e secondarie, 187-188; K. Greenawalt,
The Rule of Recognition and the Constitution, 31-35; W. Waluchow, Inclusive Legal Positivism, 76 (on secondary rules of interpretation); E. Diciotti, Regola di
riconoscimento e concezione retorica del diritto, 11-16; and compare the concept
of meta-interpretation in S. Shapiro, Legality, 304-306 and passim.
77 I use here a concept analogous to that of legal or normative ideology deployed by A. Ross, On Law and Justice, 75-76 (it consists of directives which do not
directly concern the manner in which a legal dispute is to be settled but indicate
the way in which a judge shall proceed in order to discover the directive or directives decisive for the question at issue). A quite similar concept (il principio
fondamentale) is carved out also by U. Scarpelli, Cos il positivismo giuridico, ch
VII.
78 Cf N. MacCormick, Legal Reasoning and Legal Theory, 63-64, 139-140 (what
must be essential to the internal aspect of the rule of recognition is some conscious commitment to pursuing the political values which are perceived as underpinning it); J. Raz, On the Authority and Interpretation of Constitutions: Some
Preliminaries, 334 (the rule of recognition is a normative practice).
290
There is no reason to rule out such possibilities nor do they

affect the theoretical picture such as it has been drawn
here. It is reasonable to assume, anyway, that the bulk of
officials adhere to the system on ethical-political grounds.
This is meant to be an empirical, realistic assumption, not
a necessary conceptual truth of legal systems.
Fifth, what does ensure the unity of a legal system? For
sure, while the rule of recognition is the object of ethical-political acceptance (and so to a certain extent each official will try to mould the rule of recognition as closely as
possible to his own preferences), it is clear that each official
on his own part is not entirely free to choose a rule of recognition whatsoever. Every single official is integrated in an
overall institutional structure that is the result of a historical process as well as of power relations deeply entrenched in the relevant society. So in normal circumstances (i.e., barring the case of the anarchist, of the
foreign agent, and so on) each official will either adhere
wholeheartedly to the existing legal system,79 or will strive
for a compromise between his fundamental ideals of political legitimacy and the existing legal order. In other words,
he will strive for a compromise between the legal system as
he would like it to be, and the legal system as it emerges
from the prevailing practice of the other officials.80
I take this to be, by and large, also Harts view on the
subject. Albeit usually reluctant to investigate the reasons
officials may have for practicing a certain rule of recognition,81 Hart eventually conceded that such reasons include
79 This seems to be the paradigmatic instance of what MacCormick calls willing acceptance: see H. L. A. Hart, 109. MacCormick further analyses the concept of
acceptance of the law in The Concept of Law and The Concept of Law, 183-185.
80 For a similar point, see K. Greenawalt, The Rule of Recognition and the Constitution, 8-9; J. Waldron, Law and Who Needs Rules of Recognition?, 333-334.
81 Hart indeed excludes, obviously enough, that such reasons can qualify as legal reasons; but then he deliberately leaves unexplored the matter of what kind of
reasons exactly they might be (granted they are not legal reasons), since he believes that such an enquiry would be outside the scope of analytical, descriptive legal theory: The Concept of Law, 107-108; for a discussion of this point, see J. Dickson, Is the Rule of Recognition Really a Conventional Rule?.
291
GIORGIO PINO
also the fact that other officials practice a certain rule of

recognition.82 Now, it is surely true that the fact of an existing practice (legal or otherwise) is not per se a reason to
participate in it.83 But it is equally true that, if someone
wanted to take part in a certain legal practice (because he
has some ex ante views on the moral desirability of that
practice, or because it allows him to get a nice paycheck, or
a desirable social status, and so on), then hed better play
by the existing rules even also with a view to reinterpreting them, pressing for their reform, and so on. And in this
sense it is obviously true, indeed almost trivially true, that
among the reasons a judge might have to play along in the
legal practice there is also the fact that other relevant actors concur in the same practice. Wouldnt that be so, one
would have indeed some difficulty in distinguishing a judge
from, say, a science fiction writer.
In the end, then, convergence on by and large the same
rule of recognition by various officials will be the result of
something like an overlapping consensus: officials may
well have partly different views on the legitimacy of the existing legal order, but they converge on some fundamental
features of the existing practice (for instance, on a set of
undisputed sources of law).84 Would not such a basic convergence obtain, it would be difficult indeed to talk of an
existing legal system.
Accordingly, the rule of recognition is not entirely a datum
for those who practice it, but rather it is constantly subject to
change,85 because existing practice is constantly under the
pressure of competing ideals of legitimacy among officials. Indeed, one can view some long-standing debates within the judiciary and the legal academia, such as the debate between
Postscript, 255, 267.
As A. Marmor is right in emphasising: see Positive Law and Objective Values,
22, 33.
84 This is assumed even by Dworkin, when he refers to the preinterpretive
stage (the raw legal materials on which there is unchallenged widespread consensus): see Laws Empire, 65-67.
85 J. Raz, The Identity of Legal Systems, 94.
82
83
292
originalist, textualist, and living-tree approaches to constitutional interpretation, exactly in this way as attempts at
implementing (and asserting the superiority of) partly different rules of recognitions. They boil down to different ways to
identify (part of) the law that has to be applied by courts.86
On the whole, the resulting image of the legal system resembles less a pyramid than Neuraths boat.
The rule of recognition, then, far from being a list of criteria of validity, is the range of criteria and considerations
that guide the judge in his law-applying functions: it can
direct the judge to the application of valid legal norms or,
as the case may be, towards the application of some other
kind of standard, or even to the application of an extra-legal
norm; and it will include also the preference for some interpretive methodology. On the whole, in contemporary, complex legal systems, product of long historical stratification,
with multiple sources, and complicated interactions with
other normative systems (such as foreign legal systems, international organizations, and so on) it is quite implausible
that the rule of recognition resemble anything like a
well-structured, axiomatic set of criteria; more likely, as
Jeremy Waldron says, it will work as an array of
(defeasible) normative considerations,87 that will work more
or less in the way of the search for a reflective equilibrium.
Still, this normative array will surely be recognizable, and
its operations even predictable, to a certain extent: it will
represent the way in which officials conceive of their institutional role, and the way in which they regularly carry it on.
V. BIBLIOGRAPHY
ALEXANDER, L. and SCHAUER, F., Rules of Recognition,
Constitutional Controversies, and the Dizzying De86 A similar point is developed by L. Alexander and F. Schauer, Rules of Recognition, Constitutional Controversies, and the Dizzying Dependence of Law on Acceptance, 181-187.
87 J. Waldron, Who Needs Rules of Recognition?, 339.
293
GIORGIO PINO
pendence of Law on Acceptance, in ADLER, M. and

HIMMA, K. E. (eds.), The Rule of Recognition and the
U.S. Constitution, Oxford U. P., Oxford, 2009.
ATIENZA, M. and Ruiz MANERO J., A Theory of Legal Sentences, Kluwer, Dordrecht, 1997.
BELL, J., Comparing Precedent, Cornell Law Review, vol.
82, 1997, 1243-1278.
BOBBIO N., Norme primarie e norme secondarie (1968), in
Studi per una teoria generale del diritto, Giappichelli,
Torino, 1971.
, Norme secondarie (1975), in Contributi ad un dizionario giuridico, Giappichelli, Torino.
BULYGIN, E., Sobre la regla de reconocimiento (1976), in
ALCHOURRN, C. E. and BULYGIN, E., Anlisis lgico y
Derecho, Centro de Estudios Constitucionales, Madrid, 1991.
CARACCIOLO, R., El sistema jurdico. Problemas actuales, Centro de Estudios Constitucionales, Madrid, 1988.
CELANO, B., La regola di riconoscimento una convenzione?, Ragion pratica, 21, 2003.
COHEN, L., Critical Notice, Mind, LXXI, 1962.
COLEMAN, J., Authority and Reason, in GEORGE, R. P. (ed.
by), The Autonomy of Law. Essays on Legal Positivism,
Clarendon Press, Oxford, 1996.
, The Practice of Principle: In Defence of a Pragmatist
Approach to Legal Theory, Oxford U. P., Oxford, 2001.
DICIOTTI, E., Regola di riconoscimento e concezione retorica
del diritto, Diritto & Questioni Pubbliche, 7, 2007.
DICKSON, J., Is the Rule of Recognition Really a Conventional Rule?, Oxford Journal of Legal Studies, vol. 27,
No. 3, 2007.
DWORKIN, R., The Model of Rules I (1967), in Taking Rights
Seriously, Duckworth, London, 1978.
294
, Laws Empire, Bellknap, Cambridge (MA), 1986.

FERRAJOLI, L., Principia Iuris. Teoria del diritto e della democrazia. vol. 1: Teoria del diritto, Laterza, Roma-Bari,
2007.
FINNIS, J., On Harts Ways: Law as Reason and as Fact,
American Journal of Jurisprudence, 52, 2007.
FULLER, L. L., The Morality of Law, Yale U. P., New Haven,
1964, 1969.
GREEN, L., The Authority of the State, Clarendon Press,
Oxford, 1988.
, The Concept of Law Revisited, Michigan Law Review,
94, 1996.
, Positivism and Conventionalism, Canadian Journal
of Law and Jurisprudence, XII, 1, 1999.
GREENAWALT, K., The Rule of Recognition and the Constitution (1987), in ADLER, M. and HIMMA, K. E. (eds.), The
Rule of Recognition and the U.S. Constitution, Oxford
U.P., Oxford, 2009.
GUASTINI, R., Sur la validit de la constitution du point de
vue du positivisme juridique, in TROPER, M. and
JAUME, L. (eds.), 1789 et linvention de la constitution,
Bruxelles-Paris, 1994.
, Conoscenza senza accettazione, in GIANFORMAGGIO,
L. and JORI, M. (eds.), Scritti per Uberto Scarpelli,
Giuffr, Milano, 1997.
, On Legal Order: Some Criticism of the Received
View, Ethical Theory and Moral Practice, 2000, vol. 3,
Number 3.
, The Basic Norm Revisited, paper presented at the
conference Legal Science and Legal Theory, Oxford,
September 8th 2010.
HACKER, P. M. S., Harts Philosophy of Law, in HACKER, P.
M. S. and RAZ, J. (eds.), Law, Morality, and Society:
295
GIORGIO PINO
Essays in Honour of H. L. A. Hart, Clarendon Press,

Oxford, 1977.
HART, H. L. A., Positivism and the Separation of Law and
Morals (1958), in Essays in Jurisprudence and Philosophy, Clarendon Press, Oxford, 1983.
, The Concept of Law, Clarendon, Oxford, 1961, 1994.
, Lon L. Fuller: The Morality of Law (1965), in Essays
in Jurisprudence and Philosophy, Clarendon Press,
Oxford, 1983.
, Kelsens Doctrine of the Unity of Law (1968), in
Essays in Jurisprudence and Philosophy, Clarendon
Press, Oxford, 1983.
, Postscript, in The Concept of Law, Clarendon,
Oxford, 1994.
HIMMA, K. E., Making Sense of Constitutional Disagreement: Legal Positivism, the Bill of Rights, and the
Conventional Rule of Recognition in the United States, Journal of Law and Society, 4, 2003.
KELSEN, H., General Theory of Law and State, Harvard U. P.,
Cambridge (MA), 1945.
KRAMER, M., Where Law and Morality Meet, Oxford U. P.,
Oxford, 2004.
LAGERSPETZ, E., The Opposite Mirrors: An Essay on the Conventionalist Theory of Institutions, Kluwer, Dordrecht,
1995.
LEITER, B., Legal Realism and Legal Positivism Reconsidered (2001), in Naturalizing Jurisprudence: Essays on
American Legal Realism and Naturalism in Legal Philosophy, Oxford U. P., Oxford, 2007.
LYONS, D., Ethics and the Rule of Law, Cambridge U. P.,
Cambridge, 1984.
MACCORMICK, N., Legal Reasoning and Legal Theory, Oxford
U. P., Oxford, 1978, 1994.
296
, H. L. A. Hart, Edward Arnold, London, 1981.

, A Moralistic Case for A-Moralistic Law, Valparaiso
Law Review, vol. 20, 1, 1985.
, The Concept of Law and The Concept of Law, in
GEORGE, R. P. (ed. by), The Autonomy of Law. Essays
on Legal Positivism, Clarendon Press, Oxford, 1996.
and SUMMERS, R. (eds.), Interpreting Precedents: A
Comparative Study, Dartmouth, Aldershot, 1997.
MARMOR, A., Positive Law and Objective Values, Clarendon
Press, Oxford, 2001.
Philosophy of Law, Princeton U. P., Princeton, 2011.
MORESO, J. J., Legal Indeterminacy and Constitutional Interpretation, Kluwer, Dordrecht, 1998.
MUNZER, S., Legal Validity, Martinus Nijhoff, The Hague,
1972.
NAVARRO P. and MORESO J. J., Applicability and Effectiveness of Legal Norms, Law and Philosophy 16, 1997.
NAVARRO, P. et al., Applicability of Legal Norms, Canadian
journal of law and jurisprudence, 2004, 2.
PERRY, S., Where Have All the Powers Gone? Hartian Rules
of Recognition, Noncognitivism, and the Constitutional and Jurisprudential Foundations of Law, in
ADLER, M. and HIMMA, K. E. (eds.), The Rule of Recognition and the U.S. Constitution, Oxford U. P., Oxford,
2009.
PINO, G., Norme e gerarchie normative, Analisi e diritto,
2008.
, Diritti e interpretazione. Il ragionamento giuridico nello
Stato costituzionale, il Mulino, Bologna, 2010.
, La aplicabilidad de la norms jurdicas, forthcoming
2012.
297
GIORGIO PINO
POSTEMA, G., Coordination and Convention at the Foundations of Law, Journal of Legal Studies, vol. 11, No. 1,
1982.
RAZ, J., The Concept of a Legal System, Clarendon Press,
Oxford, 1970, 19802.
The Identity of Legal Systems (1971), in The Authority of Law, Oxford U.P., Oxford, 1979, 2009.
, Practical Reason and Norms, Oxford U. P., Oxford,
1975, 1990.
, The Functions of Law (1973), in The Authority of
Law, Oxford U. P., Oxford, 1979, 2009.
, Legal Positivism and the Sources of Law in The Authority of Law, Oxford U. P., Oxford, 1979, 2009.
, The Purity of the Pure Theory (1981), in The Authority of Law, Oxford U. P., Oxford, 1979, 2009.
, Authority, Law, and Morality (1985), in Ethics in the
Public Domain. Essays in the Morality of Law and Politics,
Oxford U. P., Oxford, 1994.
, On the Authority and Interpretation of Constitutions: Some Preliminaries (1998), in Between Authority and Interpretation, Oxford U. P., Oxford, 2009.
ROSS, A., On Law and Justice, London, Stevens, 1958.
, Validity and the Conflict between Legal Positivism
and Natural Law (1961), in Normativity and Norms:
Critical Perspectives on Kelsenian Themes, edited by S.
PAULSON, B. Litschewski Paulson, OUP, 1999.
SCARPELLI, U., Cos il positivismo giuridico, edizioni di Comunit, Milano, 1965.
SCHIAVELLO, A., Perch obbedire al diritto? La svolta convenzionalista e i suoi limiti, ETS, Pisa, 2010.
SHAPIRO, S., On Harts Way Out, Legal Theory, 4, 1998.
, Law, Plans, and Practical Reason, Legal Theory, 8,
2002.
298
, What Is the Rule of Recognition (and Does It

Exists)?, in ADLER, M. and HIMMA, K. E. (eds.), The
Rule of Recognition and the U.S. Constitution, Oxford
U. P., Oxford, 2009.
, Legality, Bellknap Press, Cambridge (MA), 2011.
STAVROPULOS, N., Interpretivist Theories of Law, in ZALTA
Edward N. (ed.), The Stanford Encyclopedia of Philosophy, Fall 2008 edn, http://plato.stanford.edu/ar
chives/ fall2008/entries/law-interpretivist/.
WALDRON, J, Normative (or Ethical) Positivism, in COLEMAN,
J. (ed.), Harts Postscript: Essays on the Postscript to
The Concept of Law, Oxford U. P., Oxford, 2001.
Law, in Oxford Handbook of Contemporary Philosophy, edited by F. Jackson and M. Smith, Oxford U.
P., Oxford, 2005.
, Who Needs Rules of Recognition?, in ADLER, M. and
HIMMA, K. E. (eds.), The Rule of Recognition and the
U.S. Constitution, Oxford U. P., Oxford, 2009.
WALUCHOW, W., Inclusive Legal Positivism, Oxford U. P.,
Oxford, 1994.
, Four Concepts of Validity, in ADLER, M. and HIMMA,
K. E. (eds.), The Rule of Recognition and the U.S. Constitution, Oxford U. P., Oxford, 2009.
299
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
H. L. A. HART AND THE INVENTION OF LEGAL

PHILOSOPHY
Dan PRIEL*
Resumen:
En este ensayo se plantea la idea de que en cierto modo la filosofa del
derecho al menos en la forma en que es entendida entre los filsofos
analticos del mundo angloparlante es en gran medida una invencin
de H. L. A. Hart. Con la obra de Hart, en efecto, la bsqueda del concepto o naturaleza del derecho, se ha consolidado como un objeto de investigacin independiente, que conscientemente pretende deslindarse de las
cuestiones polticas o morales. Al definir el campo de estudio de la filosofa del derecho en este sentido, Hart no slo se aparta del trabajo de Thomas Hobbes y de Jeremy Bentham, cuyos compromisos polticos son evidentes, sino tambin del proyecto aparentemente ms cercano de John
Austin. Despus de sealar la diferencia entre los proyectos de Austin y
de Hart, el autor pone en cuestin la direccin que ha tomado la filosofa
del derecho al ser encabezada por Hart.
Palabras clave:
Jurisprudencia analtica, filosofa del derecho, positivismo jurdico, naturaleza del derecho, H. L. A. Hart.
Assistant Professor, Osgoode Hall Law School.
301
DAN PRIEL
Abstract:
In this essay I argue that in some sense legal philosophy, at least as the
term is now understood among analytic jurisprudents in the Anglophone
world, is to a large extent a creation of H. L. A. Harts work. It is with him
that the search for the concept or the nature of law was one established as
an independent object of inquiry, that consciously tried to avoid moral or
political questions. In framing the province of jurisprudence in this way
Hart not only departed from the work of Thomas Hobbes and Jeremy
Bentham, whose political commitments are clear, but also from the seemingly much closer enterprise of John Austin. After demonstrating this difference between Austins enterprise and Harts, I criticize the direction legal
philosophy has taken following Harts lead.
Keywords:
Analytical Jurisprudence, Legal Philosophy, Legal Positivism,
Nature of Law, H. L. A. Hart.
302
H. L. A. HART AND THE INVENTION OF LEGAL PHILOSOPHY
SUMMARY: I. Introduction. II. Austins Novelty. III. Harts

