The Very Idea of Legal Positivism: Stanley L. Paulson

The very idea of legal positivism
Stanley L. Paulson
Introduction
Much in recent discussions on legal positivism suggests
that the controversy surrounding the notion turns on the
distinction between inclusive and exclusive legal positivism.1
As a point of departure in distinguishing them, the
separation principle is helpful.2 At the most general level, the
separation principle – as Kenneth Einar Himma neatly puts
it – denies ‘that there is necessary overlap’ between the law
and morality.3 The separation principle counts, then, as the
contradictory of the morality principle, according to which
there is ‘necessary overlap’ between the law and morality,
however this might be explicated.4 What the legal positivist’s
1 An excellent discussion, chock full of arguments, is Kenneth Einar Himma,
‘Inclusive Legal Positivism’, in The Oxford Handbook of Jurisprudence and
Legal Philosophy, ed. Jules Coleman et al. (Oxford: Oxford UP, 2002), 125-65.
See, too, the detailed statements in Matthew H. Kramer, In Defense of Legal
Positivism (Oxford: Oxford UP, 1999); Kramer, Where Law and Morality Meet
(Oxford: Oxford UP, 2004).
2 H.L.A. Hart, ‘Positivism and the Separation of Law and Morals’, Harvard
Law Review, 71 (1957-8), 593-629, repr. in Hart, Essays in Jurisprudence and
Philosophy (Oxford: Clarendon Press, 1983), 49-87.
3 Himma, ‘Inclusive Legal Positivism’ (n. 1), 125.
4 ��
The claimed ‘necessary overlap’ between the law and morality is under-
Revista Brasileira de Estudos Políticos | Belo Horizonte | n. 102 | pp. 139-165 | jan./jun. 2011
140 THE VERY IDEA OF LEGAL POSITIVISM
denial of the morality principle comes to can be refined, we

are told, by appealing to the distinction between inclusive
and exclusive legal positivism. Inclusive legal positivism
leaves open the possibility that in a given legal system
there may or may not be necessary overlap between the law
and morality, while exclusive legal positivism recognizes
no possibility of necessary overlap.5 Again – this time in
Matthew H. Kramer’s words – ‘[t]he separability of the legal
realm and the moral realm, as opposed to their ineluctable
separation, is the condition [that the inclusive legal positivist]
seeks to highlight.’6
One can arrive at a broader perspective by opening
up the field to cover not only inclusive and exclusive legal
positivism but also non-positivism, represented by the
defence of the morality principle,7 according to which, again,
stood in a great variety of ways, as illustrated by prominent examples in

the recent literature. See e.g. Gustav Radbruch, ‘Statutory Lawlessness and
Supra-Statutory Law’ (first publ. 1946), trans. Bonnie Litschewski Paulson
and Stanley L. Paulson, Oxford Journal of Legal Studies, 26 (2006), 1-11; John
Finnis, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980), 2nd
edn. (Oxford: Oxford UP, 2011); Finnis, Philosophy of Law (Collected Es-
says, vol. IV) (Oxford: Oxford UP, 2011); Robert Alexy, The Argument from
Injustice (first publ. 1992), trans. Bonnie Litschewski Paulson and Stanley
L. Paulson (Oxford: Clarendon Press, 2002); Lon L. Fuller, ‘The Forms and
Limits of Adjudication’, Harvard Law Review, 92 (1978-9), 353-409, repr. (with
omissions) in Fuller, The Principles of Social Order, ed. Kenneth Winston, 2nd
edn. (Oxford: Hart Publishing, 2001), 101-39.
5 See Joseph Raz, The Authority of Law (Oxford: Clarendon Press, 1979); Raz,
Ethics in the Public Domain (Oxford: Clarendon Press, 1994), at 210-27 et pas-
sim.
6 Kramer, In Defense of Legal Positivism (n. 1), 114 (emphasis added).
7 The nomenclature ‘legal non-positivism’ is congenial in suggesting a generic
reading of the term, such that legal positivism and legal non-positivism
together exhaust the field. That is, on the generic reading ‘legal positiv-
ism’ and ‘legal non-positivism’ are correctly read as contradictories. For an
illuminating statement of various legal theories, grouped together under
the ‘positivist’ and ‘non-positivist’ rubrics, see Alexander P. d’Entrèves,
‘Two Questions about Law’, in Existenz und Ordnung. Festschrift für Erik
STANLEY L. PAULSON 141
there is necessary overlap between the law and morality. It

is clear that any two of these three views stand in a relation
of contrariety.8 For example, the cover statements giving
expression to non-positivism and inclusive legal positivism
cannot both be true, but they might well both be false, and
then the cover statement giving expression to exclusive legal
positivism would be true.
Say what you will about inclusive versus exclusive legal
positivism – some defend the distinction, others dismiss
inclusive legal positivism as a non-starter.9 I want in any
case to argue that a more fundamental distinction within
the positivist camp lies elsewhere. The distinction I have
in mind is that between legal positivism qua naturalism
and legal positivism without naturalism. Even though,
for institutional reasons, legal positivism has largely been
discussed in a vacuum, there is a standing presumption that
there are ties between legal positivism and ‘positivism writ
large’ in the greater philosophical tradition – or, as it would
be put in present-day philosophical circles, ties between
legal positivism and naturalism. What sorts of ties? In the
first two parts of the paper, I offer an answer. In Part One,
I draw on John Austin’s legal philosophy and argue that
it reflects the greater philosophical rubric, positivism writ
large or – my substitution again – naturalism. And, in Part
Two of the paper, I address and defend my substitution of
naturalism for positivism writ large.
Wolf zum 60. Geburtstag, ed. Thomas Würtenberger et al. (Frankfurt:
��
Klos-
termann, 1962), 309-20, repr. in d’Entrèves, Natural Law, 2nd edn. (London:
Hutchinson, 1970), 173-84.
8 See Robert Alexy, ‘On the Concept and the Nature of Law’, Ratio Juris, 21
(2008), 281-99, at 285-7.
9 For criticism, see e.g. Stefano Bertea, ‘A Critique of Inclusive-Positivism’,
Archiv für Rechts- und Sozialphilosophie, 93 (2007), 67-81; Scott Shapiro, ‘Law,
Morality, and the Guidance of Conduct’, Legal Theory, 6 (2000), 127-70.
Kramer replies to Shapiro in Where Law and Morality Meet (n. 1), at 45-75.
Specifically, in Part One of the paper, two theses are of

special interest, with the second thesis following from the
first. My first thesis: Austin’s naturalism – his reduction,
at two junctures, of ostensibly juridico-normative concepts
to matters of fact (namely, to habit and to fear) – is, as he
contends, sufficient to make out his case on the nature of
law. My second thesis, following from the first: If Austin’s
move is sufficient, then no thesis respecting a non-contingent
link between morality and the law can be necessary to
the explication of the nature of law. Taken together, these
two theses make a point, I should like to think, of genuine
significance. That is, if these two theses are indeed correct and
if Austin’s legal philosophy is representative of traditional
legal positivism, then the celebrated separation principle
is not doing the lion’s share of the work in legal positivist
circles after all. Rather, the separation principle is simply a
corollary of naturalism, the overriding view.
In Part Two of the paper, I take up the substitution of
naturalism for positivism writ large. In first thinking about
how to sort out species of legal positivism, I assumed I
would be working with positivism writ large as the greater
philosophical stage on which legal positivism finds its
place. A fair bit of reading disabused me of this notion. To
be sure, to speak of positivism writ large might well be ap-
propriate if I were directing my remarks to developments
in philosophy in, say, the mid-nineteenth century. At that
point in time, the older Hegelian consensus in Europe had
been altogether displaced by scientific positivism. I am
thinking, for example, of Hermann von Helmholtz, known
for his pioneering work in physics and physiology, as well
as for his efforts in recasting Kant’s theory of knowledge in
a modern, that is, positivistic idiom.10 All of this in the mid-
10 See e.g. Hermann von Helmholtz, ‘Über das Sehen des Menschen’ (lecture
nineteenth century.11 By contrast, ‘positivism’ is a term of

abuse in philosophical circles today. Jürgen Habermas writes
that positivism in philosophy proceeds from ‘scientistic
presuppositions’,12 and Bernard Williams writes that ‘to fall
back on positivism’ with an eye to avoiding interpretation
is ‘an offence against truthfulness’.13
In our time, the view that continues to enjoy a great
reception in philosophical circles is naturalism. Willard
Van Orman Quine, the ‘father of contemporary naturalism’
as one writer calls him,14 understands naturalism as the
appeal to the sciences. Naturalism, Quine tells us, assimilates
epistemology to ‘empirical psychology’.15 Quine’s view is
not, however, the only view of naturalism. Naturalism is
greater than Quine, thanks not least of all to the extraordinary
role he played in begetting it in its contemporary form.
Quine’s view counts today as one prominent characterization
in Königsberg 1855), in Helmholtz, Vorträge und Reden, 4th printing, 2 vols.

(Braunschweig: Friedrich Vieweg, 1896), vol.1, 85-117.
11 Helmut Holzhey öffers a three-fold characterization of philosophical
positivism at mid-century: first, knowledge stemming from the sciences is
privileged, while the philosopher’s claims respecting knowledge are dis-
puted, second, knowledge of reality (Wirklichkeitserkenntnis) is restricted to
what can be drawn from sense experience, and, third, thought is understood
solely in terms of the ‘subjective’ function of interpretation along with the
ordering of the elements of sense experience. Helmut
��
Holzhey, ‘Der Neu-
kantianismus’, in Helmut Holzhey and Wolfgang Röd, Die Philosophie des
ausgehenden 19. und des 20. Jahrhunderts [Teil] 2. Neukantianismus, Idealismus,
Realismus, Phänomenologie (Munich: C. H. Beck, 2004), 30.
12 Jürgen Habermas, Knowledge and Human Interests (first publ. 1968), trans.
Jeremy J. Shapiro (Boston: Beacon Press, 1971), 88.
13 Bernard Williams, Truth and Truthfulness (Princeton: Princeton UP, 2002),
12.
14 Penelope Maddy, Second Philosophy. A Naturalistic Method (Oxford: Oxford
UP, 2007), 4. A single important figure in contemporary legal philosophy
writes expressis verbis within the framework of Quinean naturalism, namely:
Brian Leiter, Naturalizing Jurisprudence (Oxford: Oxford UP, 2007).
15 See n. 31 below.
of naturalism, and David Hume’s view counts as another. I

return to Hume in Part Two of the paper.
Finally, in Part Three of the paper, I turn to legal
positivism without naturalism. Here the overriding figure
is Hans Kelsen. While Kelsen is of course defending the
separation principle, his position represents a wholesale
rejection of naturalism, which, he insists, is wrong-headed.
Thus, the idea that the separation principle is but a corollary
of naturalism can scarcely be attributed to him.
My greater thesis, then, is that Austin and Kelsen
represent two poles within legal positivism, namely, legal
positivism qua naturalism and legal positivism without
naturalism. The position represented by Hans Kelsen’s
colossus16 is, I think, peculiar to him. By contrast, any of a
host of other figures in the jurisprudential tradition can be
substituted for John Austin, a point to which I return.
1. John Austin
It is no accident that Austin’s statement of the separation
principle is found in a footnote to the text of Lecture V – a
fairly lengthy footnote, to be sure, where Austin carefully
sets the stage for a reply to William Blackstone:
Sir William Blackstone…says in his ‘Commmentaries’, that the
laws of God are superior in obligation to all other laws…that
human laws are of no validity if contrary to them…. Now, he may
mean that all human laws ought to conform to the Divine laws.
If this be his meaning, I assent to it without hesitation…. But the
meaning of this passage of Blackstone, if it has a meaning, seems
rather to be this: that no human law which conflicts with the Divine
law is obligatory or binding; in other words, that no human law
16 The editor of the Hans Kelsen Werke (Tübingen: Mohr Siebeck, 2007 ff.),
Matthias Jestaedt, estimates that Kelsen’s published writings run to 17,500
pages.
which conflicts with the Divine law is a law…. Now, to say that
human laws which conflict with the Divine law are not binding,
that is to say, are not laws, is to talk stark nonsense.17
Relegating to a footnote what we, influenced by