Inversion. IV. The Invention of Legal Philosophy.
V. Conclusion: Harts Faustian Bargain.
I. INTRODUCTION
In the course of a rather heated exchange in the pages of
the Times Literary Supplement Leslie Green and Brian Leiter
wrote: no one holds the view Professor [Brian] Simpson ascribes to us, namely that legal philosophy as a serious subject was invented by Hart.1 Well, this is, more-or-less, the
thesis I am going to defend here. This is not because I have
not heard of Aristotle, Bentham, Cicero, Duguit, Ehrlich,
Frank, Grotius, Hobbes, or Ihering. I have, in fact, written
about the fact that much of the plight of contemporary jurisprudence is the result of how little attention is paid to
the work of historical figures, who were often more original
and interesting than the thinkers whose works legal philosophers have analyzed to dust. My point, rather, is that with
Hart the English-speaking world has seen the creation of a
new way of thinking about jurisprudence, and that in doing
so he, in some sense, created a new subject. But whereas
Green and Leiter think of pre-Hart jurisprudence as a
dilettantish pastime for law teachers and retired judges, an
undisciplined jumble of history, speculative sociology, legal
doctrine and party politics, [that] became [with Hart] a
technical and rigorous branch of philosophy,2 I see Harts
contribution to jurisprudence in much more negative
terms.
1 Green, Leslie J. & Brian, Leiter, Letter to the Editor, H. L. A. Hart, Times Literary Supplement, April 15, 2005, p. 15.
2 Green, Leslie J. & Brian, Leiter, Letter to the Editor, H. L. A. Hart and The
Concept of Law, Times Literary Supplement, March 11, 2005, p. 15. By the way, it
should not be forgotten that philosophy in general from the first half of the twentieth century often appears quite amateurish by todays standards. Look at old volumes of the leading philosophical journals from those years and much of what appears there seems insufficiently rigorous by todays standards.
303
DAN PRIEL
This essay is part of a larger attempt to assess Harts impact on jurisprudence, and in particular that of his book
The Concept of Law,3 a book whose continued influence I
find less of a cause for celebration than others. My view,
one I will not defend here, is that Harts substantive theory
of law as articulated in The Concept of Law is largely a failure. In part it is because of the major issues Hart left out of
the book, which I think are crucial for any successful theory of law; in part this is because on many important issues Hart does address, his views are not very clear; and in
part it is because whenever he is clear, his explanations are
almost invariably less than successful. The Concept of Law
has the further demerit of seriously simplifying and misrepresenting the views of opposing views, which renders Harts
critiques of those views not very illuminating. Though many
of the misrepresentations in Harts account of the work of
legal realists, natural lawyers, and even earlier legal positivists have by now been pointed out, there is no doubt that
many have taken their first (and often last) impression of
these schools of thought from Hart, something that did not
bode well for the subject. True, The Concept of Law was
meant as a student textbook, and this inevitably required
some simplification at the expense of clarity of exposition.
But what one finds in Harts book are serious distortions of
fundamental ideas of other legal theorists, some of which
are still with us.
What then is Harts major contribution? I think Harts
most lasting influence on the field is the one that is probably least often noticed, and that is the setting of the boundaries of what actually belongs in the subject. The domain of
jurisprudence today is largely understood in the terms that
were defined by Hart. Harts legacy is best defined by what
jurisprudence is and, importantly, is not taken to be
3 Hart, H. L. A., The Concept of Law, 2nd ed., Oxford, Clarendon Press, 1994.
Further references to the book are made to Concept, parenthetically in the text. Citations throughout are to the (slightly differently paginated but otherwise identical) second edition.
304
about. But Hart did not begin with a clean slate: as is clear
to any reader of The Concept of Law, despite Harts criticisms of Austin, Harts own theory is built on Austinian
foundations. The purpose of this essay is to describe how
Hart determined the province of jurisprudence by subtly
but fundamentally shifting the ideas he took from Austin.
II. AUSTINS NOVELTY
Mr. Austin once said of himself, that if he had any special intellectual vocation, it was that of untying knots. So
said Mill of his former teacher. And as Mill added, this description fit Austin very correctly.4 Indeed. With seemingly
unbounded amounts of pedantry he sought to tell us what
positive law is. This required clarity of mind and well-kept
conceptual books, and at times it looks as though Austin
cared for little else. Ad nauseum one reads him telling us
what law (and other things) mean when they are properly
so called, strictly so called, or when they are used in the
proper acceptation of the term. When it came to law, this
required untying the knot that tied it to morality. This untying was not because there was no, substantively, necessary connection between the two. Austin never utters the
expression no necessary connection between law and morality, and his famous slogan the existence of law is one
thing, its merit or demerit is another,5 does not come remotely close to confirming it. What he does say, in fact, is
that [t]he proper purpose or end of a sovereign political
government or the purpose or end for which it ought to exist is the greatest possible advancement of human happiness (Province, p. 242), and this claim, if true, arguably
establishes some kind of connection between law and mo4 Mill, John Stuart, Austin on Jurisprudence, The Collected Works of John
Stuart Mill, Vol. 21, Toronto, University of Toronto Press, 1984, 167-205, p. 168
(first published 1863).
5 Austin, John, The Province of Jurisprudence Determined, Wilfrid E. Rumble
ed., Cambridge, Cambridge University Press, 1995, p. 157 (first published 1832).
Further references to this book will be made to Province parenthetically in the text.
305
DAN PRIEL
rality. On the other hand unlike contemporary legal philosophers Austin did not try to identify those features that all
legal systems necessarily have in common. He explicitly
limits his claims to the ampler and maturer legal systems,
whatever the proper acceptation of that term may be.6
It is true that he says that [w]ith the goodness and badness of laws [jurisprudence] has no immediate concern,
(Study, p. 350) but that is a methodological device needed,
he thought, for the clarity of explanation. As he put it:
Of laws or Rules there are various classes. Now these classes
ought to be carefully distinguished. For the confusion of them
under a common name, and the consequent tendency to confound Law and Morals, is one most prolific source of jargon,
darkness, and perplexity. By a careful analysis of leading
terms, law is detached from morals, and the attention of the
student of jurisprudence is confined to the distinctions and
divisions which relate to law exclusively. (Study, p. 355,
emphasis added.)
Or in a different formulation:
The matter of jurisprudence is positive law . But positive
law is often confounded with objects to which it is related
by resemblance, and with object to which it is related in the
way of analogy: with objects which are also signified, properly and improperly, by the large and vague expression law.
To obviate the difficulties springing from that confusion, I
begin with determining the province of jurisprudence, or
with distinguish the matter of jurisprudence from those various related objects trying to define the subject of which I
intend to treat, before I endeavor to analyse its numerous
and complicated parts (Province, p. 18).
To understand how Austin sought to keep law and morality separate in this sense, we need to see his position in the
line of thinkers that begins with Hobbes and ends with
6 Austin, John, On the Study of Jurisprudence, Lectures on Jurisprudence,
London, Murray, Vol. 3, 1863, 349-75, p. 349. Further references to this essay will
be made to Study parenthetically in the text. See also Austin, Province, p. 165.
306
Hart and contemporary legal positivism. I think a key to

understanding the way he offered a non-political account of
sovereignty, departing in this respect from what one finds
in the work of Hobbes, and also (and perhaps more surprisingly) Bentham.
Hobbes identifies the essence of a commonwealth with
the existence of a person or body that hath the use of so
much Power and Strength conferred on his, that by terror
therefore, he is inabled to conforme the wills of all to maintain peace. This person of whose Acts a great Multitude, by
mutuall Covenants one with another, have made themselves every one the Author is the sovereign.7 Immediately
after these words come a chapter dedicated to enumerating
the rights and limits on sovereign power. For this reason,
even though Hobbes favors granting sovereigns very broad
powers, his account of sovereignty is not tantamount simply to whomever happens to be in power (as long as he
protects his subjects). It is part of the contractarian story
Hobbes offers as the grounds of legitimate political
authority.
Bentham is interesting as well. It is true that his account
of political power as an observable habit of, and disposition to obedience: habit, speaking with respect to past acts;
disposition, with respect to future8 looks like the same sort
of explanation that (as we shall see in a moment) was later
found in Austins work. This, however, was only part of
what Bentham had to say on sovereignty, and much of it,
went beyond observations. His theory of law is tied to an
account of sovereignty, which in turn addressed both the
question of normativity (how law can bind) and the question of legitimacy (the conditions under which it succeeds
7 Hobbes, Thomas, Leviathan, Richard Tuck ed., Cambridge, Cambridge University Press, 1996, pp. 120-21 (ch. 17) (first published 1651).
8 Bentham, Jeremy, A Fragment on Government, J. H. Burns & H. L. A. Hart
(eds.), Oxford, Oxford University Press, 1988, pp. 101-102 (IV.35) (first published
1776).
307
DAN PRIEL
in binding).9 After embracing democracy relatively late in

his life Bentham came to be believe that sovereignty should
be vested in the people. But even before that his explanation of the normativity of law was grounded in the existence
of a reciprocal relationship between ruler and ruled. This
allowed him to distinguish between legislative (and executive) sovereignty and autocratic sovereignty on the basis of
the question whether the sovereign power was exercised by
rule or without rule.10 For these reasons it is a mistake to
think that for Bentham sovereignty was a matter of pure
observation. Like Hobbes, his theory of law was part of a
broader political theory, and his views on sovereignty must
be understood as one ingredient within that bigger picture.
Now consider Austin: It is with him that we find the definition of sovereignty that has come to be associated with
command theories of law. An independent political society
is one that has a sovereign: the two terms, says, Austin
are inseparably connected (Province, p. 165). The marks of
the sovereign are that [t]he bulk of the given society are in
a habit of obedience to a determinate and common superior and that the person or body who is in that position is
not in a habit of obedience to a determinate human superior (Province, p. 166). Thus, the definition of sovereignty
has been decoupled from an account of legitimacy and has
been rested on factual grounds. When read together with
Austins definition of law a kind of command made by
political superiors to political inferiors (Province, p. 18)
the result is what might have been called, had others not
9 See Postema, Gerald J., Bentham and the Common Law Tradition, Oxford,
Clarendon Press, 1986, pp. 232-37, 255, 260-62. Postema does not distinguish between normativity and legitimacy in the way I do in the text, but it is clear from his
discussion that Bentham was concerned with both. For the distinction between
the two and its significance to jurisprudence see Priel, Dan, The Place of Legitimacy in Legal Theory, McGill Law Journal, Vol. 57, 2011, 1-35.
10 See Bentham, Jeremy, An Introduction to the Principles of Morals and Legislation, Oxford, Clarendon Press, 1996, p. 263 note 4 (first published 1780) for the
complete discussion.
308
claimed this title for themselves, the pure theory of law,11

as the link between sovereignty and lawmaking power has
become conceptual: the sovereign is whomever makes
commands that are habitually obeyed and does not obey
others; and those habitually obeyed commands are laws.
While Austin clearly sought to offer a non-political theory
of law, and in this regard, he does have an important role
in the story of the birth of legal philosophy, his account was
still a step away from what would happen a century later in
Harts account, because this conceptual link between law
and political power allowed Austin to offer an unambiguous
explanation of the normativity of law. There is an inseparable connection between command, duty, and sanction, he
says (Province, p. 24). Laws are commands, made by political superiors, the latter being understood in terms of
might: the power of affecting others with evil or pain (Province, p. 30, see also p. 282). Translated to the language of
modern jurisprudence Austin offers a certain theory explaining what laws are through an account of obedience to
commands. One feature of this view, already mentioned, is
that he seeks to explain the normativity of law in a manner
that is completely independent of political questions of legitimacy. The other, and the one which is of greater significance for my purposes, is that the identification of valid
legal norms is dependent on, and secondary to, his account
of normativity. For Austin the explanation of laws
normativity is not a puzzle that requires solving over and
above his account of identifying legal norms; rather, it is
the basis for identifying them and part of his theory of what
law is.
11 For a comparison between Austins and Kelsens theories see Kelsen, Hans,
The Pure Theory of Law and Analytical Jurisprudence, Harvard Law Review, Vol.
55, 1941, 44-70, pp. 54-66. Some of the criticisms Kelsen directs at Austins work
anticipate similar points found in Hart, The Concept of Law, supra note 3.
Admittedly, Kelsens theory is purer than Austins and there is no doubt that
Kelsen (to some extent via his influence on Hart and Raz) is relevant for a full picture of the state of contemporary analytic jurisprudence today. That, however, is a
question for another day.
309
DAN PRIEL
Austin defines positive laws laws that flow from human sources (Province, p. 110) as a species of commands. A command is then defined as an intimat[ion of] a
wish that [one] shall do or forbear from some act and that
intimation comes with an evil in case [one] compl[ies] not
with [the] wish (Province, p. 21). The direct link between
obligation and (threat of) sanction is clear when Austin discusses the question whether obligation exists whether the
magnitude of the eventual evil matters for the question of
obligation. Austin replies in the negative: The sanction, if
you will, is feeble or insufficient; but still there is a sanction, and, therefore, a duty and a command (Province, p.
23, emphasis in original).12 There is sanction and therefore
there is obligation.
In Austins account there is no puzzle in understanding
in what way law creates obligations, because the very definition of law seeks to remove all mystery from the matter.
Thus, while Austin famously denied that an unjust law is
not law,13 his account implied that for him a non-threatening law is not law.
This is a crucial point in Austins account: what counts
as law is identified by appealing to an account of legal obligation. One thing this point helps us see and better understand is a feature of Austins ideas that has come under
considerable attack, namely the link between obligation
and sanction. What we now see, however, is that this criticism is based on a misunderstanding of the way Austin understands the relationship between validity (law) and
normativity (coercion). As we shall see in a moment, Hart
12 See also Province, p. 118: Every duty properly so called supposes a command by which it is created. For every sanction properly so called is an eventual
evil annexed to a command. And duty properly so called is obnoxiousness to evils of
the kind.
13 Suppose an act innocuous, or positively beneficial, be prohibited by the sovereign under the penalty of death; if I commit this act, I shall be tried and condemned, and if I object to the sentence, that it is contrary to the law of God, who
has commanded that human lawgivers shall not prohibit acts which have no evil
consequences, the Court of Justice will demonstrate the inconclusiveness of my
reasoning by hanging me up (Province, p. 158).
310
understands the relationship between the two in a very different way (he does not identify laws through an account of
normativity), which is why he concluded that their
connection was contingent.
III. HARTS INVERSION
The conventional wisdom is that Hart demolished Austins most important ideas. And up to a point, this is true.
But what must not be forgotten when reading Harts criticism of Austin is the he chose Austin as the subject of his
criticism because he thought there were elements in them
he could use to develop his own ideas. For this reason what
is interesting is not so much Harts critique of Austin, but
the extent to which Hart followed Austin (cf. Hart, p. vii),
the extent to which he adopted Austins general approach
his demarcation of jurisprudence without really arguing
for it. After all, Austins was not the only way of thinking
about jurisprudence, and though he was popular and appeared in many of the books Hart hated so much, the
books form which one learns what other books contain
(Concept, p. vii), he was not the only one featured in these
books and was far from being universally admired.14
Building on Austin, then, was not forced upon Hart. The
reason Hart chose to do that was because, consciously or
not, he felt a certain affinity with Austins work: part of it
must have been their shared concern for intellectual tidiness and clear-headedness,15 and another with the absence
14 In fact, many of the criticisms of Austin now associated with Hart were made
by others many years earlier. See generally Rumble, Wilfrid E., Doing Austin Justice:
The Reception of John Austins Philosophy of Law in Nineteenth-Century England,
London, Continuum, 2005, pp. 225-41: One of the most remarkable features of the
nineteenth-century reception of Austins work is the significant extent to which is
foreshadows H. L. A. Harts highly influential criticisms of Austin (quote at 225).
15 See Hart, H. L. A., Positivism and the Separation of Law and Morals, Harvard Law Review, vol. 71, 1958, 593-629, pp. 593-94; see also Hart, H. L. A., A
View of America, Listener, Vol. 59, 1958, 89-90, p. 90; cf. White, Alan R., Austin
as a Philosophical Analyst Archiv fr Rechts- und Sozialphilosophie, Vol. 64, 1978,
379-399, for the suggestion that Austin could be considered an early proponent of
analytic (here: ordinary language) philosophy.
311
DAN PRIEL
of Hart pejoratively called metaphysics in Austins work.

And so, not surprisingly, Harts account appears, at first
(and second) look, quite similar to Austins. After all, even
in the official story as told by Hart himself Austins account contained several major deficiencies,16 which, once
corrected, resulted in Harts account. Most fundamentally,
Hart took from Austin the separation of jurisprudence from
political theory, something one does not find in the major
contributions to jurisprudence by other philosophers now
called legal positivists.
Even though in all this Hart is closer to Austin than one
would think from the space he dedicates to critiquing him,
there is a different sense, one that to my knowledge has not
been acknowledged, in which Hart did depart from Austins
approach. We have seen already that for Austin to understand what law is is fundamentally to understand how it
obligates; as argued above, for him it is from understanding
how law obligates that one can learn how to identify individual (valid) legal norms. For Hart, by contrast, the answer to the question what is law is different. Indeed, in a
sense Hart answers it in exactly the opposite way than Austin does. If Austin starts with explaining the sense in which
laws create obligations and identifies valid legal norms on
the basis of that, Hart starts with validity and then constructs from his account of validity an explanation of laws
normativity. Indeed, it is probably the structure of his argument that explains why some readers doubted he even had
an account of laws normativity. Michael Moore, for example, has argued that Hart was not concerned with the question when people have a legal obligation, but only with the
16 I will not deal here with the question whether Hart described Austin accurately. There are those who have strongly argued that he has not. See, in particular, Moles, Robert N., Definition and Rule in Legal Theory: A Reassessment of H. L. A.
Hart and the Positivist Tradition, Oxford, Blackwell, 1987, especially ch. 2. There
have also been those who have argued that Austins account has the resources to
address many of the charges raised by Hart. See Hardin, Russell, Sanction and
Obligation, Monist, Vol. 68, 1985, 403-18; Schauer, Frederick, Was Austin Right
After All? On the Role of Sanctions in a Theory of Law, Ratio Juris, Vol. 23, 2010,
1-21. I do not take a stand on this issue here as well.
312
question when people put believe they have one.17 This account fits Harts claim to be engaged in descriptive sociology (Concept, p. vi), but it comes at a high cost: this is descriptive sociology of the useless kind. It is not based on
anything that would be remotely acceptable as evidence for
a social scientific explanation: no interviews, questionnaires, collection of data or anything that could justify the
claims Hart is making, if understood in these terms.
Therefore, the alternative reading, viz. that Harts practice theory of rules is an account of normativity, looks more
fruitful, or at least more charitable. However, at first it
seems equally unpromising. If taken as a philosophical
elucidation of legal obligation it seems to say, more or
less, that people are under a social obligation when they
believe they are under social obligation. And the problem
with this view is that, as Stephen Perry once put it bluntly,
believing does not make it so.18
In fact, however, we can discern a more elaborate account in The Concept of Law. For Hart, at a minimum, A is
under a legal obligation if
(1) there are relevant others who treat certain sources as a
source of prescriptions;
(2) the relevant others have power and they use it to ensure that the prescriptions contained in the sources are generally obeyed (Concept, pp. 103-04);19 and
17 Moore, Michael S., Educating Oneself in Public: Critical Essays in Jurisprudence, Oxford, Oxford University Press, 2000, p. 85: Hart is giving us a sociology of ethics. He is analysing when people regard themselves as bound by prescriptive rules, and more particular, when people regard themselves as bound by the
social rules of obligations.
18 Perry, Stephen R., Interpretation and Methodology in Legal Theory, in
Andrei Marmor (ed.), Law and Interpretation: Essays in Legal Philosophy, Oxford,
Clarendon Press, 1995, 97-135, p. 122.
19 Why the strong demand for power that is used? Because the most minimal
case of a legal system according to Hart is one in which the As only obey the law because it creates threats. Hart claims that as a conceptual truth these are cases of a
legal system. As he puts it those rules of behaviour which are valid according to
the systems ultimate criteria of validity must be generally obeyed. [This] condition is the only one private citizens need satisfy: they may obey each for his part
only and from any motive whatever (Concept, p. 116). For the difficulties this
313
DAN PRIEL
(3) those sources contain prescriptions directed at A.
There are familiar problems of circularity here (the others

have to be what I called relevant, and they are relevant in
the sense that they are legal officials). I set them aside here.
What matters is that what gets the legal system going, what
separates a set of directives from a legal system (both words
are important here: it is a system of laws) is the existence of
a certain attitude among the officials. But with regard to
the officials, their obligation, their rule of recognition consists of nothing more than the fact that they think they are
under a certain obligation.
Is this sufficient for explaining laws normativity? Let us
spell out the argument: Harts claim is that one can be under a legal obligation when one treats the law as a threat to
be avoided, and someone else treats law as reason for action. Is this not simply people believing that they are under
an obligation? I think Harts response to this question
would be: Thats all there is to it. To look for something
more than that, for anything deeper, is to maintain remnants of the old metaphysical worldview of natural law.
Now, notice that once you have accepted Harts notion of
validity, you have ipso facto understood his sense of the
normativity of law. For a legal norm to be valid means for
it to exist, and for it to exist means for it to be obligatory in
this sense. It is worth noting that Hart wrote in the Postscript that his account aimed to give an explanatory and
clarifying account of law as a complex social and political
institution with a rule-governed (and in that sense normative) aspect (Concept, p. 239, emphasis added). Being normative is nothing more than being rule-governed, and
being rule-governed is exactly Harts account of law as
union of primary and secondary rules.
raises to Harts view see Priel, Danny, Sanction and Obligation in Harts Theory of
Law, Ratio Juris, Vol. 21, 2008, 404-11.
314
On this view if you are looking for more, you are committing a category mistake. Gilbert Ryle, who coined this
term used the example of the visitor to Oxford who, after
seeing the colleges, academic departments, libraries, administrative offices and so on still asks, But where is the
university? The answer, says, Ryle, is that once you have
seen all that the visitor saw the University has been
seen.20 In effect, Hart charges the critic who looks for more
than this, to be making exactly the same mistake. Once you
have understood Harts account of law as a union of primary and secondary rules, the normativity of law has been
seen.21
Why, you may wonder, does all this matter? So Austin
started with normativity and Hart with validity, what difference does it make? The difference may seem subtle but I
think it is important for understanding the path of analytic
jurisprudence in the last fifty years. Hart aimed to show
that good analytic jurisprudence is sociology, because once
he has elucidated the meaning of the relevant terms, there
was no more real sociological work to be done. The gathering of empirical data will not add anything of value to
what good analytic philosophy can do.22 I do not know of
any contemporary legal philosopher who accepts these
views; but even though the particular account may not
have contemporary defenders, this view has been instrumental in establishing the view that legal philosophy is a
different, closed, even isolated domain. By answering the
puzzle of the normativity of law internally, i.e. by concepSee Ryle, Gilbert, The Concept of Mind, London, Hutchinson, 1949, p. 16.
Contra Shapiro, Scott J., Legality, Cambridge, Harvard University Press,
2011, p. 103. Shapiro accuses Hart of committing a category mistake by offering a
reductive account of laws normativity. This claim is doubly odd: first, it is exactly
the point of reductive accounts to explain one phenomenon in terms of another: if
that were a category mistake, then no reductive explanation would be possible;
second, as explained in the text Hart offers an explanation that, whatever its merits happen to be, clearly avoids the charge of category mistake.
22 Of sociology as done by sociologists Hart admitted he had always been mistrustful. Hart Interviewed: H. L. A. Hart in Conversation with David Sugarman,
Journal of Law and Society, Vol. 32, 2005, 267-93, 289.
20
21
315
DAN PRIEL
tual analysis and armchair sociology, Hart has severed the

link between jurisprudence and both normative philosophy
and empirical inquiry at the point where one would have
thought them most necessary. From this it is but a small
step to the view more popular today that analytic jurisprudence and the social sciences have very little to say to each
other because of their different concerns.23
It is not my intention to defend Austins account. There is
no doubt that his work has contributed to the isolationist
trend in jurisprudence. But it is worth noting that even
Austins view is in some respect more open than Harts in
that it seeks to explain what law is from an external perspective. After Hart this is often taken to be a criticism of
Austin, but we need to distinguish here between two senses
in which a theory of law may be external. In one sense, it is
external in that it ignores the attitudes of people engaged in
law. Whether or not Austin is guilty of this charge, this
sense of externality is not my concern here. Rather, my
point is that Austin sought to give an account of law that
does not follow existing linguistic usage, that does not follow the lawyers perspective. Austin is quite clear that his
account of what law is does not correspond exactly with linguistic usage: for the purposes of his account, he says,
whether an order [is] called a law, seems to depend on
circumstances which are purely immaterial (Province, p.
27). This is no coincidence. Such discrepancy follows inevitably from any account that does not start with legal validity, that does not seek to explain what law is by identifying
what legal practitioners think law is. Whatever its other
failings might be, this is an important advantage of Austins
account has over Harts: it is only from an external perspective (in this sense) that one can that one avoids the trap of
armchair (and hence, bad) sociology.
23 See Raz, Joseph, The Authority of Law: Essays on Law and Morality, 2nd ed.
Oxford, Oxford University Press, 2009, pp. 44, 104-05; Shapiro, supra note 21, pp.
406-07 note 16.
316
IV. THE INVENTION OF LEGAL PHILOSOPHY