H. L. A. Hart, are accustomed to calling the separation
principle suggests that the real thrust of Austin’s position lies
elsewhere. As indeed it does. Austin, in Lecture VI, devotes
a good bit of attention to the straightforward reduction of
the doctrine of sovereignty to concatenations of fact. His
lines on habitual obedience are familiar:
The superior which is styled sovereignty…is distinguished…
by the following marks or characters: –1. The bulk of the given
society are in a habit of obedience or submission to a determinate
and common superior…[and] 2. That certain individual, or that
certain body of individuals, is not in a habit of obedience to a
determinate human superior.18
And, lest the significance of the appeal to habit be

missed, Austin repeats the point a number of times in Lecture
VI.19
With this scheme of Austin’s, we have the makings
of the central argument I wish to attribute to him. If his
conceptual repertoire is traceable back to the doctrine
of sovereignty and if sovereignty is reducible in turn to
concatenations of fact, then, Austin is arguing, this is
sufficient to explain the ostensibly normative material of the
17 John Austin, Lectures on Jurisprudence (first publ. 1863), 5th edn., 2 vols., ed.
Robert Campbell (London: John Murray, 1885), vol. 1, Lecture V (at pp. 214-
15) (emphasis in original), also in John Austin, The Province of Jurisprudence
Determined (first publ. 1832), ed. H. L. A. Hart (London: Weidenfeld and
Nicolson, 1954), Lecture V (at pp. 184-5) (emphasis in original).
18 Austin, Lectures (n. 17), Lecture VI (at p. 220) (emphasis in original); Austin,
Province (n. 17), Lecture VI (at pp. 193-4).
19 See Austin, Lectures (n. 17), Lecture VI (at e.g. pp. 222, 223-4, 227); Austin,
Province (n. 17), Lecture VI (at e.g. pp. 195, 198-9, 202-3).
law. And if this reduction of sovereignty to fact is indeed

sufficient, then, by hypothesis, no appeal to morality can be
necessary. In other words, Austin has in effect built right
into his reduction the thesis that there cannot be ‘necessary
overlap’ between the law and morality. And there is no
reason for the proponent of such a theory to pay special
attention to a separation principle, which has no standing
as an independent doctrine in the theory.
Before continuing with Austin, I want to underscore
the general import of what I am drawing from his theory.
Given the prominence of the separation principle as the
underlying notion in the myriad Anglo-American defences
of legal positivism over the past half century, its absence –
with a single important exception – from the lively European
debate over legal positivism a hundred years ago strikes one,
at any rate on first glance, as puzzling.20 The straightforward
explanation, however, is this. A host of fin de siècle European
legal theorists, roughly identifiable as positivists, made the
very sort of move that Austin made. That is, they claimed
that facts of nature are sufficient to explain ostensibly
juridico-normative material, and since morality cannot,
then, be necessary, they had no occasion to talk about it. In
a word, their move was naturalistic.
A good illustration is found in the work of Georg
Jellinek, the most influential figure in public law theory
(Staatsrechtslehre) on the European Continent a hundred
years ago, translated in his own day into major indo-
European languages. In some circles, it is presumed that
Jellinek is a ‘normativist’, a ‘Neokantian’. A closer look at
the texts, however, shows clearly that Jellinek’s celebrated
doctrine, ‘the normative force of the factual’, reduces without
20 The exception to the rule is Hans Kelsen, and I take up his position in Part
Three.
remainder to matters physiological or psychological. And

the reduction is one of Jellinek’s own making. As he puts
it, the ‘normative import’ of the factual counts simply as
our physiological or psychological tendency to reproduce,
in our minds, that to which we have become accustomed.21
This is closer to Hume than to anything in Neokantianism.
Again, my thesis is that Austinian naturalism – the
move from the ostensibly normative material of the law to
concatenations of fact – is standard fare for legal positivists
generally. While my thesis may appear to be obvious – and I
would be pleased if it did – it is hardly the received opinion.
For example, in his celebrated paper ‘Positivism and the
Separation of Law and Morals’22 and again in The Concept of
Law,23 Hart sets out five different doctrines under the rubric
of legal positivism – command, separation, analysis, judicial
decisions as logical deductions, and non-cognitivism. He
attributes the first three of these to Jeremy Bentham and
Austin. The doctrine that I claim is fundamental – naturalism
– does not turn up on Hart’s list, and it is not implied by
anything that does turn up there.
21 Jellinek writes: ‘To seek the basis of the normative force of the factual in
its conscious or unconscious reasonableness would be utterly mistaken.
The factual can be rationalized later, but its normative import lies in an
underived property of our nature, on the strength of which something we
are already accustomed to is physiologically and psychologically easier to
reproduce than something new.’ Georg Jellinek, Allgemeine Staatslehre, 2nd
edn. (Berlin: O. Häring, 1905), 330, 3rd edn. (1914), 338. The reduction to
fact, in Jellinek’s work at this juncture, is captured effectively by Michael
Stolleis, Public Law in Germany 1800-1914 (first publ. 1992), trans. Pamela
Biel (New York and Oxford: Berghahn, 2001), 442-3.
22 See Hart, ‘Positivism and the Separation of Law and Morals’ (n. 2), 601-2
at n. 25, repr. in Hart, Essays in Jurisprudence and Philosophy (n. 2), 57-8 at n.
25.
23 See H. L. A. Hart, The Concept of Law, 2nd edn. (Oxford: Clarendon Press,
1994), 302, at note pertaining to p. 185.
Back to Austin, as promised. I have spoken of tracing

the whole of Austin’s conceptual repertoire to the doctrine
of sovereignty and of reducing sovereignty in turn to
concatenations of fact. A comparable argument can be
generated by looking to Austin’s command doctrine, implicit
in the doctrine of sovereignty.24 The doctrine of command,
on one reading of Austin’s treatise, can be understood in
terms of three components: the commander’s intention
that a party act (or forbear from acting) in a particular way,
the commander’s expression of this intention to the party,
and – central to the doctrine – the commander’s power to
impose a sanction should the commandee fail to comply
with the directive.25 The power to impose a sanction is not,
however, to be understood as a property of the commander,
for a commander, characterized in the Austinian theory in
terms of the power to impose sanctions, might not have
such power over the particular party to whom he issues his
directive, as in the case of a sovereign’s putative command
to another sovereign. The power to impose a sanction is
to be understood, in other words, as a relation between
commander and commandee or, more generally, between
superior and inferior.26 Austin gives expression to the
relation – we might call it the ‘power relation’ – when he
writes: “The term superiority signifies might: the power
of affecting others with evil or pain, and of forcing them,
through fear of that evil, to fashion their conduct to one’s
wishes”.27
24 See Austin, Lectures (n. 17), Lecture V (e.g. at p. 177); Austin, Province (n.
17), Lecture V (e.g. at p. 132).
25 Austin, Lectures (n. 17), Lecture I (at p. 91, and see p. 89); Austin, Province
(n. 17), Lecture I (at p. 17 and see pp. 13-14).
26 Austin, Lectures (n. 17), Lecture I (at pp. 96-7); Austin, Province (n. 17),
Lecture I (at pp. 24-5).
27 Austin, Lectures (n. 17), Lecture I (at p. 96) (emphasis added) (see also p. 90:
Fear, a brute fact, is the operative notion here, and the

argument proceeds just as before. If Austin’s ‘correlative
terms’ of obligation and sanction are traceable back to their
correlative, the command,28 and if the command is reducible
in turn to concatenations of fact – in particular, to the
commandee’s fear – then, so Austin, this move is sufficient
to explain the ostensibly normative material of the law. 29
And if this move is sufficient, then no appeal to morality
can be necessary.
It is useful to dwell for just a moment on the concept of
fear. Just as no one would claim that sexual desire is acquired
through reasoning or is the product of experience, so likewise
for fear. Notwithstanding the fact that experience may shape
our responses on both fronts, the phenomena themselves
have a basis independent of experience.30 Hume speaks of
natural instinct. This mention of Hume brings me to Part
Two, the substitution of naturalism for positivism writ large.
‘that which is not feared is not apprehended as an evil’); Austin, Province

(n. 19), Lecture I (at p. 24) (emphasis added) (see also p. 16).
28 On ‘correlativity’, see Austin, Lectures (n. 17), Lecture I (at pp. 89, 96, see
also pp. 91-2); Austin, Province (n. 17), Lecture I (at pp. 14, 24, see also pp.
17-18).
29 As noted by Hart, Austin also has a second interpretation of the command.
It turns on his definition of obligation ‘as the “chance or likelihood” that one
who has been commanded to do or abstain from doing something would
suffer some evil in the event of disobedience’. See H. L. A. Hart, ‘Analytical
Jurisprudence in Mid-Twentieth Century: A Reply to Professor Bodenhe-
imer’, University of Pennsylvania Law Review, 105 (1956-7), 953-75, at 965;
H. L. A. Hart, ‘Legal and Moral Obligation’, in Essays in Moral Philosophy,
ed. A. I. Melden (Seattle: University of Washington Press, 1958), 82-107, at
95-9; Hart, The Concept of Law (n. 23), 282 note at (c), 290 note pertaining to
p. 83. For Austin’s own text on ‘chance or likelihood’, see Austin, Lectures
(n. 17), Lecture I (at p. 90); Austin, Province (n. 17), Lecture I (at p. 16).
30 I owe the example to H. O. Mounce, Hume’s Naturalism (London and New
York: Routledge, 1999), 62.
2. The substitution of naturalism for positivism

writ large
I remarked in the Introduction that Quine, in the name
of naturalism, would have us appealing to the sciences. Epis-
temology becomes ‘empirical psychology’.31 Although there
are great differences between Quine’s naturalistic enterprise
and David Hume’s, there are similarities, too. For example,
many regard Hume’s theory of human nature in Book III of
A Treatise on Human Nature as a study in moral psychology.
As one prominent interpreter of Hume puts it, ‘[t]o a large
extent, Hume’s theory of human nature is not, in our terms,
philosophical, but psychological’.32 Hume’s famous – some
will say ‘notorious’ – dictum that ‘reason is, and ought only
to be the slave of the passions’,33 is most helpfully seen as
his response to the elevated role played by reason in the
rationalist philosophies of the Cartesian tradition.34 Hume’s
tack is diametrically opposed. He looks inward:
Take any action allow’d to be vicious: Wilful murder, for instance.
Examine it in all lights, and see if you can find that matter of fact,
31 ��
‘[T]he epistemological question [is] a question within science: [how hu-
mans] have managed to arrive at science from such limited information.
Our scientific epistemologist pursues this inquiry…. Evolution and natural
selection will doubtless figure in this account, and he will feel free to apply
physics if he sees a way.’ W. V. O. Quine, ‘Five Milestones of Empiricism’
(lecture of 1975), in Quine, Theories and Things (Cambridge, Mass.: Harvard
UP, 1981), 67-72, at 72.
32 Terence Penelhum, ‘Hume’s Moral Psychology’, in Cambridge Companion
to Hume, ed. David Fate Norton (Cambridge: Cambridge UP, 1993), 117-49,
at 119. Penelhum is not alone. See also Jerry A. Fodor, Hume Variations
(Oxford: Clarendon Press, 2003), at 1-27 et passim, who suggests that Hume’s
naturalism anticipates current work in cognitive science.
33 David Hume, A Treatise on Human Nature (first publ. 1739-40), 2nd edn., ed.
P.H. Nidditch (Oxford: Clarendon Press, 1978), II.iii.3. (at p. 415).
34 See Penelhum, ‘Hume’s Moral Psychology’ (n. 32), 119-20, on which I have
drawn here.
or real existence, which you call vice. In which-ever way you take
it, you find only certain passions, motives, volitions and thoughts.
There is no other matter of fact in the case. The vice entirely
escapes you, as long as you consider the object. You never can
find it, till you turn your reflexion into your own breast, and find
a sentiment of disapprobation, which arises in you, towards this
action. Here is a matter of fact; but ’tis the object of feeling, not
of reason. It lies in yourself, not in the object. So that when you
pronounce any action or character to be vicious, you mean nothing,
but that from the constitution of your nature you have a feeling
or sentiment of blame from the contemplation of it.35
These notions – feeling, sentiment, instinct, the