As I said above, before Hart and Austin philosophizing
about law was never separated from normative questions.
There was, however, another form of theoretical inquiry
about law, the analysis of legal concepts. When this was
confined to a particular legal system, this was simply part
of the study of a particular area of law; when an attempt
was made to generalize beyond a single legal system, this
was termed jurisprudence. Here, again, Hart and Austin
take different sides. What Austin lectured his students
about what appears in his Lectures on Jurisprudence is
predominantly concerned with this sort of inquiry. There
are fifty seven lectures in the posthumously published full
set, and even this is only part of the larger project as Austin
envisaged it. The Province of Jurisprudence Determined, the
book that contemporary legal philosophers take some interest in, contains only six lectures, of which three have to do
more with moral than with legal philosophy. This creates a
somewhat skewed understanding of Austins work and the
place of the demarcation project within his overall interests.
Hart, then, has redrawn the borders of the province of jurisprudence from the way they were understood by Austin
in two important respects. The first, the shift from a
normativity-first account to a validity-first account, transformed the demarcation of a domain of inquiry (whatever
happens to be positive law is the subject-matter of jurisprudence) to a very different substantive claim, put succinctly
by John Gardner: there is no such thing as nonpositive
law.24 Harts commitment to this view is evident from the
fact that under his framework the conceptual impossibility
of non-positive law is true by definition. In the validity-first
framework something can be law only if it is identified according to some test of validity. The content of the test can
change from one legal system to another, but the existence
24 Gardner, John, Some Types of Law, Douglas E. Edlin ed., Common Law
Theory, Cambridge, Cambridge University Press, 2007, 51-77, p. 75.
317
DAN PRIEL
of the test itself is beyond question. And since the mark of a

positive law is that it is (legally) valid, then the claim becomes impossible to refute. This is not Austins view. As we
have seen above, for him limiting jurisprudence to positive
law was a demand of clarity of understanding, not a claim
about what law is.
The second is the narrowing down of domain of jurisprudence. For Austin it included (one might even say consisted in) the analysis of the legal concepts. This is the
material that makes up the bulk of his full Lectures and
for which the material found in The Province was meant to
serve only as boundary-setting introduction (Province, pp.
288-89). Following Hart the boundary-setting inquiry, now
called the search for the concept or nature of law, has
become the aim. Jurisprudence has now been reconceived
as either the descriptive or conceptual attempt to answer
the question What is law? or the normative inquiry into
the question what the law (of contract, of crimes, of torts)
should be. But the latter question was really now reconsidered as a question of moral philosophy about which the
first, conceptual, inquiry had relatively little to contribute.
The effect of this has been insufficient consideration of the
extent to which the internal workings of the law take (its
structure and organization, the content of its doctrines)
may be relevant to an answer to the more general questions
about law. Consequently, analytic jurisprudents paid little
attention not only to legal doctrine, but also to legal history,
economic theories of law, and even to political thought.
These two changes can be nicely demonstrated by considering the way the term general jurisprudence (or universal jurisprudence) shifted its meaning. For Bentham it
was used to describe those aspects of law true of all nations. It was under universal jurisprudence that the censorial line, what we would now call normative jurisprudence,
that is of greatest significance, because it is here that
there is the greatest room for disquisitions that apply to
318
the circumstances of all nations alike.25 In other words,

universal jurisprudence was primarily normative, because
it was here that the (descriptive) differences between legal
systems become less significant. With Austin general jurisprudence has come to mean the analysis of the legal concepts that are found in all legal systems. As he put it, it was
principles abstracted from positive systems [that] are the
subject of general jurisprudence (Study, p. 350); or in
another formulation General Jurisprudence is the science concerned with the exposition of the principles, notions, and distinctions which are common to systems of
law (Study, p. 351). And as we have seen, after Hart the
term general jurisprudence has come to have an even narrower meaning, one that in fact excludes much of what
Austin considered as part of the subject in his lectures.
The Concept of Law, said Hart, was written for lawyers
and primarily had them in mind,26 and it made the internal point of view that of the lawyer who accepts legal
rules as reasons for action the central perspective of his
account. At the very same Harts approach to jurisprudence
showed little interest in many of the questions that lawyers
were interested in: investigation of legal concepts, the
boundary between law and politics, different approaches to
interpretation, the role of courts in society and much else.
The result has been a form of inquiry that seeks to explain
a social practice in the way it is understood from the eyes of
those who participate in it, but does so by drawing on the
methods of philosophy. (This point illustrates in another
way in what sense Hart attempted to do philosophy and sociology at the same time.)
When presented in this way it is not hard to see why this
effort was bound to fail right from the start: since lawyers
did not concern themselves much with the question of the
nature of law, then an account that sought to build on their
attitudes was bound to have little to work with and not be
25
26
Bentham, Introduction, supra note 10, p. 295.

Hart Interviewed, supra note 22, p. 291.
319
DAN PRIEL
very enlightening. And yet, this approach has prevailed and

has had enormous influence. (It is an interesting question,
but one that I cannot address here, why this happened.) As
a result of Harts inversion of normativity and validity and
the primary place he has given to the question of validity,
one can nowadays discern two ways in which the broadly
Hartian approach to legal philosophy has been further developed. One approach, concerned with an answer to the
question What is law?, focuses more on identifying the existence conditions of law by building on a positivist account
of legal validity. The other focuses more on the question of
normativity and attempts to improve upon Harts practice
theory of rules with an alternative positivist (i.e., one that
does not rely on morality) answer.
As a result we can distinguish between two versions of legal positivism, or as I would prefer to put it, two views on
the domain of legal positivism: one could be a positivist with
regard to validity, and one could be a positivist with regard
to the explanation of laws normativity. Unfortunately, while
some self-styled positivists focus more on validity and others on normativity, the two versions of positivism are rarely
treated separately. Perhaps this is because Hart never distinguished between the two domains, possibly because he
never clearly separated between the questions of validity
and normativity and tried to be a positivist in both domains. Be that as it may, it is important to see that they
may represent two different issues and it is not obviously
clear that they easily go together: as the example of Austin
shows in some ways the two issues may reflect two conflicting approaches to legal theory; and in a different way the
work of Joseph Raz also highlights the separation of these
issues, as Raz is a positivist on the question of validity but
a non-positivist on the question of normativity. The significance of this point goes well beyond the question of correctly classifying the ideas of various legal theorists as
positivist or not. I believe that the failure to see clearly the
distinctness of these issues has been detrimental to much
320
positivist work. Potentially valuable contributions to the

question of normativity are undermined by being embroiled
with unsuccessful attempts to defend legal positivism in the
domain of validity or, worse, with the fruitless (and hopeless) effort to answer the question What is law? as a conceptual question to which legal philosophers can meaningfully contribute.
V. CONCLUSION: HARTS FAUSTIAN BARGAIN
Elsewhere I argue that contemporary positivists appropriation of Thomas Hobbes and Jeremy Bentham to the
positivist fold is misleading.27 In this essay I have argued
that in some sense this is true even of Austin. Although
Austins work is clearly closer in spirit to that of contemporary legal positivists, there are important differences between his account of law and that of his followers. The
point, however, is broader and more general than Austin
exegesis. I said at the beginning of this essay that one of
the problems with contemporary jurisprudence is that there
seems to be little awareness of work in the field prior to
1961. I hope it is clear now why this is so. It is not because
there were no useful contributions by philosophers to
thinking about law before Hart. I assume that even Harts
followers would admit that he was not nearly as original as
were, say, Hobbes or Hume. But if there is today relatively
little interest at their work on law (compared with the
amounts dedicated to Hart in both writing and teaching), it
is because by contemporary standards they were not doing
what counts as legal philosophy. Their work was perhaps
an important stepping stone one that deserves an appreciative paragraph (or footnote), but not careful study in
the way to the creation of legal philosophy properly so
27 See Priel, Dan, Jurisprudence between Science and the Humanities, Washington University Jurisprudence Review, Vol. 4, (forthcoming 2012), available at
http://ssrn.com/abstract=1566858; Priel, Dan, Towards Classical Legal Positivism (unpublished manuscript), available at http://ssrn.com/abstract=1886517.
321
DAN PRIEL
called. It is for this reason that I said in the beginning that

in a sense contemporary legal philosophy is a new subject,
one that did not exist before The Concept of Law.
The transformation Hart brought to jurisprudence is this:
one can have a subject called legal philosophy as distinct
from both law and political philosophy, but it will be possible
to keep this subject alive only by keeping out artificially, as
not part of the subject, the kind of problems and questions
that challenge this demarcation. Thus, by definition, differences between legal systems that might presumably be shown
to be a challenge to the notion of general jurisprudence (in
its contemporary sense) are ruled out as the domain of sociology (and as such irrelevant); and the separation of legal philosophy from political philosophy again guarantees that any
attempt to show that the political foundations of legal systems result in fundamentally different concepts of law is also
declared off-limits and can thus be ignored.
All this comes at a price: legal philosophy is alive but it
owes whatever life it currently has to a kind of Faustian
bargain: Legal philosophy will continue to live as a distinct
discipline, but the price for being alive is that it will be so
separate from other disciplines to seem to be existing in its
own separate world. Academic lawyers, even those with theoretical inclinations and interests rarely see the point of
many jurisprudential debates. Legal philosophers, aware of
this, sometimes respond with a sneer: other academic lawyers are not smart and sophisticated enough, or sufficiently
well-read in philosophy, to understand the issues legal philosophers are dealing with.28
28 Most of these comments, naturally, do not make it to print. But see,

Coleman, Jules L. & Leiter, Brian, Determinacy, Objectivity, and Authority, in
Andrei Marmor (ed.), Law and Interpretation: Essays in Legal Philosophy, Oxford,
Clarendon Press, 1995, 203-78, pp. 204-05, where it can be glimpsed; see also
Waldron, Jeremy, Legal and Political Philosophy, Jules Coleman & Scott Shapiro
(eds.), The Oxford Handbook of Jurisprudence and the Philosophy of Law, Oxford,
Oxford University Press, 2002, 352-81, pp.374-75, who mentions (but does not endorse) these attitudes.
322
But the truth is that moral and political philosophers,

against whom it is much harder to make these claims, are
not much interested either and (this is something I have
personally heard from more than one of them) do not seem
to understand the point of what legal philosophers do.
For all these reasons my ultimate assessment of The Concept of Law and its legacy are largely negative. Is there nonetheless nothing to be salvaged from it? I pointed out earlier
that there are in fact two distinct strands that purport to follow on Harts work, one that focuses more on the question of
validity, the other that looks more to normativity. If legal
positivism has any plausibility or significance it is, I think,
only with regard to the second question. Harts positivist approach to the question of normativity is also seen where he
follows, explicitly albeit cautiously, on the footsteps of
Hobbes and Hume in attempting a positivist answer to
questions of normativity of morality (Concept, pp. 193-200).
It is these ideas, which sit rather uneasily with other parts of
the book, but which have the strongest ties with work of theorists now considered founders of legal positivism,29 that I
think (or at least hope) should prove the longest lasting contribution of Harts work in analytic jurisprudence.
29
Cf. Priel, Towards Classical Legal Positivism, supra note 27.
323
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
HARTIAN POSITIVISM AS A (PLAUSIBLE)

ERROR THEORY
Fbio Perin SHECAIRA*
Resumen:
Los crticos del positivismo jurdico hartiano han caracterizado este enfoque como una teora del error sobre el discurso judicial concretamente
sobre la argumentacin judicial en el contexto de desacuerdos relacionados con una metodologa de interpretacin apropiada. Los crticos tambin han sugerido que el positivismo hartiano, al igual que otras teoras
del error en otros campos de la filosofa, ha de ser tomada por falsa. El
propsito de este artculo es ofrecer una explicacin ms cuidadosa de la
naturaleza y alcance del error que los positivistas hartianos atribuyen a
los jueces. Se argumentar que el positivismo hartiano, aun cuando pudiera ser entendida como una forma de teora del error, no debera ser
relacionada con, por ejemplo, teoras del error en metatica, donde el
error relevante es omnicomprensivo y misteriosamente permanente.
Palabras clave:
Positivismo jurdico, teora del error, desacuerdo de interpretacin, metatica, H. L. A. Hart.
* Helpful comments on an earlier, longer version of this paper were provided

by Matt Grellette, Nick Griffin, Karl Laderoute, Stefan Sciaraffa, Noel Struchiner,
Wil Waluchow and participants in the PhD philosophy seminar at McMaster University. I am indebted to Heather Kuiper for comments on the current version.
325
FBIO PERIN SHECAIRA
Abstract:
Critics of Hartian legal positivism have characterized it as an error theory
about judicial discourse more precisely, about judicial argument in contexts of disagreement about proper interpretive methodology. The critics
have also suggested that Hartian positivism, like error theories in other areas of philosophy, is to be presumed false. The purpose of this paper is to
give a precise account of the nature and extent of the error assigned to
judges by Hartian positivists. It will be argued that Hartian positivism, even
if it can be understood as a form of error theory, should not be likened to,
say, error theories in metaethics, where the relevant error is pervasive and
mysteriously enduring.
Keywords:
Legal Positivism, Error Theory, Interpretive Disagreement,
Metaethics, H. L. A. Hart.
326
HARTIAN POSITIVISM AS A (PLAUSIBLE) ERROR THEORY
SUMMARY: I. Introduction. II. Problems with (P2). III. Why

Believe that Judicial Interpretation is Rule-Governed? IV. Conclusion.
I. INTRODUCTION
A number of legal philosophers (henceforth the critics) recently offered an objection to Hartian positivism (henceforth
HP) that deserves careful consideration.1 What follows is a
reconstruction of the argument grounding the critics objection:
(P1) Error theories are to be presumed false;
(P2) HP is an error theory;
(C) Therefore, HP is to be presumed false.
(P1) asserts what appears to be a well-established view in
other areas of philosophy, if not yet in jurisprudence: a theory about an entrenched type of discourse or mode argument that characterizes it as systematically flawed (e.g. as
resting on a false set of assumptions) is to be presumed
false. The presumption, of course, can be defeated; but the
burden of argument is assigned to the proponents of the error theory. I will not question the truth of (P1) in this paper;
(P2) will be the main target of my criticisms.
Why would HP be regarded as an error theory? An answer to this question requires a bit of background. HP is often associated with the claim that the content of the law
depends on official consensus. This claim is ambiguous because the phrase content of the law is ambiguous. H. L.
A. Hart clearly believed that the rule of recognition has its
content fixed by official consensus. And the rule of recogni1 These are some of the critics I have in mind: M. Berman, Constitutional
Theory and the Rule of Recognition: Toward a Fourth Theory of Law, in M. Adler
and K. Himma (ed.), The Rule of Recognition and the U.S. Constitution (New York: Oxford University Press, 2009); S. Sciaraffa, The Justificatory View and Theoretical
Disagreement (forthcoming in Problema. Anuario de Filosofia y Teoria del
Derecho); and S. Shapiro, What Is the Rule of Recognition (And Does It Exist)?,
in M. Adler and K. Himma (ed.), The Rule of Recognition and the U.S. Constitution
(New York: Oxford University Press, 2009).
327
FBIO PERIN SHECAIRA
tion has the undisputed function of specifying criteria that

are used to determine which norms are valid (i.e. binding)
within a legal system.2 It is a different and more controversial question whether Hart thought that the rule of recognition also serves the function of specifying proper methods
for interpreting norms which it recognizes as valid. So,
there are two ways of understanding the claim attributed to
HP, namely, that the content of the law is determined by official consensus. On one interpretation, official consensus
determines the norms which are binding within a given legal system. On the other interpretation, official consensus
determines not only what norms are binding within a legal
system but also the proper methods for interpreting binding
norms.
I am not sure which of the two interpretations of Harts
theory is correct. I will assume, for the sake of argument,
that the second interpretation is correct, since it is on that
interpretation that the critics seem to rely. This brings us
back to the question, why would HP be deemed and error
theory? If HP claims that the content of the law is fixed by
official consensus, then it follows that whenever officials
disagree either about criteria of validity or about proper interpretive methodology, the content of the law is indeterminate. But, the critics argue, judicial disagreement about
proper interpretive methodology is a common occurrence in
legal systems like that of the US. Judges disagree about
how to determine the meaning of valid legal norms (some
judges are textualists, some are intentionalists, others are
purposivists, etc.), and here comes the important part
they do so while arguing as if they were advancing claims
about what the law is or requires. Recall that, for HP, if
judges disagree about proper interpretive methodology,
then the law is indeterminate, i.e. it does not yet require
any specific course of action (unless, of course, different
methods yield identical interpretations; but there is no
2 H. L. A Hart, The Concept of Law (Oxford: Clarendon Press, 1961), Chapters 5
and 6.
328
guarantee that that should happen). So, defendants of HP

have no option but to characterize judicial discourse in circumstances of disagreement about proper interpretive
methodology as flawed: judges talk as if they were discerning what the law requires when, as a matter of fact, the law
makes no determinate requirements. HP is an error theory
insofar as it attributes to judges a mistaken belief (or perhaps a knowingly false belief, in which case judges would
not be misguided but insincere) that underlies judicial argument in circumstances of disagreement about proper
interpretive methodology.
I have two central aims in this paper. The first one is to
show that (P2) is unjustified. HP may indeed be characterized (loosely speaking) as an error theory, but the error it
identifies is neither widespread enough nor mysterious
enough to warrant a presumption that HP is false. For one
thing, the critics seem to ignore the significant risk of
skewed analysis generated by the so-called selection effect. In addition to that, even if we focus on the limited
sample of cases used by the critics, there is reason to doubt
the claim that judges unambiguously speak as if they were
setting out to discover what the law is in circumstances of
disagreement about proper interpretive methodology. My
second aim is to show why the defenders of HP have reason
to insist that judicial interpretive practice in modern legal
systems is rule-governed and that the rules in question are
a matter of official consensus. So, I will not only make a
negative case against the critics objection but also a positive case in favor of HP.
II. PROBLEMS WITH (P2)
1. How Pervasive and Mysterious is The Error?
As I stated in the introduction, I will not be questioning
the truth of (P1). But I need to say a few words about the
presumption that is raised against error theories and about
329
FBIO PERIN SHECAIRA
the conditions that need to be satisfied if that presumption