constitution of our nature – have one looking inward. And
this, Hume would have us believe, is the source of our
psychological explanations.
This view of ‘Hume the naturalist’ comes as a surprise
to those who take their cues from a textbook account of
Hume, which has him following his empiricist predecessors,
Locke and Berkeley, while recognizing – and making the
most of – the scepticism to which their view inevitably leads.
The argument on behalf of scepticism is familiar. Empiricism
has its source in sense experience. Beliefs that stem from
sense experience do not lend themselves to justification. A
justification requires an appeal to something independent,
but there is no way, so to speak, of stepping outside sense
experience in order to appeal to something independent of
it. The result is scepticism.
This point, in explicit criticism of Hume’s scepticism,
stems, inter alia, from the philosopher Thomas Reid.
Defending his own view, Reid argues that sensory experience
is not ‘what we perceive’ but rather that ‘whereby we
perceive’.36 Norman Kemp Smith – in path-breaking papers
35 Hume, Treatise (n. 35), III.i.1. (at pp. 468-9) (emphasis in original).
36 See The Works of Thomas Reid, 8th edn., 2 vols., ed. William Hamilton (Edin-
a hundred years ago and in an extraordinary treatise on

Hume some seventy years ago – stands Reid’s interpretation
of Hume on its head.37 In Kemp Smith’s splendid words:
Hume ‘is depicted as having done no more than deliver his
successors from a bondage to which he himself remained
subject. A strangely paradoxical verdict!’38 Hume, on Kemp
Smith’s interpretation, was keenly aware of the scepticism
inherent in traditional empiricism and sought to provide an
alternative. The alternative, naturalism, is found in Book III
of Hume’s Treatise. Indeed, Kemp Smith goes on, the best
way to read Hume is to begin with Book III of the Treatise
before turning to Book I, whose scepticism will then properly
be seen as qualified by Hume’s naturalism.
3. Positivism without naturalism. The case of

Hans Kelsen
Where legal positivism qua naturalism is the point of
departure, Hans Kelsen is the spoiler. Other legal positivists
count as naturalists, arguing that since the facts are sufficient
to explain ostensibly normative material, morality cannot be
necessary. Kelsen offers no such argument, for he, unlike all
the others, is not arguing that ostensibly juridico-normative
material is reducible to fact. Kelsen defends what he terms
a normative legal philosophy.
burgh: James Thin, 1895), vol. 1, pp. 108, 112, 117, 121 et passim. The quota-
tion in the text is Mounce’s tidy summary statement, in Mounce, Hume’s
Naturalism (n. 30), 1, 54. Reid’s philosophy is presented in rich detail in
Keith Lehrer, Thomas Reid (London and New York: Routledge, 1989).
37 See, in particular, Norman Kemp Smith, The Philosophy of David Hume
(London: Macmillan, 1941). The early papers are ‘The Naturalism of Hume
(I.)’, Mind, 14 (1905), 149-73, and ‘The Naturalism of Hume (II.)’, ibid. pp.
335-47.
38 Kemp Smith, The Philosophy of David Hume (n. 37), 3.
Normativity,39 Kelsen tells us, is his alternative to other

approaches within legal philosophy, but – the rub – there
has never been any agreement on what he means here. The
interpretations of Kelsen’s idea of normativity run the gamut,
from a counter-factual interpretation of normativity40 to a
‘justified normativity’ thesis. The justified normativity thesis
is far and away the most ambitious reading of normativity in
Kelsen’s legal philosophy. It has been attributed to Kelsen
in different forms with different sorts of argument by no
fewer than four leading figures – Robert Alexy, Carlos
Santiago Nino, Joseph Raz, and, a bit earlier, by Alf Ross
–41 with, so far as I can tell, each writer developing his own
case independently of the others. I confine myself here to
39 Normativity, for many purposes a concept in its own right, has enjoyed a
good bit of attention in recent philosophy. See e.g. Joseph
��
Raz, ‘E��
xplain-
ing Normativity: On Rationality and the Justification of Reason’, Ratio 12
(N.S.) (1999), 354-379, repr. in Raz, Engaging Reason (Oxford: Oxford UP,
1999), 67-89. See also Alan Millar, Understanding People. Normativity and
Rationalizing Explanation (Oxford: Clarendon Press, 2004); John Skorupski,
The Domain of Reasons (Oxford: Oxford UP, 2010).
40 In an overview of Kelsen’s work, Robert Walter writes that coercive systems,
legal systems in particular, are to be interpreted ‘as if they were normative’.
Walter, ‘Der gegenwärtige Stand der Reinen Rechtslehre’, Rechtstheorie, 1
(1970), 69-95, at 70 (emphasis in original).
41 See Alexy, The Argument from Injustice (n. 4), at 95-123; Carlos Santiago
Nino, ‘Some Confusions surrounding Kelsen’s Concept of Validity’, Ar-
chiv für Rechts- und Sozialphilosophie, 64 (1978), 357-77, at 357-65, repr. in
Normativity and Norms. Critical Perspectives on Kelsenian Themes [hereafter:
NN], ed. Stanley L. Paulson and Bonnie Litschewski Paulson (Oxford:
Clarendon Press, 1998), 253-61; Nino, La validez del derecho (Buenos Aires:
Editorial Astrea, 1985), at 7-40 et passim; Joseph Raz, ‘Kelsen’s Theory of
the Basic Norm’, American Journal of Jurisprudence, 19 (1974), 94-111, repr.
in NN (this note, above), 47-67, and in Raz, The Authority of Law, 2nd edn.
(Oxford: Oxford UP, 2009) [hereafter: AL], 122-45; Alf Ross, ‘Validity and
the Conflict between Legal Positivism and Natural Law’, Revista Jurídica
de Buenos Aires, 4 (1961), 46-93 (bilingual printing), at 82, and see generally
at 78-82, repr. in NN (this note, above) 147-63, at 160, and see generally at
159-61.
Raz, whose statement is in some respects the most extreme

of the four. Raz begins by contrasting Hart’s position with
Kelsen’s. H. L. A. Hart is a proponent of social normativity,
understanding ‘[t]he normativity of the law and the
obligation to obey it [as] distinct notions.’ An altogether
different understanding, Raz continues, is evident in the
work of one who recognizes ‘only the conception of justified
normativity’,42 namely, Hans Kelsen. In characterizing
justified normativity, Raz writes: “[T]o judge the law as
normative is to judge it to be just and to admit that it ought
to be obeyed. The concepts of the normativity of the law
and of the obligation to obey it are analytically tied together.
Kelsen, therefore, regards the law as valid, that is, normative,
only if one ought to obey it”.43
Of course Raz sees the paradox in attributing justified
normativity to Kelsen, a thesis that places Kelsen far closer
to natural law theory than to anything found in traditional
or naturalistic legal positivism. Indeed, Raz invites attention
to the paradox, writing that although ‘Kelsen rejects natural
law theories, he consistently uses the natural law concept of
normativity, i.e. the concept of justified normativity.’44
Interpreters of a philosopher will of course pursue
what they deem to be the most promising reading of the
philosopher’s work. So far, so good, but with a caveat. Merit
lies in pursuing what the philosopher actually wrote rather
than in imposing on the philosopher’s text an interpretation
‘from without’, so to speak. As Paul W. Franks puts it in his
book on the post-Kantians, ‘[i]f we assume that historical
42 Raz, ‘Kelsen’s Theory of the Basic Norm’ (n. 41), 105, in NN (n. 41), 60, in
Raz, AL (n. 41), 137.
43 Ibid.
44 Raz, ‘Kelsen’s Theory of the Basic Norm’ (n. 41), 110-11, repr. NN (n. 41),
67, and in Raz, AL (n. 41), 144.
figures are asking or answering our questions’, we ‘run

the risk of both distorting what they say and missing an
opportunity to learn from them, whether positively or
negatively.’45
Raz, like Alexy, Nino, and Ross, can reply that he is
drawing an interpretation from the text, not imposing an
interpretation on it. The reply is a good one as far as it goes,
but it gives rise to the question of just how representative
the passages selected by Raz and the others are. As I have
argued at length elsewhere, the passages selected are in fact
not representative of Kelsen’s work. And then the point
made by Franks stands: We learn best from historical figures
when we address the questions they themselves were asking
and answering.
What, then, takes the place of justified normativity?
My answer: Kelsen’s project over many decades was, above
all, an ambitious and far-reaching attempt, first, to show that
naturalism in fin de siècle legal science is mistaken, and, second,
to develop the rudiments of an alternative theory that would
secure the autonomy (Eigengesetzlichkeit) of the law and, by
the same token, the purity (Reinheit) of legal science. And
this takes us full circle back to the question of normativity.
Kelsen’s alternative to naturalism yields a normativity thesis,
and this should come as no surprise. He has to be committed
to something that is normative in character lest he have no
alternative to naturalism after all. In sharp contrast to the
45 Paul W. Franks, All or Nothing. Systematicity, Transcendental Arguments, and

Skepticism in German Idealism (Cambridge, Mass., and London: Harvard UP,
2005), 5. Of course this idea is hardly new; I quote from Franks because his
statement of the matter is unusually perspicuous. To the same effect, see
John Rawls, Lectures on the History of Political Philosophy (Cambridge, Mass.:
Harvard UP, 2007), at 251: ‘[I]n studying the works of the leading writers
in the philosophical tradition, one guiding precept is to identify correctly
the problems they were facing, and to understand how they viewed them
and what questions they were asking.’
thesis of justified normativity, however, Kelsen’s normativity

thesis is part and parcel of his greater effort to develop an
alternative to naturalism and thereby to lend respectability
to legal science, underscoring its nomological dimension. I
call Kelsen’s thesis the nomological normativity thesis. As
Kelsen understands the thesis, its import is underscored by
a ‘law-like’, necessary or nomological connection at the very
core of his legal philosophy. In what follows, my primary
concern is to invite attention to this connection.
Kelsen’s alternative to naturalism is captured in my
reconstruction by the concept of peripheral imputation. ‘To
impute’ (Latin imputare) means to bring into reckoning, to
ascribe, to attribute. Kelsen’s German verb is ‘zurechnen’, and
‘to impute’ is a reliable translation, not least of all in light of
Kelsen’s own occasional use of the loan-word ‘imputieren’
where ‘zurechnen’ might have been expected.46
Kelsen has two doctrines of imputation. The first
of these, central imputation, is by and large a reflection of
the philosophical tradition, though Kelsen’s use of central
imputation is anything but traditional.47 The second doctrine,
46 As Kelsen writes: ‘It would be a serious misunderstanding if one wanted

somehow to impute [imputieren] to these observations [on the legal authority
of administrative agencies] the significance of a political mandate for the
greatest possible restriction of the state’s administrative activity.’ Kelsen,
Hauptprobleme der Staatsrechtslehre (Tübingen: J.C.B. Mohr, 1911), 503, and
see at 138, 194, 209, repr. in Hans Kelsen Werke, ed. Matthias Jestaedt, vol.
2 (Tübingen: Mohr Siebeck, 2008), 650, and see at 244, 306, 322. See also
Hans Kelsen, Über Grenzen zwischen juristischer und soziologischer Methode
(Tübingen: J. C. B. Mohr, 1911), at 44.
47 Central imputation serves in Kelsen’s very early work as an escape hatch
from naturalism and psychologism. Later he turns to the basic norm,
already evident in Das Problem der Souveränität und die Theorie des Völker-
rechts (Tübingen: J.C.B. Mohr, 1920), and to a Neokantian transcendental
argument, and these steps represent his effort to replace central imputation
with something more satisfactory. To be sure, central imputation survives
elsewhere in Kelsen’s work. His most extensive discussion of both imputa-
peripheral imputation, is peculiar to Kelsen. Both doctrines