is to be defeated. The idea is that error theorists must explain the origin and persistence of the discourse which they
believe to be flawed: for why would people engage in and
perpetuate a way of speaking and arguing that is based on
assumptions whose falsity can be perceived by intelligent
individuals? Again, there is no question that error theorists
must be able to carry this burden if their view is to prevail.
But it should be noted that the weight of the burden is not
identical for all kinds of error theories. In metaethics, error
theorists recognize that they have a specially heavy burden
to carry because the error they assign is pervasive. The assumption that there are such things as mind-independent
yet action-guiding moral properties/facts is arguably central to ordinary moral discourse. If that assumption is false,
and error theorists argue that it is, then moral discourse is
fraught with error.
It is a hazardous thing, however, to compare HP to error
theories in metaethics. The error that HP assigns, I will argue, is limited to a select class of legal decisions and the explanation as to why flawed discourse persists within that
class is readily available to the proponent of HP. While error
theorists about moral discourse are concerned about explaining why most everybody is getting it wrong, proponents of HP are concerned about explaining why some
judges are getting it wrong in some very special cases. (I
will later suggest a further disanalogy between HP and error theories about moral discourse. While it is quite plausible to claim that ordinary moral discourse rests on realist
assumptions, it is much less plausible to claim that judicial
discourse within that select class of cases rests on the assumption that law is determinate. I put that to the side for
now and concentrate on what makes the relevant class of
cases special).
The critics tend to look for evidence of disagreement
about proper interpretive methodology in the decisions of
appellate courts. Decisions of the US Supreme Court are
330
among the critics favorites. It is surprising that the critics

pay little attention to the perils involved in relying on decisions of high-level courts when attempting to substantiate
general claims about the nature of judicial interpretive
practice. There is no question that the decisions of a court
like the US Supreme Court represent a small fraction of all
the decisions published by American courts. More important than mere numbers, however, is the fact that this
small fraction of cases is not selected randomly from the
entire universe of legal disputes. Political scientists and
philosophers have argued persuasively that the easiness
(technical and political) of a case serves as a disincentive to
litigation and appeal.3 Cases that reach high appellate
courts are very likely to be technically complicated and
highly politically charged (these features are not unrelated;
I will argue later that the political complexity of a case may
have an impact on its prima facie technical simplicity).
The first way in which HP differs from moral error theories, then, has to do with its frugal assignment of error to
judges deciding a relatively small class of cases which are
special in that they are legally and politically hard. The second difference between HP and moral error theories is
closely related to the first. For the nature of the cases in
which there is error is such that a plausible explanation of
the origin and persistence of error immediately suggests itself. Judges in high appellate courts tend to deal with cases
that attract public attention. The technical complexity of
hard cases is specially attractive to legal scholars; their political complexity is attractive to the media and hence to the
3 For a couple of important empirical studies, the first one being specially influential, see G. L. Priest & W. Klein, The Selection of Disputes for Litigation in
Journal of Legal Studies 13 (1984) and L. Lederman, Which Cases Go to Trial?: An
Empirical Study of Predictors of Failure to Settle, in Case Western Reserve Law Review 49 (1999). For philosophical papers containing helpful discussions of the
so-called selection effect, see F. Schauer, Easy Cases in Southern California Law
Review 58 (1985); F. Schauer, Judging in a Corner of the Law, in Southern California Law Review 61 (1988); L. Solum, On the Indeterminacy Crisis: Critiquing Critical Dogma, in University of Chicago Law Review 54 (1988); B. Leiter, Explaining
Theoretical Disagreement in The University of Chicago Law Review 76 (2009).
331
FBIO PERIN SHECAIRA
general public. High appellate judges decide under the

limelight, as it were, and behind that limelight there is a
public which associates policy-making with the legislature
and is reluctant to accept the democratic legitimacy of judicial activism. If judges are coy about disclosing the creative
nature of their decisions in contexts of disagreement about
interpretive methodology, the obvious reason for this is that
they are understandably seeking to avoid the flak which
overt activism tends to elicit their lack of sincerity is specially understandable because judges are normally not
given a choice; they must decide cases regarding which the
law is silent, and they must at the same time fulfill their
stereotypical role of law-appliers.
2. Is there Error at All?
Now, what I have said so far assumes that the critics are
right to claim that, at least within the limited class of cases
on which they focus, judges argue as if they were attempting to discern what the law is. This claim is crucial, for it is
what commits HP to the view that judges are in error when
they argue about the right solution to a legal problem in
circumstances of disagreement about proper interpretive
methodology. But the claim is dubious, if not utterly false.
Judicial discourse at the high appellate level is quite elusive, and it is not at all clear that judges argue as if there
were trying to discover what the law is. Judges often cumulate doctrinal with substantive (moral, political, economic)
arguments in what has been derisively described as a
shotgun fashion: the judge will canvass all possible arguments in support of a position, repeat them for emphasis,
and present them all without any regard for how they actually hang together as a coherent, principled position.4 To
be sure, often judges will be a bit more careful about how
4
B. Leiter, supra note 3, at 1233. Leiter convincingly argues that such a strategy was employed by the majority in Riggs v. Palmer 22 NE 188 (NY 1889) precisely one of the cases which has been taken to present difficulties for HP.
332
they argue and will attempt to show that the substantive

considerations to which they appeal are connected in important ways to the relevant formal rules: a substantive
consideration may be presented as the rationale that
underlies a formal rule or as the principle that unifies a
body of formal rules.
None of this proves that judges in circumstances of disagreement about proper interpretive methodology do not argue as if they were trying to discern what the law is. The
fact that judges rarely appeal exclusively to substantive
considerations (as opposed to formal rules) suggests that
judges indeed see themselves as being bound to some extent by rules which they have not created themselves. On
the other hand, high appellate judges do frequently appeal
to considerations of justice, utility, efficiency, etc. which
cannot uncontroversially be characterized as parts of
non-discretionary attempts to discern what the existing law
requires. My point, once more, is that judicial discourse in
hard cases is elusive. Formal and substantive arguments
appear side by side and their connections are not always
clear.
Note that, for a proponent of HP, this should not come
as a surprise. If a judges predicament involves juggling the
obligation to decide a hard case and the obligation to fulfill
the role of law-applier, one should expect the judges discourse to reflect the tensions which afflict his reasoning.
Judicial discourse is ambiguous because judges are subject
to conflicting professional requirements.
III. WHY BELIEVE THAT JUDICIAL INTERPRETATION
IS RULE-GOVERNED?
The foregoing considerations suggest the following picture. If HP assigns any error it assigns it to those judges
who unambiguously argue in terms of law discovery and
application (and not very many judges do that, I surmise) in
circumstances of disagreement about proper interpretive
333
FBIO PERIN SHECAIRA
methodology. This kind of disagreement does not appear to

be widespread; it tends to occur in hard cases which tend
to attract more public attention than the average case. Any
simple comparison between HP and error theories in
metaethics (theories which assign systematic error whose
explanation is not nearly as straightforward)5 is rendered
very suspicious. Is the limited and well explained error assigned by HP to judges capable of motivating a significant
presumption against its soundness? Not to my mind.
In this section I will offer some general statements about
judicial interpretive practice that will serve a twofold function. First, they will help to show that proponents of HP
have good reason to believe that judicial interpretive practice is governed by (largely tacit) interpretive rules. Second,
once these statements are made I will be able to reinforce
the point of the previous sections by showing that some alleged examples of disagreement about proper interpretive
methodology are not in fact instances of this kind of disagreement. The limited amount of examples on which the
critics rely is probably even more limited than one may
think. A brief examination of Riggs v Palmer will help me to
establish this point.
What evidence is there that judicial interpretive practice
is rule-governed? A good part of that evidence comes from
an influential collection of essays examining the interpretive practices of high-level courts in nine different jurisdictions.6 The essays point to the conclusion that judges abide
by a broad rule (or perhaps standard, given its relative
vagueness) that creates a strong presumption in favor of
the plain meaning of formal rules.7 Where the plain mean5 Richard Joyce, for instance, a contemporary error theorist, argues that our
erroneous belief in the categorical nature of moral reasons has to be explained by
reference to our evolutionary history. The evolutionary story he tells is quite complicated and occupies a significant portion of his recent book, R. Joyce, The Evolution of Morality (Cambridge: MIT Press, 2006).
6 N. MacCormick and R. Summers (eds.), Interpreting Statutes: A Comparative
Study (Aldershot: Dartmouth, 1991).
7 The authors focus on statutory rules.
334
ing of a rule is not vague, ambiguous or otherwise indeterminate, judges will only deviate from it if they can provide
strong arguments based on the supposed intentions of the
authors of the rule or on pressing considerations of policy
and principle that more often than not are presented as describing the underlying rationale of the very rule. The interpretive rule guiding this practice might be formulated thus:
Plain meaning (when it is clear) is to be applied unless the results of so doing are very unsatisfactory (from a substantive
point of view) or unless legislative intention is clearly at odds
with plain meaning. This rule clearly has a penumbra of uncertainty (how unsatisfactory must the results be in order
to qualify as very unsatisfactory?) but it also has a core of
clear meaning and thus is capable of disposing of a large
array of cases.8
I will now turn to a particular case, the famous Riggs v
Palmer, in order to provide more focused evidence for the
claim that judicial interpretive practice is rule-governed in
legal systems like that of the US. Riggs will also be the
source of an important lesson about how to identify genuine instances of disagreement about interpretive methodology. In Riggs, the Court of Appeals of New York did not allow the defendant, Palmer, to inherit under the will of his
grandfather, whom he had murdered. The court made this
decision in spite of the fact that the plain meaning of the
relevant statutes did not invalidate the grandfathers will.
Plain statutory meaning, in this case, was arguably displaced by the substantive principle that no man may profit
from his own wrong. There are two interesting facts about
Riggs. One is simply that plain meaning was discarded, a
fact which may motivate suspicion about my claim that
judges share a strong presumption in favor of plain meaning. The second fact is that there was a dissent in which
8 Note that the vagueness of this formulation of the rule is partly a function of
the fact that it purports to encompass the subtly different practices that occur in
all nine jurisdictions. The practices of each jurisdiction can be described somewhat more precisely.
335
FBIO PERIN SHECAIRA
the idea of discarding plain meaning was contested: this

fact motivates the idea that Riggs involved judicial disagreement about proper interpretive methodology. I argue in
what follows that Riggs does not provide evidence against
the presumption of plain meaning and also that Riggs did
not involve disagreement about proper interpretive methodology. I will make these arguments in turn. Consider what
Frederick Schauer has said about Riggs:
We can find numerous examples of courts allowing killers to
take property that became available to them solely because
of their own culpable actions, including cases involving a killer of the testator who was found not guilty by reason of insanity, a killer of the testator who was convicted of voluntary
manslaughter, murderers whose acts of murder caused property to pass to their children although not directly to themselves, a murderer convicted of being an accessory after the
fact but not of actually wielding the murder weapon, a murderer who did not kill a testator but instead as remainderman killed the holder of the life estate... In all of these cases,
all falling only slightly short of first and second degree murder, courts have allowed culpable killers to inherit, and have
treated the Riggs v. Palmer principle, whether embodied in a
statute or in the common law, as an exception to be construed narrowly, notwithstanding the broad potential implications of the no man may profit from his own wrong principle.9
Again, focusing too narrowly on a case or small group

thereof can lead to a skewed analysis. Once we take a step
back and consider a sequence of cases that are quite similar to Riggs, an interesting pattern suggests itself. It is true
that the judges in Riggs were willing to deviate from the
plain meaning of the applicable statutes. But other judges
in similar cases were not so willing. This lends credibility to
the view that, instead of casually being ignored by the majority in Riggs, the statutes were found to make a prescrip9
F. Schauer, The Limited Domain of Law, in Virginia Law Review 90 (2004),
1937-1938 (footnotes omitted).
336
tion that was morally too objectionable to be enforced.

Heirs often do rotten things and are allowed to get away
with it, but murder, the judges believed, cannot be treated
so leniently.
Let me consider now the dissent in Riggs. Judge Gray
complained about the opinion of the majority in the following terms: ... the legislature has, by its enactments, prescribed exactly when and how wills may be made, altered
and revoked, and, apparently, as it seems to me, when they
have been fully complied with, has left no room for the exercise of an equitable jurisdiction by courts over such matters.10 It might appear, thus, that Gray is a committed textualist while the majority makes room for the exercise of
equitable jurisdiction in certain cases. But this is dubious; Grays supposed textualism is abandoned in a decision
handed down on the very same day as Riggs. In Bockes v
Temple, Gray claims: It is an elementary rule that statutes
are to be interpreted according to their intent. The intention
of the legislature is undoubtedly the great principle which
controls the office of interpretation...; and he qualifies that
by saying that [i]t is only where the literal acceptation of
the words used will work a mischief, or some absurd result,
or where some obscurity in the sense compels it, that we
need resort to extrinsic aids of interpretation.11
So, did Gray have inconsistent views about legal interpretation? Not necessarily. Note that in Riggs Gray affirmed
that public policy did not mandate a ruling against
Palmer, since he was already being punished for his crime
(22 NE 188 (NY 1889) at 519). It appears, then, that Gray
did not see the result of applying the plain meaning of the
statutes in Riggs as mischievous or absurd. Otherwise, he
might have been willing to join the majority, for his decision
in Bockes v Temple identified mischief and absurdity as sufficient conditions for the use of extrinsic aids of interpreta-
10
22 NE 188 (NY 1889) at 519.
11
Cited in Leiter, supra note 3, at 1433.
337
FBIO PERIN SHECAIRA
tion (which presumably includes intention and perhaps

also moral considerations).
What does this analysis of Grays decision teach us?
First, it provides additional evidence in support of the view
that judicial discourse can elude if analyzed superficially.
One who reads Riggs and nothing else may come away with
the mistaken impression that Gray was a hard core textualist. Second, and relatedly, it shows how the critics should
not be too quick to classify judicial disagreements like the
one in Riggs as instances of disagreement about proper interpretive methodology. It is just as plausible to say that
the judges in Riggs were all abiding by the rule that gives
precedence to plain meaning but allows for its defeat in exceptional circumstances. That is to say, the judges in Riggs
were not employing different interpretive strategies but
were disagreeing about how to apply a shared interpretive
strategy in a particular case. How is this possible? Because
the interpretive rule that governs judicial practice has a
penumbra of uncertainty: it will not always be uncontroversial whether the conditions for the defeat of plain meaning
have been satisfied.
IV. CONCLUSION
Allow me to summarize my argument. First, consider its
negative aspect. Error theories should indeed (I concede
without argument) be presumed false. But this statement
refers to theories that assign systematic error whose origin
and persistence is not easily explained. HP is hardly an error theory in this sense. It assigns errors to those (few)
judges who speak unambiguously in terms of law discovery
in cases that are exceptionally complicated and attract an
inordinate amount of attention.
As to the positive aspect of my argument, there is evidence, provided by analyses that are much more systematic
than the critics, indicating that judicial interpretive behavior is rule-governed. The broad rule in question has a pen338
umbra of uncertainty and it is likely that some of the putative examples of disagreement about interpretive methodology provided by the critics are in fact instances of disagreement that arise under the rules penumbra. Recall
Harts famous example of a rule that did not permit vehicles
in a public park. Disagreement might arise as to whether a
bicycle is an instance of a vehicle and hence whether it
should be permitted in the park. This kind of disagreement
does not concern the critics, for it is not disagreement
about proper interpretive methodology. All parties to the debate about bicycles may be convinced that the plain meaning of the rule should be enforced. The problem is that vehicle is a vague term and that bicycle is a borderline case.
Similarly, in Riggs, the judges were probably not really disagreeing (although their rhetoric may suggest otherwise)
about what interpretive method to apply. Instead, they disagreed about the application of a second-order interpretive
rule with fuzzy edges. The question was simply whether
Palmers behavior was rotten enough to allow for the defeat
of plain meaning.
339
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
HART ON JUDICIAL DISCRETION

Roger A. SHINER1
Resumen:
El concepto de derecho (1994) de H. L. A. Hart, contiene muchos pasajes
que se han convertido en conos de la teora jurdica. Este artculo se
concentra en el captulo 7, seccin 1 y 2, y en los comentarios que realizara Hart sobre la discrecin judicial en el contexto de la bien conocida
crtica de Dworkin a esta misma idea en su ensayo El modelo de las
normas. Especficamente, el artculo emprende tres proyectos. El primero de ellos consiste en defender la importancia del esquema fundamental
que ofrece Hart en el captulo 7 de su obra, esto es, la creacin judicial
de normas en el Common Law. Hart representa tal creacin de normas
como un balance de certeza y flexibilidad; y l est en lo correcto al presentarlo de tal modo. El segundo proyecto consiste en argumentar que la
crtica de Dworkin al modelo positivista de la creacin judicial de normas
en el Common Law, como un ejercicio de discrecin fuerte, no est sustentada. La idea central para el significado de discrecin fuerte de
que los tribunales no estn sujetos a criterios establecidos por la autoridad en cuestin no puede ser establecida. El tercer proyecto consiste en
argumentar que Hart es su peor enemigo. El lenguaje, las metforas e
imgenes que emplea para exponer su teora de la creacin judicial de
normas en el Common Law abre la puerta a la crtica dworkiniana. Si tomamos en serio el lenguaje, las metforas y las imgenes, encontraremos a un Hart del tipo formalista o deductivista sobre la adjudicacin,
justo como el que l mismo est abiertamente cuestionando en el captulo 7 de El concepto.
University of British Columbia Okanagan and Okanagan College.
341
ROGER A. SHINER
Palabras clave:
Discrecin judicial, Common Law, creacin judicial de normas, positivismo jurdico, escepticismo ante la regla, formalismo, textura abierta, autoridad, certeza, flexibilidad, H. L. A.
Hart, R. Dworkin.
Abstract:
H. L. A. Harts The Concept of Law (Hart 1994) contains many passages
that have become iconic for legal theory. This essay focuses on Chapter 7,
sections 1 and 2, and Harts comments about judicial discretion in the context of Ronald Dworkins well-known attack on the idea of judicial discretion in his essay The Model of Rules. Specifically, the paper undertakes
three projects. The first project is to defend the importance of the fundamental picture that Hart presents in Concept, Chapter 7 of Common-Law
judicial rule-making. Hart represents such rule-making as a balance of certainty and flexibility, and he is correct to do that. The second project is to
argue that Dworkins attack on the positivist model of common-law judicial
rule-making as an exercise of strong discretion fails. The idea, central to
the meaning of strong discretion that courts are not not bound by standards set by the authority in question cannot be established. The third is
to argue that Hart is his own worst enemy. The language, metaphors and
images he uses to present his account of common-law judicial rule-making
open the way to Dworkins critique. They also reveal Hart, if the language,
metaphors and images are taken seriously, to be precisely the kind of
formalist or deductivist about adjudication that he is ostensibly in Concept
Chapter 7 criticising.
Keywords:
Judicial Discretion, Common Law, Judicial Rule-Making, Legal
Positivism, Rule-Scepticism, Formalism, Open Texture, Authority, Certainty, Flexibility, H. L. A. Hart, R. Dworkin.
342
SUMMARY: I. Introduction. II. Dworkin and Strong Discretion.