purport to offer an alternative to causal explanation, and
both, for this reason, are of unusual significance in Kelsen’s
legal philosophy – central imputation in Kelsen’s very early
work, and peripheral imputation thereafter. I shall confine
my discussion to the latter, peripheral imputation, for this is
the doctrine that underlies Kelsen’s nomological normativity
thesis.
Kelsen tells us that peripheral imputation links ‘material
facts’ (Tatbestände). As he puts it in the Allgemeine Staatslehre,
in what counts as an early statement of the doctrine: ‘[P]
eripheral imputation always leads from one material fact to
nothing other than another material fact.’48 A comparable
statement is found in the first edition of the Reine Rechtslehre.
At the end of a section devoted to the doctrine of central
imputation, Kelsen contrasts that doctrine with peripheral
imputation. Central imputation, he writes,
is an entirely different operation from the peripheral
imputation mentioned earlier, where a material fact is
connected…to another material fact within the system,
that is, where two material facts are linked together in the
reconstructed legal norm.49
Two questions arise. What exactly would Kelsen have
us understand by material facts? And how might peripheral
imputation, linking material facts, be formulated? As to the
first question, Kelsen answers in terms of ‘legal condition’
tion doctrines – central and peripheral – is found in a lengthy essay: Kelsen,

‘Unrecht und Unrechtsfolge im Völkerrecht’, Zeitschrift für öffentliches Recht,
12 (1932), 481-608, see at §§ 1-2 (pp. 481-504), § 5 (pp. 525-9), § 7 (pp. 537-44).
48 Hans Kelsen, Allgemeine Staatslehre (Berlin: Julius Springer, 1925), § 12 (d)
(p. 65).
49 Hans Kelsen, Introduction to the Problems of Legal Theory, a translation of
the first edition of Kelsen’s Reine Rechtslehre (1934) by Bonnie Litschewski
Paulson and Stanley L. Paulson (Oxford: Clarendon Press, 1992) [hereafter:
LT], § 25 (d) (pp. 50-1).
and ‘legal consequence’, more precisely, in terms of the

state of affairs counting as the legal condition in a particular
instance and, in Hohfeldian parlance, the legal position that
emerges as its legal consequence. This seems to be an odd
fit, for we do not ordinarily think of a legal consequence
as a material fact (Tatbestand). Rather, in a hypothetically
formulated legal norm, a material fact falling within the
scope of the antecedent clause of the norm triggers the legal
consequence, establishing, in Kelsen’s doctrine, the legal
position of liability that counts as the legal consequence.
It is, however, material facts, thus understood, that
Kelsen brings together in introducing peripheral imputation.
He writes that ‘[i]f the mode of linking material facts is
causality in the one case, it is imputation in the other.’50 What
is more, he uses ‘legal condition’ and ‘legal consequence’
alongside ‘cause’ and ‘effect’ as the respective relata of these
very same ordering principles or relations, imputation and
causality.51 That is, he understands their relata as species of
the genus ‘material fact’.
To shed light on Kelsen’s expansion of the notion of
material facts as the relata of peripheral imputation, it is
perhaps helpful to point to his effort to provide as close a
parallel as possible to the principle of causality. Since he
assumes material facts to be indisputably the relata in the case
of causality, so likewise, he is arguing, material facts serve
as the relata in the case of peripheral imputation. Kelsen
wishes to underscore a law-like, necessary or nomological
relation in the law running parallel to the law-like, necessary
or nomological relation manifest in causality, and his
development of this parallel is a central part of his effort to
reply to naturalism and, by the same token, to turn the legal
50 Kelsen, LT (n. 49), § 11 (b) (p. 23) (emphasis added).

51 See Kelsen, LT (n. 49), § 11 (b) (at pp. 23-4).
science of his day into something scientifically respectable.52

If it can be shown that aspects of the fundamental ordering
principle of the natural sciences are reflected per analogiam
in the fundamental ordering principle of legal science, then,
so Kelsen, the parallel will indeed enhance the status of legal
science qua science.53
I turn now to the second question, which speaks to the
formulation of peripheral imputation. One proposal for a
formulation might read (with an ‘and if …’ clause inserted
between parentheses as a shorthand reference to the other
conditions associated with a legal proceeding):
Formulation I: If an act of a certain type takes place (and if),
then the actor or a surrogate54 is liable for that act.
This formulation is ruled out, however, by Kelsen’s
stipulation that peripheral imputation links material facts,
where the latter material fact is understood to be the liability
imputed to the legal act. To adopt formulation I as a
representation of peripheral imputation would be to confuse
peripheral imputation with central imputation.
The alternative is a ‘subjectless’ counterpart to
formulation I, that is to say, a formulation that does not
include an ascription to a legal subject:
Formulation II. If an act of a certain type takes place (and if),
then that act is treated as ‘liability ascribing’.
52 See generally Kelsen, Über Grenzen zwischen juristischer und soziologischer

Methode Grenzen (n. 46), at 1-15 et passim; Horst Dreier, Rechtslehre, Staatsso-
ziologie und Demokratietheorie bei Hans Kelsen, 2nd printing (Baden-Baden: No-
mos, 1990), at 1-15 et passim; Horst Dreier, ‘Hans Kelsen’s Wissenschaftspro-
gramm’, Die Verwaltung, Beiheft 7: Staatsrechtslehre als Wissenschaft, ed.
Helmuth Schulze-Fielitz (Berlin: Duncker & Humblot, 2007), 81-114.
53 I take up the parallel in the name of methodological forms, see the text at
nn. 61-6 below.
54 Here I am using ‘surrogate’ to cover all the variations on the theme of
vicarious and collective liability, see the text at nn. 58-9 below.
Its counterintuitive character aside, a point to which