III. The Importance of Harts Account of Common-Law Judicial Rule-Making. IV. Hart Versus
Hart: Strong Discretion after All? V. References.
I. INTRODUCTION
H. L. A. Harts The Concept of Law (Hart 1994) contains
many passages that have become iconic for legal theory. In
this essay, I want to focus on Chapter 7, sections 1 and 2,
and Harts comments about judicial discretion. I also want
to revisit and review Ronald Dworkins well-known attack
on the idea of judicial discretion in his essay The Model of
Rules (Dworkin 1978, Chapter 2). Specifically, I have three
projects in this paper. The first is to defend the importance
of the fundamental picture that Hart presents in Concept,
Chapter 7 of common-law judicial rule-making. The second
is to argue that Dworkins attack on the positivist model of
common-law judicial rule-making as an exercise of strong
discretion fails. The third is to argue that Hart is his own
worst enemy. The language, metaphors and images he uses
to present his account of common-law judicial rule-making
open the way to Dworkins critique. They also reveal Hart, if
the language, metaphors and images are taken seriously, to
be precisely the kind of formalist or deductivist about adjudication that he is ostensibly in Concept Chapter 7
criticising.
Let me begin with a reminder of the passages in Concept I
have in mind. Famously, in the context of what Hart identifies as our twin handicaps of relative ignorance of fact
and relative indeterminacy of aim (128) and of the open texture of legal language, he speaks of the courts as
rule-making authorities who exercise a discretion (132),
of the law as being developed by the courts (135). He says
that at the margin of rules and in the fields left open by
the theory of precedents, the courts perform a rule-producing function (135) that is also referred to as a creative
343
ROGER A. SHINER
function (136). Then there is the following section on

rule-scepticism, where Hart distinguishes between how
laws function and the game of scorers discretion, not by
denying that in adjudication there is any judicial discretion
at all, but by asserting it falls short of turning adjudication
into scorers discretion. In the Postscript published in the
second edition of Concept (Hart 1994), Hart continues to
emphasize that the law is indeterminate or incomplete,
and that judges exercise a limited law-creating discretion
(272), and he refers to legally unregulated cases which
constitute gaps in the law (252, 272) which are filled by
the exercise of a law-creating discretion (252). Judging by
the tenor of his remarks in the final section of the Postscript, Hart remained fully convinced of the correctness of
his analysis of adjudication as containing this discretionary
aspect. As said, I will assess the justification for Harts
confidence later. Let me begin now with Dworkins critique.
II. DWORKIN AND STRONG DISCRETION
Dworkins equally famous critique of the doctrine of judicial discretion found in his frontal attack on legal positivism in The Model of Rules seems to me under-theorized.
Again, I will briefly remind us of the details. His
broad-brush characterization of positivism (Model, 17) involves reducing positivism to three key tenets. The term
discretion appears twice in the characterization in the
second tenet, according to positivism, when a case does
not fall clearly under a rule, the judge or other legal official
decides the case by exercising discretion, by reaching beyond the law for some other sort of standard: and in the
third tenet, positivism is alleged to believe that, when the
judge does so reach beyond the law and exercise discretion,
he is not enforcing a legal right. Later in the essay,
Dworkin supplements these comments with his distinction
between three senses of discretion the first weak
sense, according to which to say that a decision requires
344
the use of discretion is to say no more than that the decision requires the use of judgment, not of the mechanical
application of a standard: the second weak sense, according to which to say that a decision requires the use of discretion is to say no more than that the decision the decision-maker takes is not subject to review and possible
reversal by any other official: the third, or strong, sense,
according to which to say that a decision requires the use
of discretion is to say that the decision-maker is simply
not bound by standards set by the authority in question
(Model, 32), or that the decision-makers decision is not
controlled by a standard furnished by the particular authority we have in mind when we raise the question of discretion (Model, 33).
I lay aside here the second of the weak senses: it plays
no role in this essay. I am concerned about the interplay
between the first weak sense and the strong sense.2
Dworkins claim, as we know, is that when positivists talk
about discretion, they have to be using the term in the
third, strong sense, and that as a result their theory must
fall.
The Meaning of Weak and Strong
Dworkins imagery of weak and strong is notably different from Harts imagery. When he is not using the plain
terminology of regulated and unregulated disputes,
Harts imagery is essentially spatial. He uses expressions
like gaps in the law, interstitial powers, open texture,
areas of conduct, area of open texture and judicial activity within it, at the margins of rules and in the fields left
2 I am going to stop now using the inverted commas around weak and
strong, because it will become tedious. But I do want to emphasize that these
terms are Dworkins creation. It must not be supposed that he is drawing on some
previously established or accepted use of these terms in English that has the
meaning he attributes to it. Somewhat the same might be said of Harts use of discretion as a technical term, although his use of it is far closer to the ordinary use
of the word in English.
345
ROGER A. SHINER
open by precedent, a large and important field left open,

nothing to circumscribe the area of open texture, core
and penumbra, within a rule, lying on the border of a
rule, and so on. This imagery is not entirely consistent internally. Most are compatible with Dworkins notorious
comment that discretion is like the hole in a doughnut,
[not existing] except as an area left open by a surrounding
belt of restriction (Model, 31) the comments about gaps
in the law, interstitial powers and the like. But the idea
of a core of certainty surrounded by a penumbra of
uncertainty where discretion is exercised is different.
Dworkins imagery of weak and strong, on the other
hand, is not spatial imagery, but rather draws on the idea
of natural forces. The wind may be blowing weakly or
strongly: the current in a river may be flowing weakly or
strongly: the undertow on the shore may be weak or strong.
By easy extension, persons may be physically weak or
strong according to how much force they are able to exercise. By a further easy extension peoples character may be
described as weak or strong accorded to how easily they are
able to apply mental or emotional force to resist temptation.
In philosophy we are familiar with other easy extensions:
an argument is weak or strong according to the degree of
support it offers for a conclusion; a theory may be weak if it
is too easily confirmed, or strong if it is hard to confirm but
confirmed nonetheless. However, how do we get from there
to the idea of discretion being weak or strong? Dworkin
himself is no help with this. He baldly introduces the terminology by stipulation without giving any account of why the
terms weak and strong might be appropriate. For example, indeed we may well speak of a decision-maker having
discretion when the decision they have to take demands the
use of judgment. But what about that makes it appropriate
to call this kind of discretion weak discretion? Dworkin
does not say. It certainly is not that in these cases the decision-maker has to apply less mental force or is subject to
less rational or emotional pressure from outside.
346
I can offer a plausible reconstruction. Fred Schauer writing a couple of decades or more after Dworkin points out
that rules function as devices for the allocation of power,
and that according to the forms of language in which rules
are expressed, rules may allocate more or less power
(Schauer 1991, 15862). Schauers thought is this. Imagine
one person or body of persons wanting to control the behavior of another person or body of persons. Suppose the
would-be controller expresses their wish for the would-be
controllee to comply in vague and general language: a parent might say to a child, Go out and have fun: Ill see you
later. The parent hands over virtually all control over the
future behavior of the child to the child: the parents rule
allocates almost all power to the child. On the other hand,
the parent might say something very specific: all right, go
and play with your friends: but dont go further than the
end of the road, dont talk to strangers, be back by 5.00
pm, and phone me every hour so that I know you are all
right. Here the parents rule allocates most of the power to
the parent. Likewise, in the case of a legislative body delegating power to an administrative body, the legislature
might pass a very precise body of rules and regulations and
charge the administrative body with the application of these
rules exactly as written. In such a case, most of the decision-making power rests still with the legislature. Or the
legislature might pass a body of rules that is full of terms
like fair, reasonable, due and so on, terms which leave
a lot of work to be done by the administrative body applying
them to determine exactly how they are to be applied in
particular cases. Here the legislature is essentially handing
over a lot of power to determine outcomes to the administrative body. This structure can be linked to weak and
strong discretion, especially if we consider Dworkins examples (Model, 32). If a lieutenant orders a sergeant to take
the five most experienced men on patrol, the discretion
given to the sergeant is weak in that most of the power to
determine who goes on patrol rests with the lieutenant. If
347
ROGER A. SHINER
the lieutenant tells the sergeant to pick any five men for
patrol he chooses, the discretion given to the sergeant is
strong in that all of the power to determine who goes on
patrol rests with the sergeant.3
This interpretation of weak and strong might seem to
bring Dworkins schema into line with a different spatial
image used by Hart in speaking of judicial discretion
where he speaks of courts or judges discretion being
narrow or wide (cf. Concept, 135, 273). He is speaking
primarily of judicial law-making in the domain of common-law precedent, where the terminology of narrowing
or widening is well established. Actual cases involve actual fact-situations, and necessarily only some of the features of any given fact-situation will be material to the legal
issue involved. A court might decide that some given feature of the fact-situation of a case is no longer to be one
that is part of the set of features that create the legal rule
for which the case stands. In such an instance, the court is
said to narrow the rule extracted from the precedent case.
Or the court might decide to add into the set a feature that
was not previously considered part of the set: then the
court is said to widen the rule for which the precedent
stands. Such narrowing or widening will have an effect on
the power of future courts to control how the precedent affects the particular case that has fallen to them to decide.
So one might think that when a court narrows a precedent
that it automatically weakens the discretionary decision-making powers of future courts, and when it widens a
3 Incidentally, one of Dworkins examples is very odd. He imagines a dog show
judge who has discretion to judge airedales before boxers if the rules do not stipulate an order of events. But that isnt going to happen. In North America, anyway,
airedales would not be judged alongside boxers, as they belong to different groups.
Airedales are in the Terrier group, and boxers the Working group. Within groups,
breeds are judged alphabetically. Even if an airedale won the Terrier group and a
boxer the Working group, in the final Best of Show group again the order of judging
groups is specified. So Dworkin has picked an example which is guaranteed to be
counterfactual. Over things like this dog show judges have no discretion: the
American Kennel Club rules spell it all out. Of course, in judging this or that dog to
be the best boxer, say, certainly weak discretion would be involved.
348
precedent it strengthens such powers. But that is not necessarily so. The effect of narrowing a precedent might just
as easily be to leave more room for discretionary decision-making by future courts, and the effect of widening a
precedent to leave less room. It is not possible to decide
what is actually the case in the abstract. Everything will
turn on the specifics of the case in question and the area of
law in question.
Nonetheless, even if we can in this way reconstruct a
plausible meaning for the terminology of strong and weak
discretion, this does not further Dworkins critique of positivism. For one thing, the home for the terminology we have
found applies only in one narrow area of adjudication, a
specific established practice of courts in relation to reasoning from precedent. Dworkin clearly intends the object of
his critique to be some alleged feature of adjudication tout
court. Moreover, the terminology of narrow and wide applies to the structure of the verdict itself in a case, not to
the process of reasoning by which the verdict was reached.
There is no necessary connection between a verdict containing a widening of a precedent and the process of arriving at that verdict being one of the exercise of strong discretion, nor is there any connection between a verdict
containing a narrowing of a precedent and the process of
arriving at that verdict being one of the exercise of weak
discretion. The process in both instances is, if it is, a matter of strong or weak discretion for entirely different reasons from those for which the verdict would be called a
widening or a narrowing of precedent or precedents
before the court.
Not Bound by Standards Set by the Authority in Question
Whatever we can make of the terminology of weak and
strong in general, Dworkins use of the terms depends
crucially on the phrase I have highlighted as the title of this
subsection. It seems to me that this phrase stipulates what
349
ROGER A. SHINER
is to be meant by the expression strong discretion, rather

than that the phrase articulates some antecedently existing
meaning of the expression. So what is it for a decision-maker to be not bound by standards set by the authority in question? Dworkins example looks straightforward. The lieutenant says to the sergeant, Pick any five
men for patrol you choose. The lieutenant is in the military
hierarchy the authority in question: the sergeant has
stringent duties of obedience to the lieutenants commands.
The lieutenants command leaves it entirely up to the sergeant whom he actually picks to go on patrol: the sergeant
is not bound by any standards set therefore.4 However, if
this is the paradigm example of strong discretion, how do
we get from there to the idea that when judges or courts exercise discretion in the manner in which Hart urges that
they do, they are in the position of the sergeant being told
by the lieutenant to pick any five men he wishes to go on
patrol? It is not so clear what the way is.
First, it is clear that the military example works so well
because the military is the perfect example of what Joseph
Raz has called an institutionalized normative system (Raz
1975, Chapter 4). The norms of military behaviour are formally determined by the institution of the military, and
they are interlinked so as to form a system. The norms in
their content place much emphasis on hierarchy and obedience: they create clear and precise command structures in
that way. Legal systems are also institutionalized normative
systems, and within the typical common-law legal system
certainly there are hierarchies of a kind. If you take the typical triadic structure of trial court, appeal court and supreme court, trial courts are bound by the decisions of appeal courts and the supreme court, appeal courts by
decisions of the supreme court. When statute and legisla4 Dworkin points out, as we know, that strong discretion does not amount to
license. There are relevant certain standards of rationality, fairness and effectiveness (Model, 33: cf. also Hart, Concept 128-9, and his comments about reasonable social aims). I assume this qualification by Dworkin is unproblematic and applicable.
350
tion are concerned, courts are bound by decisions of the

legislature that create laws. Administrative tribunals may
exercise delegated powers, but they do so within limits determined by the legislature. So far so good, then: it looks
like the military paradigm may have applicability to the
case of legal decision-making by courts.
I believe, though, that that is as far as the parallel can go
namely, the case of one part of the legal system being
clearly subordinate by the rules or conventions of the system to some other part of the system. So in Canada a provincial court of appeal might be like a lieutenant, a trial
court like a sergeant. The Supreme Court, or Parliament,
might be like a colonel or a general within a hierarchical
structure. The problem is that the cases that Hart is most
concerned with do not have this hierarchical character to
them. Hart focuses, as I have noted, a lot on the role of
courts in creating precedents for later courts at the same
level to follow. The British Columbia Court of Appeal in
2005, say, sets a precedent that the British Columbia Court
of Appeal in 2011 has to respect. The 2005 Court affects
the decision-making activities of the 2011 Court. The Supreme Court of Canada in 2011 reasons by taking into account previous decisions of the Supreme Court of Canada.
There is no lieutenant/sergeant-type hierarchy here.
To see the implications of the above, consider another
one of Dworkins examples the boxing referee who is required by the rules of the sport of boxing to award a round,
or a match, to the fighter who has been the most aggressive
(Model, 32). This referee, Dworkin says, is in the same position as the sergeant told to pick the five most experienced
men. The rules of the sport constrain his judgment: his discretion is weak, not strong. Now, we could make this example fit the lieutenant/sergeant paradigm exactly, if we
think, not of the rules of boxing in the abstract, but of the
Rule Committee of the Canadian Amateur Boxing Association, or of the World Boxing Association as being the relevant decision-makers who create the constraint on the ref351
ROGER A. SHINER
eree. But this would be forced: the rules themselves have a

life as social rules independent of the moment of their creation by a Rule Committee. So in a sense it looks like we
have a model here for how the law itself, rather than some
specific legal body, might plausibly be said to create or affect the decision-making of a court or a judge. The rules
themselves, rather than some body of rule-makers, occupy
the position of the lieutenant in the analogy.
Nonetheless, law is still different. The rules of precedent,
say, are part of the law: but so also are the judges and
courts deciding cases according to precedent, and the legislatures rule-making by legislation. The legal system comprises all of these activities and agents: they exist all on the
same plane of the hierarchy. Dworkins account of strong
discretion supposes that the place of the authority setting
the standards that affect the discretion of judges is occupied by the law. If there is law, courts discretion can at best
be weak: if there is no law, then the courts discretion is
strong, because there are no standards set by the relevant
authority. However, as Harts analysis in Concept, Chapter
7 makes very clear, the practice of legal decision-making
according to precedent, and the consequent law-making
function that courts can and do perform, is an on-going
part of the legal enterprise, at least in common law systems. It does not make sense to distinguish between the
law as the relevant authority and the courts as norm-subjects of this authority, because the courts and the practice
of courts is part and parcel of the law. The paradigm of the
lieutenant and sergeant and their hierarchical relation
breaks down at precisely the point where Dworkin needs it
in order to make sense of his concept of strong discretion.
In short, Dworkins concept of strong discretion is little
more than an arbitrary and question-begging construct.
The image of strong and weak cannot be rooted in any
plausible non-imagistic account of legal practice. The examples Dworkin gives to illustrate the notion of strong discretion fail to fit legal practice. While it is true that Hart em352
phasizes the existence of judicial discretion, no grounds

have been given by Dworkin for thinking such discretion
must be strong in Dworkins technical sense.
III. THE IMPORTANCE OF HARTS ACCOUNT OF COMMON-LAW
JUDICIAL RULE-MAKING
For several years now, in teaching undergraduate philosophy of law classes, I have conducted the following experiment. I have given the class the official definitions of
strong and weak discretion (though without saying what
Dworkin is doing with the terms), and then had them look
at the relevant passages in Chapter 7 where Hart talks
about judicial discretion. I have then asked the students
which kind of discretion do they think Hart is talking
about? Inevitably the class is split between strong and
weak discretion, with a tendency to favour the latter. It is
far from intuitive that Hart is committed by the remarks
made in Chapter 7 about judicial decision-making to judicial discretion being strong. Suppose, then, that, despite
all the talk about gaps in the law, open texture, interstitial
judicial powers and the like, the correct way to think about
the discretion that Hart ascribes to judges is that it is a
form of weak discretion. It is a scalar form of discretion
it may be narrow or wide, but never strong.
I think as a matter of fact such an interpretation is faithful to the fundamental picture that Hart is trying to present
in Concept Chapter 7 of common-law judicial rule-making.
While he does urge us to take note of a function of courts
properly characterized as rule-making, he is just as insistent that this rule-making function is limited: the life of
the law consists to a very large extent in the guidance both
of officials and private individuals by determinate rules ...
which do not require from them a fresh judgment from case
to case (Concept, 135: his emphasis). That is, the other
cases the ones where rule-making occurs are ones
that do involve fresh judgment, just as Dworkins notion
353
ROGER A. SHINER
of weak discretion implies. As I have argued elsewhere,5

precedential reasoning in common law courts really does
consist to a form of reasoning that is genuinely reasoning
while not being mechanical or deductive reasoning. Courts
have a duty to take into account prior legal materials: it is
absurd to suggest they are not bound by standards set by
the authority concerned. On the other hand, those legal
materials do not necessitate one specific verdict: a court
must exercise judgment, and does so, as already noted, by
narrowing, widening or simply following precedent. Hart
very clearly in Concept Chapter 7 is trying to draw the
attention of the theorist to these straightforward facts.
The question must then be faced, however, of what effect
the acknowledgment that common-law judicial rule-making
involves weak, not strong, discretion will have on Dworkins
critique of legal positivism. Will Dworkin still be able to argue that positivism is fundamentally mistaken even though
he fails to establish securely a necessary commitment by
positivism to strong judicial discretion? Dworkins own argument on this point is remarkably tendentious: The proposition that when no clear rule is available discretion in the
sense of judgment must be used is a tautology. It has no
bearing, moreover, on the problem of how to account for legal principles (Model, 34). The positivists, he continues,
speak as if their doctrine of judicial discretion is an insight
rather than a tautology. These remarks amount to little
more than a complaint that positivism is not in fact as stupid as he needs it to be in order to refute it. It seems to me
extremely unlikely that Hart is taking his remarks about
discretion to be profound theoretical insights. Recall the famous comment in the Preface to Concept that it is to be
considered a work in descriptive sociology. Philosophy frequently proceeds by the tactic of describing the familiar,
when clouds of theory have caused us to lose sight of the
familiar. It is clear in Chapter 7 that some of Harts main
philosophical opponents are the rule-sceptics, those who
5
See Shiner 1992, 837; Shiner 2005, 4456.
354
deny any certainty anywhere in the legal system. How can

such views be better countered than by simple straightforward reminders of just how much certainty there is in legal
rule-making and rule-following? Others of Harts opponents
are the formalists, those who find complete certainty everywhere in the legal system. How can such views be better
countered than by simple straightforward reminders of just
how much flexibility there is in legal rule-making and rulefollowing?
Hart though sees that there is something about how adjudication in the common law operates that gives rise to
rule-scepticism, and that a proper theorization of the common law will bring this something out. The something is
the role of discretionary rule-making by courts. Dworkin
may have, and clearly does, different expectations as to
what an analytical theory of law is about. But it is hardly
an unprejudiced objection to Hart that Hart does not do
what Dworkin thinks he ought to be doing, as long as Hart
is in fact doing what he thinks he ought to be doing. As for
the comment that Harts analysis does not contribute anything to the problem of how to account for legal principles,
it must be remembered that this is Dworkins problem, not
Harts. It is absurd to criticise a doctrine for not solving a
certain problem, when in the eyes of those propounding the
doctrine there is no such problem.6
IV. HART VERSUS HART: STRONG DISCRETION AFTER ALL?
So far I have argued in Section II that Dworkins attempt
to saddle legal positivism with a necessary commitment to
common-law judicial rule-making as strongly discretionary
is a failure. I have argued in Section III that it is plausible
6 We must be careful, though. All the above comments demand of Dworkin is
that he gives a reason independent of the disagreement between himself and positivism for showing that there really is a problem of how to account for legal principles, and then that positivism fails to contribute to it. However, to pursue this
would take the paper beyond its mandate.
355
ROGER A. SHINER
to see common-law judicial rule-making as involving weak

discretion, that Hart has said important things that accurately represent this feature of common-law judicial
rule-making, and that there is no reason to think that a
theoretical commitment to common-law judicial rule-making as involving weak discretion is at all problematic for legal positivism. It remains now in this fourth section of this
essay to address the remaining issue for this paper, the difficulties Hart makes for himself by the language, metaphors
and images he uses to lay out his view.
There is a passage almost at the end of the Postscript in
which Hart seems to acknowledge that discretionary judicial law-making is after all to be regarded as strong in
Dworkins sense (Concept, 274-5). He says essentially that
of course when judges do make law they do not push
away their law books and start to legislate without further
guidance from the law. They do proceed by analogies with
existing law and they do cite general principles or general
aims or purposes that an area of existing law can be plausibly seen to embody. But these considerations, he says, defer [but do] not eliminate the moment for judicial law-making. Analogies and general principles may not dispose of
the matter, and that is the point at which the judge must
proceed by the exercise of discretion. In other words, discretionary law-making by judges begins at the moment
when the ability of existing law to be dispositive ends. That
comes very close to saying that discretionary law-making by
judges is not bound by any standards set by the authority
concerned. The key question here seems to me to be not
the question whether Dworkin is after all correct to attribute to positivism in general and Hart in particular adherence to a doctrine of strong discretion. The key question
rather is whether there is common to both Hart and
Dworkin some misapprehension about the nature of judicial reasoning, especially in hard cases. I shall argue that
there is, but we must approach the matter somewhat
indirectly.
356
At the beginning of Model, after he has introduced his

three tenets of positivism, Dworkin tries to show how the
vies of both Austin and Hart embody these tenets. On page
18, he proposes to find a doctrine of judicial discretion in
Austin. In my view, Dworkin is at best anachronistic and at
worst just mistaken: Austin doesnt use the word discretion, and no such doctrine can be extracted from what he
does say. Austins view, as we know, is that laws proper are
commands of the sovereign. Austin however is very well
aware that, if I may put it thus in a philosophically neutral
way, judges do develop the law in the areas of common-law
precedent and the interpretation of statutes. Austin in fact
regards such development as potentially of great social
value (cf., e.g., Austin 1954, 1901). Judges have a role in
making up for the negligence or the incapacity of the
avowed legislator. Judge-made law is brought within the
requirements of the Command Theory by the notion of the
sovereigns tacit command. After the judge has ruled, the
sovereign could countermand the ruling: if the sovereign
does not, then the rule has the status of a command of the
sovereign, a tacit command (cf., e.g., ibid., 32). This framework, as Austin explicates it, does not yield any kind of judicial discretion. It is very clear that Austin does not use
the idea of tacit command of the sovereign indiscriminately. Austin was a utilitarian as regards political morality: the major constraint on whether any given piece of judicial law-making counts as a tacit command of the sovereign
is whether the law as made makes normative sense. Moreover, it is also clear (cf. Morison 1982, 789, 1045, and
elsewhere) that Austin appealed as well to some more formal notion of coherence with existing law as a criterion for
the sovereigns tacit commands. Austin relied on the idea
that the new law must be based on analogies with existing
law. According to Morison, Austin seems to have thought of
analogy as a form of logical link, even one that would support a degree of codification of the law, although he left little evidence as to exactly what he had in mind. If we ask of
357
ROGER A. SHINER
Austins view, then, the same question as we asked of

Harts, which is the more plausible of Dworkins two kinds
of discretion to find in Austin, it would seem to be at most
weak, not strong. Judicial law-making operates within
standards set by the authority concerned, not outside
them.
The point of this brief historical excursus is that it shows
how far Hart goes to present his view in ways that expose
him to Dworkins philosophical concerns. It is a commonplace that Concept in 1961 was much influenced by issues
and doctrines in philosophy at the time ordinary language philosophy, Friedrich Waismanns concept of open
texture, Wittgensteins ideas about family resemblances
and following rules, the emphasis in both Wittgenstein and
John Wisdom on case by case argument and on the law as
the prime repository of such argument, and the like.7 Those
ideas though in terms of their internal logic are fully compatible with what one might call an Austinian view of judicial law-making that is, a view in which such law-making is underwritten by such logical, or quasi-logical, ideas
as analogy and coherence. It is still judicial law-making,
but it is not strongly discretionary.8
Chapter 7 of Concept is fully compatible with such a view.
Hart has two opponents in this Chapter, formalism and
rule-scepticism. As against formalism, he argues that the
law is not rigid all the way down. The law is flexible enough
to be able to respond to both lifes and its own
indeterminacies. As against rule-scepticism, he argues that
this flexibility does not go all the way down either. Most of
the time the law applicable to a fact situation is clear, and
certain enough to be a guide to action. Hart could have expressed these ideas without using the spatial metaphors
that he does in fact use. He could, for example, have talked
about legal reasoning affording an opportunity to courts to
7 For some details, see Lacey 2004, 128, 13941. Specifically, Hart attended
Waismanns lectures in philosophy of language at Oxford. See also Schauer 2011.
8 See here again Shiner 2005, 49-56.
358
engage in law-making by adjusting and amending, by expanding or contracting the law as unforeseen circumstances, or unnoticed indeterminacies, require. He could
even have summarized this power of the courts as a discretionary power, a power or right to decide or act according
to ones own judgment. The expression according to ones
own judgment, as Dworkin himself emphasizes more than
once, can be given a perfectly innocent interpretation in
terms of reliance on ones own judgment simply in order to
make any judgment at all, as opposed to relying on ones
own judgment just because it is ones own judgment (see,
for example, Dworkin 1978, 124).
If Hart had expounded his view in terms such as these, it
would have been very hard for Dworkin to get traction for
any claim that Hart or any other positivist is committed to
what Dworkin calls strong discretion, to saying that a
court exercising such a discretionary power is like a sergeant being told to pick any five men he likes to go patrol in
not being bound by any standards set by the authority in
question. Harts specific interest in Waismanns views and
Waismanns image of the open texture of language in that
sense has not served Hart or legal positivism well. This image and those associated with it (see the second paragraph
of this essay for a more detailed list) create the association
of discretionary decision-making with gaps and space, with
there being no law there where the decision-making takes
place, and thus pave the way for Dworkins attack. These
images have helped to obscure, rather than illuminate, the
aspects of legal reasoning and common law adjudication
that Hart quite rightly thought in the context of the conflict
between formalism and rule-scepticism that he needed to
emphasize.9
Moreover, to the extent that Hart thinks these images
must be taken seriously that there really are gaps in the
law, that courts really do reason in open fields, that they
9 For more on the role of spatial metaphors in the construction of the conflict
between positivism and anti-positivism, see Shiner 1992, 316-21.
359
ROGER A. SHINER
create law Hart shows himself, not to be an opponent of

formalism and rule-scepticism, but to accept the framework
within which such theories operate. There is a fundamental
tension between the good Hart who sees that common-law
judicial rule-making and adjudication is misrepresented by
the formalist and rule-sceptical extremes; and the bad
Hart who thinks that where courts must go beyond existing law there is no law but only spaces and gaps. Consider
again Austin and his idea that common-law judicial
rule-making is based on the finding of analogies. The kind
of reasoning that goes on in precedential decision-making
(see again the passages referred to in notes 5 and 7 above)
is case-by-case reasoning. It is not deductive or formal, and
it is not not reasoning at all. The logic of a piece of legal
reasoning that supports a conclusion by reference to analogies, to similarities and differences with precedent cases, is
misrepresented by characterizing such reasoning as operating in an open field or in a gap. That image belongs to
the formalist or deductivist who believes that if a premise
does not imply or even entail its conclusion then the proffered argument is no argument at all, or to the rule-sceptic
who believes exactly the same thing the Nightmare/Noble Dream commonality that Hart so cogently describes
elsewhere (Hart 1983, Chapter 4). Harts images and metaphors in his story about common-law judicial rule-making
are the images and metaphors of the formalists and the
rule-sceptics, even though the underlying theoretical points
he wishes to make are against those theorists. Dworkins
attribution to legal positivism of a commitment to strong
discretion relies on those images and metaphors. Whatever
trouble Hart gets into as a result of opening himself to
Dworkins critique by relying on these images and metaphors is Harts own responsibility. It is not the responsibility of legal positivism as such.
Strip those images and metaphors away, and what is left
is an important account of common-law judicial rule-making, as a certain kind of reasoning that balances the laws
360
need for certainty against the laws need for flexibility. To

appreciate the theoretical value of Harts discussion of common-law judicial rule-making in Concept Chapter 7, we
need to focus on what Hart does in this Chapter, not on the
language in which he speaks.10
V. REFERENCES
AUSTIN, John [1832] 1954, The Province of Jurisprudence Determined. Ed. H. L. A. Hart. London: Weidenfeld and
Nicholson.
DWORKIN, Ronald M. 1978, Taking Rights Seriously. 2nd edn.
Cambridge, MA: Harvard University Press.
HART, H. L. A. 1983, Essays in Jurisprudence and Legal Philosophy. Oxford: Clarendon Press.
[1961] 1994, The Concept of Law. 2nd. Ed. Penelope
A. Bulloch and Joseph Raz. Oxford: Clarendon Press.
LACEY, Nicola 2004, A Life of H. L. A. Hart: The Nightmare and
the Noble Dream. Oxford: Oxford University Press.
MORISON, W. L. 1982, John Austin. Jurists: Profiles in Legal
Theory. London: Edward Arnold.
RAZ, Joseph, 1975, Practical Reason and Norms. London:
Hutchinson.
SCHAUER, Frederick F. 1991, Playing By The Rules: A Philosophical Examination of Rule-Based Decision-Making
in Law and in Life. Clarendon Law Series. Oxford: Clarendon Press.
10 An earlier version of this paper was presented to the Special Workshop on

the Fiftieth Anniversary of The Concept of Law, IVR World Congress, Frankfurt, in
August 2011. I am very grateful to Imer Flores for organizing the Workshop and inviting my participation. Thanks to my fellow panelists in the Workshop for comments Tom Campbell, Perluigi Chiassoni, Imer Flores, Noam Gur and Eerik
Lagerspetz, and also to Bartosz Greczner for helpful discussions. The research
leading to this paper has been supported by a grant from the Social Sciences and
Research Council of Canada: I thank the Council.
361
ROGER A. SHINER
SCHAUER, Frederick, 2011, On the Open Texture of Law.

Available at http://ssrn.com/abstract=1926855, 13th
September 2011.
SHINER, Roger A. 1992, Norm and Nature: The Movements of
Legal Thought. Clarendon Law Series. Oxford: Clarendon Press.
2005, Legal Institutions and the Sources of Law. Vol. III
of A Treatise of Legal Philosophy and General Jurisprudence, ed. Enrico Pattaro, Gerald J. Postema, and Peter Stein. Dordrecht: Springer.
362
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
LESSONS FROM HART

W. J. WALUCHOW
Resumen:
En este ensayo el autor defiende a H. L. A. Hart de dos crticas recurrentes a su teora de las normas sociales y a las obligaciones que regularmente acompaan a estas normas. Ambas crticas, sostiene el autor,
descansan sobre malos entendidos o caracterizaciones equivocadas de lo
que Hart en realidad se propuso. Estos malentendidos son explicados
plausiblemente a travs de la falta de capacidad de los crticos de apreciar plenamente dos de las valiosas lecciones que Hart deseaba expresar
en su lectura inaugural. En primer lugar, palabras como norma y obligacin no deberan ser sacadas de sus distintos contextos de uso para
someterlas a definiciones abstractas y al anlisis filosfico. En segundo
lugar, cuando se analizan enunciados en los cuales figuran tales palabras, tanto en la teora como en la prctica del derecho, resulta primordial tener en mente las diversas funciones que esos enunciados puedan
tener y los diferentes propsitos para los cules esos enunciados fueron
proferidos.
Palabras clave:
Normas sociales, obligacin, jurisprudencia analtica, filosofa
del lenguaje, H. L. A. Hart.
Abstract:
In this paper, I defend H. L. A. Hart against two prevalent criticisms of his
views on social rules and the obligations with which these rules are often
associated. These criticisms, I argue, rely on misunderstandings or
363
W. J. WALUCHOW
mischaracterizations of what Hart actually intended. These misunderstandings are plausibly explained by a failure on the part of his critics to appreciate fully two of the valuable lessons Hart sought to communicate in his inaugural lecture. First, words like rule and obligation should not be
removed from their various contexts of use and subjected to abstract philosophical definition and analysis. Second, when analyzing assertions in
which such words figure in both legal practice and theory, it is crucial to
bear in mind the different functions these assertions can be made to serve
and the different purposes to which they can be put.
Keywords:
Social Rules, Obligation, Analytical Jurisprudence, Philosophy
of Language, H. L. A. Hart.
364
LESSONS FROM HART
SUMMARY: I. Thoughts on Methodology. II. Rules and Social

Practices. III. Rules of Obligation. IV. Are Social Rules Nothing but Social Practices? V. Himma,
Hart and Austin: Still Apart after All These Years.
VI. Concluding Thoughts.
I. THOUGHTS ON METHODOLOGY
Upon assuming the Oxford Chair of Jurisprudence, H. L. A.
Hart presented his inaugural lecture in which he warned of
certain methodological traps into which legal philosophers
sometimes fall. In issuing these cautions, Hart saw himself
as following in the footsteps of his positivist predecessor,
Jeremy Bentham, a philosopher Hart much admired, and
from whom he thought we had much to learn. It is vital,
Hart wrote, that we
attend to Benthams warning that we should not, as does
the traditional method of definition, abstract words like
right and duty, State, or corporation from the sentences
in which alone their full function can be seen, and then demand of them so abstracted their genus and differentia.1
On the contrary, we must never take these words alone

but consider whole sentences in which they play their characteristic role. We must take not the word right but the
sentence You have a right, not the word State, but the sentence He is a member or an official of the State. 2 Furthermore, Hart suggested, we should always bear in mind that
one and the same word or sentence, used in one context,
can play a very different role, and mean something quite
different, in another. Smith has a right to be paid 10
said by a judge in deciding a case has a different status
1 Hart, Definition and Theory in Jurisprudence, in Essays in Jurisprudence

and Philosophy (Oxford: OUP, 1983), 31.
2 Ibid., 26.
365
W. J. WALUCHOW
from the utterance of it out of court, where it may be used

to make a claim, or an admission and in may other ways.3
We can discern at least three different recommendations
in these thoughts on methodology. First, in doing philosophy of law we should avoid the traditional, Aristotelian
method of definition per genus and differentia. Second, no
other method of definition, if applied to singular words like
right, duty corporation or law is likely to shed much
light on the phenomena we seek to understand, at least in
the case of a multi-faceted and somewhat puzzling phenomenon like law. Third, one and the same word or string
of words can be used to serve different functions or roles in
different sentences, as when a judge states that The defendant is liable for damages and I later use this very same
string of words in reporting that authoritative verdict. In
other words, we very much need to be sensitive to the context of utterance when attempting to build an understanding
of a phenomenon like law by examining the words and concepts typically employed in its analysis and/or practice.
There is no doubt much to dispute in these recommendations.4 But my aim in this paper is not to criticize Hart.
Rather, I want to pick up on his important point that, in attempting to build a theory of law partly by looking at how
its key terms and concepts are typically employed within its
practice, it is absolutely crucial to bear in mind the varying
contexts in which these terms and concepts can be used.
More specifically, we should always ask whether the assertion in question is being made from what Hart famously
termed the internal point of view or from what he referred
to as the external point of view.
The importance of differing perspectives from which assertions can be made, so seemingly obvious, is one, unforIbid., 28.
See, e.g., Jonathan Cohen, Definition and Theory in Jurisprudence Proceedings of the Aristotelian Society, Supplementary Volumes, Vol. 29, Problems in
Psychotherapy and Jurisprudence (1955), pp. 213-264. See also, P.M.S. Hacker,
Definition in Jurisprudence The Philosophical Quarterly, Vol. 19, No. 77 (Oct.,
1969), pp. 343-347.
3
4
366
LESSONS FROM HART
tunately, that legal scholars do not always seem to keep

firmly in mind. And among the most egregious offenders are
certain critics of Harts views. Time and again one encounters positions ascribed to Hart that he could only have held
had he, for some inexplicable reason, simply forgot his very
own recommendations. Its as if Hart, in analyzing some of
his key concepts, had ignored the importance of context of
utterance and the attendant danger of attempting to understand words in isolation from the sentences in which they
can be employed. Some of the most glaring examples are
found in discussions of two notions central to Hartian legal
thought the notion of a social rule and that of the obligation with which such a rule is sometimes associated. My
aim is this paper is to make some headway towards rectifying this situation. I want to help clear up some common
misunderstandings of Harts views on these two matters
that are plausibly explained by failures fully to appreciate,
as Hart himself almost always did, the importance of context and perspective when doing legal theory.
II. RULES AND SOCIAL PRACTICES
In The Concept of Law, Hart presents himself as building
a theory of law in response to perceived deficiencies in the
views of his positivist predecessor John Austin. In Harts
view, one of Austins most fundamental mistakes was his
failure to provide the theoretical apparatus with which to
explain what is commonly called the normativity of law. As
is well known, Austin attempted to reduce law to the empirically discernible commands of a sovereign who is habitually obeyed, and who has the power and intention to inflict
sanctions should his commands be disobeyed. In effecting
this reduction, Hart argued, Austin was unable to explain
adequately the ways in which many people think about law
and the normative vocabulary they employ when engaged in
various activities either within, or in some way related to,
legal practice. For example, those subject to law do not al367
W. J. WALUCHOW
ways view themselves as simply obliged to obey its directives upon pain of penalty. They often view themselves as
under its obligations and as possessing the rights and powers that law, via its secondary rules, creates and facilitates.
With respect to those laws that impose obligations on them,
they view themselves as bound by the relevant norms, not
by any sanctions that might or might not be inflicted
should they fail to comply. In order to capture and explain
this normative dimension of law, its internal aspect, we
need, Hart claimed, to replace the notion of a command
with that of a rule.
Yet if law is fundamentally a matter of rules, then a question naturally arises for Hart, as it does for any theorist
who views law this way: Where do legal rules come from?
Harts answer was that many legal rules are brought into
existence through the use of formal procedures like legislative enactment, a procedure made possible by a systems
secondary, power-conferring rules. But this raises a further
question: Where do these secondary rules come from and
from where do they get their normative force? According to
Kelsen, we must presume a basic norm that authorizes
the entire process, that gets it off the ground, so to speak.
Wanting to anchor legality in social reality, Hart proposed
instead that the relevant basic rules are neither presumed
nor deliberately created via specified formal acts or procedures. Rather, they arise informally by way of the behaviour and attitudes the practices of those who willingly
observe them and whose rules they are. Harts famous rule
of recognition is, of course, the foremost example of these
special rules in virtue of which the normative, i.e. rule-governed practice of law comes into being. It is not postulated
or presumed as in Kelsens work. Nor is it formally enacted
as in the case of the Canadian Criminal Code or the Canadian Charter of Rights. On the contrary, it exists within and
as a result of the actual social practices of those whose rule
it is. It is, in short, a social rule. Such a rule exists when
there is a pattern of behaviour accompanied by the appro368
LESSONS FROM HART
priate critical, reflective attitude. Members of the group who

observe the rule not only engage in a fairly widespread and
uniform pattern of behaviour, they also take an internal
point of view toward that pattern:
What is necessary is that there should be a critical reflective
attitude to certain patterns of behaviour as a common standard, and that this should display itself in criticism (including self-criticism), demands for conformity, and in acknowledgements that such criticisms and demand are justified, all
of which find their characteristic expression in the normative
terminology of ought, must, and should, right, and
wrong.5
When these conditions are met, we can meaningfully say

that the group engages in the practice of observing a social
rule.
In his recent book, Legality, Scott Shapiro addresses this
theory of social rules and accuses Hart of making a category mistake.
Social rules cannot be reduced to social practices because
rules and practices belong to different metaphysical categories. Rules are abstract objects. They are like games, numbers, plots, propositions, and concepts they are objects of
thought, not entities that exist within space and timePractices, on the other hands, are concrete events. They take
place within the natural world and causally interact with
other physical events.6
Reducing rules to social practices that is, claiming that

a social rule just is nothing but a social practice is a mistake, Shapiro believes. The mistake is revealed in a number
of ways not least of which in our way of talking about
practicesWe say that practices exemplify, embody, conform to, are structured by, and are grounds for rules. This
5
6
Hart, The Concept of Law, 2nd ed. (Oxford: Clarendon Press, 1994), 57.
Scott Shapiro, Legality (Cambridge: Harvard University Press, 2011), 103.
369
W. J. WALUCHOW
strongly suggests that rules are standards that guide conduct, not the conduct itself.7
Shapiros charge that Hart has made a category mistake
echoes criticism leveled years earlier by Ronald Dworkin in
The Model of Rules II. Here, Dworkin also claims that
Harts social rule theory mistakenly takes a social rule to be
nothing but a social practice and views the assertion or expression of such a rule as nothing over and above a factual
statement to the effect that the practice-conditions for the
existence of the rule are satisfied. The existence of a social
rule, and therefore the existence of the duty, is simply a
matter of fact.8 Dworkin goes on to argue that people who
invoke social rules in evaluating behaviour do not view
them this way. This is certainly true of a person who finds
the social practice pointless, or silly, or insulting.9 Such a
person may believe that it does not even in principle justify
asserting any duties or normative rules of conduct, and in
that case he will say, not that it imposes a duty upon him
which he rejects, but that, in spite of what others think, it
imposes no duty at all.10 It is true, Dworkin adds, that people who invoke rules often do assume a social practice such
as Hart describes as an essential part of their justification
for doing what they do. This might be, for example, because
they do not wish to disturb settled expectations and cause
the harm that might result from the frustration of those expectations. In such cases, the social practice helps to justify a rule which [a] normative judgment states.11 But
Harts theory doesnt put things this way: it erroneously assumes that the rule just is the practice. In other words,
Hart has committed the category mistake that Shapiro,
years later, accused him of making. The existence of a soIbid.
Ronald Dworkin, Taking Rights Seriously (Cambridge: Harvard University
Press, 1978), 49-50.
9 Ibid., 58.
10 Ibid.
11 Ibid., 57.
7
8
370
LESSONS FROM HART
cial practice is an empirical matter of fact that can help justify the assertion of a normatively binding rule of behaviour
and the normative judgments it supports. But such a practice can never itself be or constitute a normative rule.
III. RULES OF OBLIGATION
Included within a societys social rules will be some that
are thought to impose duties and obligations.12 These are
the rules which the societys members generally view as vitally important to social life or some crucial part of it, and
which they are willing as a result, and if necessary, to back
up with serious social pressure. These rules are also characteristically recognized as requiring conduct that, while
benefitting others, may come in conflict with self-interest.
Hence obligations and duties are thought of as characteristically involving sacrifice or renunciation.13 Rules prohibiting gratuitous infliction of harm and the keeping of promises typically count among a societys obligation rules; rules
governing fashion and polite discourse usually do not.
Harts theory of obligation has, like his theory of social
rules, been subject to extensive critique over the years.14
Dworkin, for example, correctly pointed out that Hart was
wrong to think that one cannot sensibly assert an obligation unless there is a social rule that supports its existence.
Vegetarians, for example, can meaningfully claim that everyone has a moral obligation not to eat meat, despite the
fact that doing so is an entrenched practice within virtually
all societies. In the Postscript, Hart acknowledges the force
of Dworkins criticism, as well as the validity of Dworkins
12 In most of his discussions of obligation and duty, Hart refers to the two interchangeably. I will follow Harts lead here.
13 The Concept of Law, 87.
14 See, e.g., Richard Bernstein, Professor Hart on Rules of Obligation Mind,
New Series, Vol. 73, No. 292 pp. 563-566; Roscoe E. Hill, Legal Validity and Legal
Obligation The Yale Law Journal, Vol. 80, No. 1 (1970), pp. 47-75; Michael Green,
Legal Realism as Theory of Law, William and Mary Law Review, Vol. 46, No. 6
(2005), pp. 1915-2000.
371
W. J. WALUCHOW
claim that he, Hart, had failed carefully enough to distinguish between the concurrent acceptance of a rule and the
acceptance of a conventional rule. A community displays a
concurrent [acceptance] when its members are agreed in
asserting the same, or much the same, normative rule, but
they do not count the fact of that agreement as an essential
part of their grounds for asserting that rule. It displays [acceptance of a conventional rule] when they do so.15 To this,
Hart responds:
My account of social rules is, as Dworkin has also rightly
claimed, applicable only to rules which are conventional
But the theory remains as a faithful account of conventional
social rules which include, besides ordinary social customs
(which may or may not be recognized as having legal force),
certain important legal rules including the rule of recognition, which is in effect a form of judicial customary rule existing only if it is accepted and practiced in the law-identifying and law-applying operations of the court.16
Yet another criticism of Harts theory of obligation is that

it fails to mark a significant advance over Austin.17 According to this line of argument Hart, just like Austin before
him, has presented a theory that reduces obligation to the
threat of sanction, in Harts case serious social pressure to
conform. On Harts view those subject to an obligatory social rule end up being obliged, not obligated, and so Hart
himself falls prey to the devastating criticisms he launched
against Austins theory. In a recent paper, Kenneth Himma
draws attention to this line of criticism and turns it on its
head. Far from being a basis for rejecting Harts theory of
social obligation, this aspect of Harts view actually reveals
the great insight contained within it. The centrality of serious social pressure as an obliging force is actually a strong
point in favour of Harts theory of social obligations. Upon
15
16
17
Taking Rights Seriously, 53.

Ibid., 256.
So far as I know, Hart never did accept the validity of this criticism.
372
LESSONS FROM HART
making these points, Himma goes on to develop his own sophisticated and subtle theory of legal obligation that, as his
title suggests, brings Austin and Hart together in ways few
would have imagined.
Hart explains the binding character of social obligations in
terms of considerations ordinary persons are likely to regard
as having normative significance.Social pressure in the
form of a hostile reaction is something people with ordinary
psychological characteristics tend to regard as having normative force[I]t is an empirical fact that ordinary persons
tend to dislike criticism and hostility and are willing to take
at least minimal steps to avoid it[I]t is clear that Hart views
social pressure as being a necessary constituent of social
obligation. As Hart puts the view, such pressure is the primary characteristic of obligation.18
So the binding force of social obligation is explained in

terms of the characteristic desire of human beings to avoid
social disapproval and pressure to conform. The normative
force of an obligation rule consists in a kind of sanction,
just as it did in Austin.
IV. ARE SOCIAL RULES NOTHING BUT SOCIAL PRACTICES?
At the outset I noted Harts claim that context and point
of view play important roles in legal theory. Words should
not be yanked from their contexts of utterance and subjected to abstract definition; assertions should not be analyzed independently of the differing points of view from
which they can be made. But this seems precisely to be
whats going on in the critiques of Harts theories to which I
have just drawn attention or at least this is what I will
endeavour to show in the remainder of this paper. In so doing, I hope both to rescue Hart from the charge of egregious
18 Ken Himma, Hart and Austin Together Again for the First Time: Coercive
Enforcement and the Theory of Legal Obligation, presented at The Nature of Law,
conference held at McMaster University, May 2011, p.p. 11-12. Internal reference
to Hart is from The Concept of Law, 87.
373
W. J. WALUCHOW
error and to lend support to the wisdom of heeding at least

some of his inaugural lessons.
Lets begin with the charge, issued in effect by both
Shapiro and Dworkin, that Hart has, in identifying rules
with social practices, committed an elementary category
mistake. What could possibly lead someone to make such a
charge? The first thing to note is the way in which each
theorist characterizes Harts views. Heres how Shapiro
puts it: Hart tries to demystify the creation of social rules
by reducing them to social practices. In other words, social
practices generate rules because these rules are nothing but
social practices.[S]ocial rules are just social practices.19
With this definition in hand, Shapiro goes on easily to fashion an argument against Hart that reduces, essentially, to
this: rules are abstract normative objects; social practices,
on the other hand, are concrete events in the world; therefore rules cannot be social practices. Notice now how
Dworkin describes Harts theory:
Duties exist when social rules exist providing for such duties. Such social rules exist when the practice conditions for
such rules are met. These practice conditions are met when
the members of a community behave in a certain way; this
behavior constitutes a social rule, and imposes a dutyThe
existence of a social rule, and therefore the existence of the
duty, is simply a matter of fact.20
What could possibly lead one to take Hart as saying the

sort of things attributed to him by Dworkin and Shapiro in
these passages, as saying that a rule just is nothing but a
social practice and that the existence of a social rule, and
any obligation or duty to which it might give rise, is simply
a matter of social fact? The short answer is that Hart undoubtedly thought that, in some (potentially misleading)
sense and from one important perspective, these things can
19
20
Legality, 95, authors emphasis.

Taking Rights Seriously, 49-50.
374
LESSONS FROM HART
sensibly be said to be true. Lets focus, for the moment, on

the claim that a social rule exists, leaving till later the claim
that a social rule just is a social practice. When a sociologist or legal historian asserts the existence of a social rule
within a particular society, his claim is true only if the social facts Hart describes obtain, only if there is in fact the
converging behaviour coupled with the appropriate beliefs
and attitudes that the existence of a social rule requires.
From such an external, theoretical point of view, the existence of a social rule, accepted by its practitioners from the
internal point of view, is determined by establishing the appropriate social facts, as is the existence of the duties with
which it is associated.21
Lets now shift gears, however, and consider the social
rule from the point of view of one who is subject to it, accepts its requirements, and invokes the rule in making normative assertions.22 When such a person asserts the existence of a duty under the social rule she might, of course,
mean nothing more than what the legal historian means.
This might be the case, for example, were she in the business of comparing the obligation rules practiced within her
society with those practiced in another. But then again she
might not. She might, in other words, assert that she actually does have an obligation to act in a particular way and
cite the rule in defence of her claim and any actions she
might take under its authority. But does she not thereby
imply or presuppose her belief that the sheer existence of
the rule within her society constitutes a sufficient reason
21 Notice how obscured this point becomes if we follow Shapiros lead and, in effect, take Hart to have defined a rule simply as a social practice.
22 Hart is well aware that one who is subject to a rules requirements might reject the rule, thereby assuming what Hart also calls, the external point of view.
One who thus rejects the rule and its requirements will dispute the normative
claims of those who accept the rule. What he will not do, however, is dispute the
claim of the historian. He rejects that he is subject to a valid normative requirement; not that there is a social rule which purports to impose one. For sake of simplicity, I will restrict my discussion to the differences between the historian and
one who is subject to a rule and accepts it as establishing a valid normative requirement.
375
W. J. WALUCHOW
for acting as she does? Would she, in justifying her behaviour by citing the rule that licenses or requires it, mean to
assert nothing over and above what the historian asserts?
Certainly not. As made plain in the Postscript, if the rule is
a conventionally based social rule, then part of the reason
behind her assertion will be the social facts Hart draws to
our attention and to which the sociologist or historian will
appeal in justifying his claim that the social rule exists. But
it will almost certainly be only part of the reason. Another
part might well consist in the various values, perhaps
moral, perhaps not, realized in practicing the particular social rule in question.23 More importantly, whatever that additional element is, the fact remains that the claim being
made by our rule supporter is not the same as the claim
made by our legal historian; it is not, as Hart put it in his
response to Dworkin, a statement of the external sociological fact that the practice-conditions for the existence of the
rule are satisfied.24
So what kind of statement is it? The simple answer is
that its a normative statement, signalling the speakers acceptance of the requirement that she behave as the social
rule prescribes. One who wishes to dispute this particular
statement will not necessarily (though she might) point to
the absence of the practice-conditions historians could
quibble over. She is more likely, however, to appeal to the
absence of sound or valid reasons for doing as the social
rule prescribes. She might, to use Dworkins very own
words, claim that the rule is pointless, or silly, or insulting25 in which case her view will not be that there is a duty
which she rejects but that there is no duty at all. And this
23 Hart claims that individuals can have any number of reasons for accepting
and abiding by a social rule. See The Concept of Law, p. 257. This any reasons thesis has been disputed by a number of theorists in regard to legal obligations. See,
e.g., Raz, Hart on Moral Rights and Legal Duties, Oxford Journal of Legal Studies,
Vol. 4, No. 1, pp. 123-31, and Shapiro, Legality, p. 114.
24 The Concept of Law, 256.
25 Taking Rights Seriously, 58.
376
LESSONS FROM HART
will be because, in her view, the rule does not establish a

valid normative requirement.
So from the internal point of view, the existence of a
duty-imposing rule, i.e. one that actually does impose the
obligation its practitioners take it to impose, is never merely
a matter of sociological fact. But how can this be if, as
Shapiro maintains, Hart has reduced social rules to social
practices claimed that a rule just is a social practice? If
social rules are nothing over and above social practices,
then if the requisite social practice exists, surely the duty
exists. And this must be acknowledged from whatever point
of view one cares to occupy. It is true that occasionally Hart
comes perilously close to saying things that commit him to
the reduction Shapiro attributes to him. For example, in his
response to Dworkin Hart says this: Rules are conventional social practices if the general conformity of a group to
them is part of the reasons which its individual members
have for acceptance.26 Careful consideration of this claim
does not, however, necessarily lend itself to the interpretation Shapiro would have us accept, that for Hart a social
rule is reducible without remainder to a social practice.
Consider, once again, how Harts claim, that rules are
conventional social practices, looks depending on the perspective from which it is made. From the external point of
view of an historian or legal sociologist, a social rule could,
I suppose, loosely be said to be a conventional social practice in which general conformity is part of the reason its
members cite for acceptance, criticism and so on. This is
because such a theorist will be concerned neither with the
question whether the rule really does provide the reason for
action its adherents believe it to provide, nor with any
metaphysical questions concerning the ontological status of
rules. Rather, she will more likely be concerned with the
empirical conditions under which rules can be said to exist
within groups or societies. And given this concern there is
no great harm in treating the rule as nothing over and
26
The Concept of Law, 255, my emphasis.
377
W. J. WALUCHOW
above a social practice. But from the internal point of view,

i.e. the point of view of one who actually cites the rule as
part of her justification for what she does, the rule will
never be sensibly viewed in this way. It will neither be constituted by nor reduced to the social practice in which she
participates. Rather it will be what the social practice creates and establishes and helps to justify in the way of behaviour.
Yet if the rule is not identical with the social practice itself ( except in the very loose, and potentially misleading,
sense associated with the legal historian or sociologist) then
exactly what is it? To this Hart gives no clear answer of
which I am aware. But there seems nothing in what he tells
us that should lead us to think that he would reject the
thought that the rule is, as Shapiro says, the abstract object expressed by or exhibited in the social practice the internal participant takes as providing part of her reason for
behaving as she does. It is an abstract object created by
and expressed in that social practice in much the same way
that some other rule perhaps the very rule in question
under different circumstances might be created and expressed by someones writing a string of words on a piece of
paper. As the old saying goes, there are many ways to skin
a cat; there are also many ways to create and express one
and the same rule. Writing is one of them; behaving in a
particular way is another. It is perhaps worth stressing that
none of this need be denied from the external point of view.
When our legal historian or sociologist cites social practice
as revealing the existence of a social rule, she need not be
taken to have identified the rule with that practice. What
she cites may be taken to constitute empirical evidence that
a social rule exists, i.e. a rule having the distinctive character of having been created, and in that sense constituted by,
the kind of social practice she describes. But the object created need not be viewed as identical with or reducible to the
practice that expresses it and by which it comes into being
any more than a statutory rule need be viewed as identi378
LESSONS FROM HART
cal with or reducible to the acts of legislators or the strings

of words that find their way into statute books.
V. HIMMA, HART AND AUSTIN:
STILL APART AFTER ALL THESE YEARS
Lets now turn to Harts views on obligation rules. Himma
suggests that we can arrive at the following insight from
Harts thoughts. The binding force of social obligation can
be explained in terms of the characteristic desire of human
beings to avoid social disapproval and pressure to conform.
Thus, the normative force of an obligation rule consists in
a kind of sanction, just as it did in Austin. To be clear,
Himma does not maintain that the wish to avoid serious social pressure necessarily explains a persons motives in accepting a social rule under which he has an obligation, nor
does he saddle Hart with this view. Social pressure is said
to explain how the ruleobligates, and not why [individuals subject to it] accept [the] rule.27 In other words, we
must, on this Hartian inspired theory of obligation, distinguish between reasons for accepting a rule, on the one
hand, and the rules normative force, on the other. Its the
pressure, not the reasons, that constitute the rule as obligatory and provide it with its so-called normative force.
I do not wish here to dispute either the soundness of this
intriguing theory of social obligation or the account of legal
obligation that Himma goes on to develop on its footing. Instead, I want to ask, as before, whether one can attribute
such a view to Hart. And as before, I shall begin by looking
at what actually Hart says, bearing in mind, once again,
the importance of the point of view from which assertions
concerning obligations can be made.
In presenting this thoughts on obligation, Hart cites three
distinct, but related characteristics:
27
Himma, 17.
379
W. J. WALUCHOW
1. 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.28
Two further characteristics of obligation are then said to

go naturally together with this primary one.
2. The rules supported by this serious social pressure are
thought important because they are believed to be necessary
to the maintenance of social life or some highly prized feature of it.29;
and finally
3. [I]t is generally recognized that the conduct required by
these rules may, while benefitting others, conflict with what
the person who owes the duty may wish to do. Hence obligations and duties are thought of as characteristically involving sacrifice or renunciation.30
Hart goes on to add this intriguing thought.

The figure of a bond binding the person obligated, which is
buried in the word obligation, and the similar notion of a
debt latent in the word duty are explicable in terms of these
three factors, which distinguish rules of obligation or duty
from other rules. In this figure, which haunts much legal
thought, the social pressure appears as a chain binding
those who have obligations so that they are not free to do
what they want.31
With these words of Hart in mind, lets now consider the

following questions. (a) What, in Harts view, is an obligation rule? And (b) Under which rules do obligations arise
28
29
30
31
The Concept of Law, 86.

Ibid., 87.
Ibid.
Ibid.
380
LESSONS FROM HART
within a community? Put this way, one might naturally answer these questions much as Himma indicates and go
on to conclude that Hart has not, in fact, taken a significant step beyond Austin. But one would come to these answers, I submit, only if one ignored the important lessons of
Harts inaugural lecture that is, only if one yanked the
words obligation and obligation rule out of context and
portrayed Hart as providing a definition in answer to question (a); and only if one fashioned a Hartian answer to
question (b) that for some reason ignored the differing contexts from which the question can be asked and answers to
it provided. But of course Hart explicitly rejected the utility
of definitions in legal theory, and so it would be unfair to
view him, in the passages cited above, as answering a question such as (a), just as it would be unfair to take him as
answering the question What is a social rule? by replying
simply that it is a social practice of a particular kind. As for
question (b), here the importance of context comes acutely
to the fore because we get very different answers depending
on the point of view from which the question is put.
Lets begin with the external, theoretical point of view of
our legal historian or sociologist. On this reading, the question asks how, from the perspective of such an external observer of a society and its workings, one could distinguish
social rules that function as obligation rules from those
that do not.32 Harts answer, of course, is that we would
look to those rules that bear the features he mentions: they
require personal sacrifice, are accompanied by serious social pressure and demands for conformity, and there is a
widespread belief that these responses are appropriate
because the rule protects what is taken to be a crucial
feature of social life or some highly prized feature of it.
Now consider these important points, so seemingly obvious,
yet in need of emphasis: the question so framed calls for a
32 Notice how Hart phrases his analysis: Rules are conceived and spoken of as
imposing obligations when; These rules are thought important because they
are believed to be necessary and so on (my emphasis).
381
W. J. WALUCHOW
descriptive answer drawing on relevant social facts about

behaviour and beliefs. And this is precisely what Hart provides. The existence of an obligation rule is, from this external perspective, a matter of sheer social fact. Its a matter of whether or not the right practice conditions obtain. It
follows from this that, in citing the belief that the rule is
valuable because of its role in contributing to the maintenance of social life or some valuable feature of it, the external theorist who follows Harts lead in no way commits himself to the truth of the belief he reports. On the contrary, he
might view the belief as blatantly false because he, unlike
those who accept it, views the rule as pointless, or silly, or
insulting.
Now consider question (b) from the internal point of view,
that is from the point of view of one who views the rule as
actually providing him and others with sound or valid reasons for action.33 Will such a person answer question (b) by
citing the features Hart describes? Will he cite the serious
social pressure to conform and the widespread belief that
the rule promotes or protects important values? Of course
he wont. Hell cite the important values. From his internal
perspective, its not the serious social pressure that produces the chain that binds, that provides the normative
force that Himma seeks. Instead its the values in light of
which such pressure and criticism are believed to be warranted or justified. That the shared social rule is, in his estimation, actually necessary to social life, or some highly
prized feature of it, is a good part of the reason why it
counts among those rules that actually impose obligations
on him.34 From his perspective, the existence of obligation
33 Bear in mind that, for reasons of simplicity, I have left out the so-called external perspective of one who is subject to the rules but rejects them, perhaps because he believes that the feature of social life protected e.g., patriarchy is in no
way valuable.
34 Harts any- reasons-thesis entails that the values invoked by those who
consider themselves bound by the rule may differ from person to person. But for
any such person, except perhaps one who acts from unthinking habit, or who is
under the spell of unadulterated rule-fetishism, there will be some value or other
to which appeal will be made.
382
LESSONS FROM HART
is not a matter of sheer social fact; it is not simply a matter

of the right practice conditions being met. And finally, any
attempt to define his obligation in terms of sheer social
practice, to reduce it to a constellation of behaviour, attitudes and beliefs, is to commit a category mistake.
VI. CONCLUDING THOUGHTS
In this paper, I defended H. L. A. Hart against two prevalent criticisms of his views on social rules and the obligations with which they can be associated. These criticisms, I
argued, rely on misunderstandings or mischaracterizations
of what Hart actually intended. These misunderstandings
are plausibly accounted for by a failure on the part of his
critics to appreciate fully two of the valuable lessons Hart
sought to communicate in his inaugural lecture. Words like
rule and obligation should not be removed from their various contexts of use and subjected to abstract philosophical
definition and analysis. Such a methodology, Hart advised,
only serves to obscure and confuse. Furthermore, when analyzing assertions in which such words figure in both legal
practice and theory, it is crucial to bear in mind the different functions these assertions can be made to serve and
the different purposes to which they can be put. An assertion that invokes a social rule, made from the external, theoretical point of view of a sociologist or legal historian, can
bear a meaning that differs greatly from one made from a
different point of view most notably the point of view of
one who views the rule invoked as providing him with reasons for action, perhaps of an obligatory nature. In drawing
these points to the readers attention, I hope to have made
some small headway in re-affirming the thought that, from
H.L.A Hart, we still have much to learn.35
35 Thanks to Otto Phillips, Fabio Shecaira, Imer Flores, Juan Vega, Ken Himma
and Enrique Caceres for helpful comments on earlier drafts of this paper.
383
PROBLEMA
Anuario de Filosofa
y Teora del Derecho
LESSONS FROM HART: A REPLY TO WALUCHOW

Kenneth Einar HIMMA*
Resumen:
En este breve ensayo, intento responder a la crtica de Wilfrid Waluchow
sobre mi teora positivista de la obligacin jurdica, crtica que aparece
en su colaboracin de este nmero, Lessons from Hart. Waluchow sostiene, con perspicacia y elegancia, que mi teora fracasa al no distinguir entre el punto de vista interno y externo, y que ofrece una explicacin de la
obligacin jurdica desde el punto de vista externo. Intento refutar esta
interesante crtica al argumentar que la distincin entre el punto de vista
interno y externo fue hecho dentro del anlisis de Hart de los conceptos
centrales del derecho y que yo intento proporcionar una explicacin desde el mismo punto de vista que toma Hart cuando analiza conceptos jurdicos, el de la metodologa tradicional del anlisis conceptual.
Palabras clave:
Obligacin, obligacin jurdica, normas sociales, punto de vista interno, punto de vista externo.
Abstract:
In this brief essay, I attempt to reply to Wilfrid Waluchows critique of my
positivist theory of legal obligation, which appear in his contribution to this
volume, Lessons from Hart. Waluchow argues, with insight and elegance,
that my theory fails to distinguish between the internal and external point
of view, and gives an account of legal obligation from the external point of
*
himma@uw.edu.
385
KENNETH EINAR HIMMA
view. I attempt to rebut this interesting criticism by arguing that the distinction between the internal and external point of view was made inside
Harts analysis of the concepts central to law, and that I am attempting to
give an account from the same point of view that Hart takes when he analyzes legal concepts that of the methodology of traditional conceptual
analysis.
Keywords:
Obligation, Legal Obligation, Social Norms, Internal Point of
View, External Point of View.
386
SUMMARY: I. Introduction. II. Waluchows Critique. III. Reply

to Waluchow.
I. INTRODUCTION
In his outstanding paper Lessons from Hart,1 Wilfrid
Waluchow does me the honor of critiquing my attempt to
provide a comprehensive positivist account of legal obligation that covers both the rule of recognition that governs
the conduct of officials and the primary rules of obligation
that bind citizens.2 The starting points for my argument are
(1) that, as positivism insists, law is a social artifact all the
way down; and (2) that, according to Hart, 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.3
The basic idea is that the quote in (2), understood as a
conceptual claim about the nature of social obligation, provides an adequate grounding for a comprehensive theory of
legal obligation Hart was obliged to give but never did, explaining only the second-order obligations of officials defined by the rule of recognition. Hart accomplished this by
recourse to what I take to be his theory of social norms and
obligations but, notably, had little to say about the legal
obligations of citizens.
This is problematic for two reasons. First, when we talk
about the obligations of law in legal practice, we are most
often speaking of the legal obligations of citizens generated
by the primary legal norms valid under the rule of recognition. This is not to say that the duties defined by certain
norms comprising the rule of recognition are not of theoretWil Waluchow, Lessons from Hart, Problema, n. 5, 2011.
Kenneth Einar Himma, A Positivist Theory of Legal Obligation, Problema, n.
5, 2011.
3 H. L. A. Hart, The Concept of Law (Revised ed.) (Oxford: Oxford University
Press, 1994), 86.
1
2
387
KENNETH EINAR HIMMA
ical importance, but Hart notes, correctly, that Austins basic starting point that law creates obligations is correct
and Austins theory was concerned with the legal obligations of citizens. This means not only that Hart lacks a
comprehensive theory of legal obligation, but also arguably
ignored the most important part of the problem.
Second, there is an element of unfairness involved. Hart
rejected Austins theory of law partly on the ground that
Austin couldnt distinguish a gunman from the state. That
is to say, that he found Austins account of legal obligation
confused the notion of being obligated with the notion of being obliged: one is obliged, not obligated, to conform to the
commands of a gunman and the legal system, on Harts
view of Austin, is nothing more than a gunman writ large.
According to Hart, Austin had explained no more than how
citizens are obliged to comply with the law. It is true that
Austin overlooked the rule of recognition and hence lacked
an essential piece of the story of legal obligation, believing
implausibly that sovereigns were incapable of legal limitation. Nonetheless, it seems unfair, among other things, to
reject Austins account of legal obligation, which applies
exclusively to citizens, and not offer something better.
In consequence, Hart lacked a comprehensive theory of
the normativity of law, one of the most important problems
in conceptual jurisprudence no matter how it is conceived. Theorists like Jules Coleman believe that it is sufficient to resolve the issue to explain how it is logically possible for law to create obligations or, otherwise put, make
intelligible laws ability to create legal obligations. On his
view, law purports to, but often fails, to create legal obligations.4 I have challenged this view elsewhere5 but, even if
4 Jules Coleman, Conventionality and Normativity, in Enrique Villanueva
(ed.), Legal and Political Philosophy. Social, Political, & Legal Philosophy, Volume 1,
(Amsterdam/New York, NY: 2002).
5 Kenneth E. Himma, Conceptual Jurisprudence and the Intelligibility of
Laws Claim to Obligate, in Michael ORourke, Joseph Keim-Campbell, and Harry
Silverstein (eds.), Topics in Contemporary Philosophy (Volume III): Law and Social
Justice (Boston: MIT Press, 2005).
388
this is the correct formulation of the problem of explaining

legal normativity, Hart has offered no comprehensive
account of even explaining this much.
Of course, as my contribution makes clear, I believe law
necessarily creates legal obligations and this is part of the
explanation of laws distinctive normativity. The specific
thesis of the paper is that the legal obligations of officials in
their capacities as officials and the first-order legal obligations of citizens under the primary rules can be explained
in terms of Harts theory of social obligation.6 What constitutes the binding force of a secondary rule of recognition on
officials (and hence laws distinctive normativity as it applies to officials) is, so to speak, the authorization of social
pressure as a response to non-compliance. As for the binding force of legal obligation as it pertains to citizens, it is
the authorization at least in modern municipal legal systems of which we are aware of coercive enforcement
mechanisms that constitutes the binding force of an
obligation and provides reasons for acting in accordance
with the law.
Now, it is important to note at the outset that the theory
of legal obligation I defend does not reduce obligation to the
authorization or appropriateness of formal or informal social pressure. There are other conditions that have to be
satisfied:
A norm is legally obligatory if and only if (1) it the behaviour it
requires is made mandatory (2) by a valid legal norm (in a legal system, of course) (3) that is exclusionary and (4) backed
by social pressure (5) because the norm protects what is
taken to be a particularly important feature of society.
6 It is, I should note, no part of my project as Waluchow makes clear, if I did

not, to attribute this thesis to Hart. My project is not an interpretive project. It is
charitable of Waluchow to acknowledge this because I think he offers very good
reasons to think that my view shouldnt be attributed to Hart although, as I mentioned, I think there is some reason to think he would have held this view.
389
KENNETH EINAR HIMMA
My claim in the paper is that it is the authorization of coercive enforcement mechanisms that constitutes a first-order
legally valid norm as having the binding force characteristically associated with obligations as a general matter. Moreover, this binding force is a way of describing the distinctive
reasons for action that obligations provide. Thus, a first-order legal obligation exists when conditions (1) through (5) are
satisfied. But the binding force and special reasons that legal obligations provide are constituted by authorization of
coercive enforcement mechanisms.
II. WALUCHOWS CRITIQUE
Waluchows subtle and important objection to my thesis
is that my account fails to overlook the idea that what constitutes the binding force of a legal obligation depends on
what point of view we take towards the obligation. As
Waluchow puts this insightful objection much more elegantly than I can, I will quote him here:
Lets now consider the following question[]. Under which
rules do obligations arise within a community? Put this way,
one might naturally answer these questions much as Himma
indicates and go on to conclude that Hart has not, in fact,
taken a significant step beyond Austin. [But] here the importance of context comes acutely to the fore because we get
very different answers depending on the point of view from
which the question is put.
Lets begin with the external, theoretical point of view. On
this reading, the question asks how, from the perspective of
an external observer of a society and its workings, one could
distinguish rules that function as obligation rules from those
that do not. Harts answer, of course, is that we would look
to those rules that bear the features he mentions: they require personal sacrifice, are accompanied by serious social
pressure and demands for conformity, and there is a widespread belief that these responses are appropriate because
390
the rule protects a crucial feature of social life or some

highly prized feature of it..
Now consider [the] question from the internal point of
view, that is from the point of view of one who views the rule
as actually providing him and others with sound or valid
reasons for action. Will such a person answer [the] question
by citing the features Hart describes? Will he cite the serious
social pressure to conform and the widespread belief that
the rule promotes or protects important values? Of course he
wont. Hell cite the important values. From his internal perspective, its not the serious social pressure that produces
the chain that binds, that provides the normative force that
Himma seeks. Instead its the values in light of which such
pressure and criticism are believed to be warranted or justified. That the shared social rule is, in his estimation, actually necessary to social life, or some highly prized feature of
it, is a good part of the reason why it counts among those
rules that actually impose obligations on him. From his perspective, the existence of obligation is not a matter of sheer
social fact; it is not simply a matter of the right practice conditions being met.7
This subtle and insightful criticism, as I understand it, is

that my account of the binding force of an obligation incorrectly takes the external point of view towards the rule, explaining it in terms of a disinterested observer (i.e., one who
neither accepts, endorses or supports the legal system or
the rule) who sees only that the rule generating the obligation is supported by some form of social pressure. The disinterested observer does not regard himself as bound by the
rule and simply discerns a signal that the rule is considered obligatory by those persons in the legal system. On
Waluchows view, the disinterested observer identifies a
characteristic that all legally obligatory norms might share
but this characteristic cannot explain the binding force of
the legal obligation.
Waluchow, Lessons from Hart, Problema, n. 5, 2011, p. 380.
391
KENNETH EINAR HIMMA
To understand the binding force of a legal obligation, one

must consider the attitude of someone who takes the internal point of view towards the legal system that is from
the point of view of one who views the rule as actually providing him and others with sound or valid reasons for action. If you ask such a person why he is bound by the legal
rule, he will respond that it is the importance of the value
protected by the rule and not the (formal or informal) social pressure that is authorized for violations of the rule.
For someone who takes the internal point of the view toward the rule, it is the value that binds that provides the
special normative force that legal obligations provide.
III. REPLY TO WALUCHOW
As it turns out, Waluchows helpful critique assumes a
plausible proposition that I reject namely, that the project
of conceptual analysis of legal concepts requires, at least
some of the time, that the theorist articulate the content of
concepts from the vantage point of someone who takes the
internal point of view towards the law. I disagree on this
proposition not because I believe that the conceptual theorist must take the external point of view with respect to
conceptual theorizing about law. But rather because the
distinction between internal and external point of view occurs inside Harts analysis of the relevant legal concepts.
Of course, the differences between how people view the
law from the external point of view and how they view it
from the internal point of view is important but these are
primarily sociological questions that require empirical analysis that is no part of Harts theory. Harts contribution to
this enterprise was to make the distinction so as to explain
the concept of a social norm, among other things. But it is
important to note that how people view law, legal normativity, legitimacy, and so on from a certain point of view are
not conceptual issues.
392
Conceptual analysis, of course, adopts an approach that

presupposes something like a point of view; but it is misleading to characterize it terms of an internal or external
point of view. Conceptual analysis is grounded in the language we use to impose a conceptual framework on the
world, and hence is grounded in something like conventions governing language. But there does not seem to be a
distinction between an internal point of view towards a language and an external point of view. Regardless of how
much English one knows or speaks, it is incorrect to say
that bachelor means law enforcement official. No one
with any knowledge of the language of the two terms would
simply report that English speakers typically regard that as
incorrect. That definition is simply incorrect.
My aim was to give a conceptual analysis of the notion of
legal obligation as it functions in all possible legal systems. There are, of course, disputes about what the appropriate methodology is: e.g., is the methodology purely descriptive or normative? These distinctions do not track the
distinction between the external point of view and the internal point of view. I am an advocate of traditional conceptual
methodology, and that might be a mistake, but not because
I adopt an external point of view.
But Waluchows analysis is intriguing and deserves a
more focused response than the above. Waluchow argues
that someone who takes the internal point of view towards
a legal system or a law will view the relevant law as having
binding force in virtue of the values it protects not in
terms of the authorization of coercive force.
I think there are two problems with this idea. First,
someone who views the binding force of the law as explained by the importance of values it protects is having a
moral reaction and is likely reporting the feeling that the
law imposes a moral obligation. If you ask someone who
takes the internal point of view whether the authorization
of coercive enforcement mechanisms is part of what makes
a law mandatory, obligatory, or something one must do (in393
KENNETH EINAR HIMMA
stead of merely something one should do), I would be surprised if anyone would answer no. The authorization of
coercive enforcement mechanisms is always a reason for
doing what the law requires even if it is not the best reason,
and it seems to be absolutely essential to the explanation of
the binding force of the obligation law generates. Of course,
someone who merely takes the external point of view towards the law will give the same answer.
It is important to acknowledge, of course, that many people will criticize behaviour simply because it breaks the law.
But people who criticize law-breaking on this ground view
the system as legitimate and the legal norm as being within
the scope of laws legitimate authority. Morally legitimate authorities are commonly supposed to generate content-independent obligations to obey, but these obligations are moral
in character. Accordingly, most people who criticize breaking a law simply because the law requires it are making
moral assertions that presuppose the legitimacy of the legal
system. But this is not a requirement of Harts theory, and
simply could not be because many people do not view laws
even in a system that is generally regarded as legitimate as
necessarily imposing legitimate requirements. For example,
nearly half of US citizens think it is illegitimate for the state
to redistribute income for the purpose of alleviating poverty
Second, if I am correct in thinking mandatory legal
norms necessarily create legal obligations, as Hart seems to
have, then it would have to be a conceptual truth that people in a legal system take the internal point of view towards
the law. But this is false. It is clearly possible that people
within a legal system take only the external point of view
towards the legal system or the law and hence would reject
the values the law promotes. And, indeed, Hart himself expressly acknowledges that it is not a conceptual truth that
citizens take the internal point of view towards the law or
the legal system.
Here it is crucial to note that if legitimate legal systems
give moral reasons to obey, it is uncontroversial among po394
litical theorists that law as such does not so much give an

even prima facie moral reason to obey the law. One simply
cannot presume that law as such is something that will be
accepted for moral reasons. Indeed, not only did Hart deny
that citizens must take the internal point of view towards
the law or the rule of recognition as an essential condition
for the existence of a law or legal system; he clearly pointed
out that officials can take the internal point of view towards
the law for any reason at all. It need not be the case that
even officials take the internal point of view towards the law
because they believe the content is morally acceptable.
Thus, even the officials need not claim that (or even purport) that citizens have a moral reason to obey the law. The
most they need claim is that they claim to have a legal reason, which under my theory, is entirely reducible to prudential reasons.
395

Flores Fabra 50th Anniversary of Harts Concep of Law

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Flores Fabra 50th Anniversary of Harts Concep of Law

Transféré par

Droits d'auteur :

Formats disponibles

PROBLEMA

INSTITUTO DE INVESTIGACIONES JURDICAS

INSTITUTO DE INVESTIGACIONES JURDICAS

De las opiniones sustentadas en los trabajos responden exclusivamente

Formacin en computadora: Karla B. Templos Nuez

Dra. Elvia Luca Flores valos

Este nmero incluye un cd-rom con el contenido de los nmeros

Enrique Cceres Nieto / Imer B. Flores / Juan Vega Gmez

Nmero de reserva al ttulo en Derechos de Autor:

DISCUSIN: EN TORNO A EL CONCEPTO DE DERECHO

Rewriting Harts Postscript: Thoughts on the Development of Legal Positivism . . . . . . . . . . . . . . .

Harts The Concept of Law as a Study in Social Philosophy

Farewell to the Rule of Recognition . . . . . . . . . . .

H. L. A. Hart and the Invention of Legal Philosophy. . .

Hartian Positivism as a (Plausible) Error Theory . . . .

Hart on Judicial Discretion . . . . . . . . . . . . . . .

Lessons from Hart: A Reply to Waluchow . . . . . . . .

P ROBLEMA . A NUARIO DE F ILOSOFA Y T EORA DEL

H. L. A. HART: SUPERVISOR, MENTOR, FRIEND,

So began the acknowledgments page of my book, Inclusive

W. J. Waluchow, Inclusive Legal Positivism (Oxford: Clarendon Press, 1994),

sion that often obscured the depth of his understanding;

H. L. A. HART: SUPERVISOR, MENTOR, FRIEND, INSPIRATION

What does Herbert think of this? to which I replied: He

pass muster. But I didnt know quite what to do about it.

H. L. A. HART: SUPERVISOR, MENTOR, FRIEND, INSPIRATION

phoned Hart who agreed to read my paper. Soon thereafter,

work. Not only did H. L. A. Hart read my paper; and not

H. L. A. HART: SUPERVISOR, MENTOR, FRIEND, INSPIRATION

what he had to say: I was out for a ride and thought Id

erything to help me meet an urgent deadline, I might never

ternal viewpoint will regard such behavior as an array of

condition for the existence of any such system (or, at least,

scope of his reflections as he pondered a number of other

order to deprive him of his unjust gain. Proponents of this

which maintains that no mode of conduct can legitimately

for Rawlss prioritization of certain liberties is a liberal ideal

where the action or event is abnormal. Unlike the criterion

FINNIS, John 1980, Natural Law and Natural Rights, Oxford,

REWRITING HARTS POSTSCRIPT: THOUGHTS

REWRITING HARTS POSTSCRIPT

SUMMARY: I. Introduction. II. Re-reading The Concept

1 Campbell, Tom, Prescriptive Legal Positivism: Law, Rights and Democracy,

3 Hart, H. L. A., Essays in Jurisprudence and Philosophy, Oxford, Clarendon,

REWRITING HARTS POSTSCRIPT

Such a reading of The Concept of Law comes up against

that this reading of Hart, explains some of the appeal of

Punishment and Responsibility, Oxford, Clarendon Press, 1968.

REWRITING HARTS POSTSCRIPT

repeated efforts to undermine poor evaluative arguments

the outline of a morally desirable system of law rather than

REWRITING HARTS POSTSCRIPT

without recourse to evaluative judgments and first order

reading lies in his straightforward account of the transition

REWRITING HARTS POSTSCRIPT

that must be satisfied for a primary rule to be considered

not.18 While such evaluations are not unequivocal, they

REWRITING HARTS POSTSCRIPT

tages of social institutions is a recurring theme of Harts

REWRITING HARTS POSTSCRIPT

burdens, for fairness he notes (with an eye to the context of

Ibidem, pp. 158-59.

Yet, Hart is not required to stop there and in fact he does

REWRITING HARTS POSTSCRIPT