I return below, formulation II captures the import of
peripheral imputation. A defensible formulation must
reflect a necessary connection between the two material
facts. And if the formulation is confined, as here, to the act
and to liability – the liability imputed to the act – then the
link is indeed necessary. As Kelsen writes:
If there is the necessity of an absolute ‘must’ when the law of nature
links cause and effect, so there is the equally rigorous ‘ought’
when the law of normativity (Rechtsgesetz) sets out the synthesis
of conditioning and conditioned material facts. In the sphere of
the law or in ‘legal reality’,…delict is linked to punishment with
the same necessity as, in the sphere of nature or in ‘natural reality’,
cause is linked to effect.55
This is close to being right as a statement of Kelsen’s

position, though one wrinkle has to be ironed out. Kelsen
cannot be claiming a necessary link between the delict and
the actual imposition of punishment. That would not make
good sense, for, as we know and as Kelsen makes perfectly
clear elsewhere, ‘in the system of nature, punishment may
fail to materialize for one reason or another’.56 It is not
punishment but criminal liability – and by the same token
civil liability – that figures in the law-like, necessary or
nomological link. In the most general terms, liability serves
in this formulation as the second relatum, the second ‘material
fact’. The relation of liability to the act to which it is imputed
is a necessary relation. By contrast, the actual imposition of
punishment in the criminal law and the actual execution of
judgment in the civil law is a contingent matter.57
55 Hans Kelsen, ‘“Foreword” to the Second Printing of Main Problems in the

Theory of Public Law’, trans. in NN (n. 41), 3-22, at 5 (in the last latter sentence,
the quotation marks are in the original text).
56 Kelsen, LT (n. 49), § 11 (b) (pp. 25).
57 The point made here can be compared with Hart’s argument directed
Still, formulation II seems counterintuitive in imputing

liability to the act, not to the actor. We are accustomed
to distinguishing between the imputation of liability
individually on the one hand and collectively on the other.58
In the first case, the imputation of liability is either to the actor
or, under the rubric of vicarious liability, to a surrogate. In
the second case, liability is imputed, say, to the insurance
company.
Why does Kelsen restrict himself to the imputation of
liability to the act rather than to the actor? Kelsen’s restriction
can be explained, I think, by the contingent element
presupposed in identifying the liable party. As understood
in this or that jurisdiction, the character of the liable party –
actor, surrogate, or collective body – is a contingent factor,
a question of legal policy, not legal science. This point
strengthens Kelsen’s hand in insisting that the necessary link
be limited to the imputation of liability to the act.
In any case, it is precisely this necessary relation
between act and liability that represents the core of what
I am calling Kelsen’s nomological normativity thesis. The
relation is nomological in being necessary or law-like, and
it is normative in being non-causal. Further permutations
stemming from imputation, thus understood, will then be
forthcoming where liability is ascribed to a person, triggering
to Austin’s claim that nullity is a sanction. Hart replies that nullity and
sanction are conceptually distinct. Specifically, he points out that a nullity
follows necessarily from the failure to satisfy the conditions of the legal
arrangement (Jones purports to marry Sally, but the ‘marriage’ is null and
void, for he is already married), whereas the actual imposition of a sanction
is a contingent matter. See Hart, CL (n. 23), at 33-5, and Austin, Lectures on
Jurisprudence (n. 17) Lecture XXIII (at p. 457), Lecture XXVII, (at pp. 505 f.).
58 See Kelsen, LT (n. 49), § 13 (at p. 27); Hans Kelsen, General Theory of Law
and State, trans. Anders Wedberg (Cambridge, Mass.: Harvard UP, 1945), at
59, 69-71; Hans Kelsen, Reine Rechtslehre, 2nd edn. (Vienna: Franz Deuticke,
1960) at § 28 (c) (pp. 125-6).
the empowerment of a legal organ to follow through with

whatever sanction is called for.
What remains now is to cast imputation in terms that
invite attention to the underpinnings of the nomological
normativity thesis. At some points in his work, Kelsen treats
imputation as a Kantian or Neokantian category by analogy
to the category of causation.59 The transcendental argument
that Kelsen adduces in the name of imputation qua category,
however, is not sound. If Kelsen nevertheless utilizes
imputation in his philosophy as in my sketch above, then
its foundation requires a closer look. My suggestion is that
Kelsen’s concept of peripheral imputation be conceptualized
as a methodological form, specifically, the methodological form
peculiar to legal science. The notion is drawn from the work
of the Baden Neokantian Heinrich Rickert.
In the last chapter of his treatise, The Object of
Knowledge,60 Rickert distinguishes the constitutive categories
of objective reality – for example, the category of permanence
– from the methodological forms of the various standing
disciplines. Rickert’s basic idea is that objective reality,
constituted transcendentally, must be sharply distinguished
from the processing (Bearbeitung) of the material given in
objective reality. Objective reality, Kant’s phenomenal
world, is constituted by means of the categories of reality,
while the processing of the material of objective reality is the
work of the standing disciplines, which are grounded in their
respective methodological forms. Rickert offers lawfulness
59 See Kelsen, LT (n. 49), § 11 (b) (at p. 23).

60 See Heinrich Rickert, Der Gegenstand der Erkenntnis, 2nd edn. (Tübingen and
Leipzig: J.C.B. Mohr, 1904), at 205-28, Rickert, Der Gegenstand der Erkenntnis,
6th edn. (Tübingen: J.C.B. Mohr, 1928), at 401-32. See also Heinrich Rickert,
Die Grenzen der naturwissenschaftlichen Begriffsbildung (first publ. 1902), 5th
edn. (Tübingen: J.C.B. Mohr, 1929), at 283-4, 373-7, et passim.
(Gesetzlichkeit) as an example of a methodological form in

the natural sciences.61 In fact, the example has to be taken
as the genus of methodological forms in the natural sciences
generally, for it has application to all of them.
In The Object of Knowledge, Rickert begins with the
constitutive categories of reality:
The unique significance of…the forms that have been discussed
in terms of the examples of causality and permanence requires
that they be given a special name, one that distinguishes them as
original forms in contrast to methodological forms. Building on
the expression ‘objective reality’, we could speak…of ‘objective
forms of reality’. But we prefer…the term ‘constitutive’. In that
these particular forms constitute what is presupposed as finished
product or as real material of cognition, ‘constitutive’ designates
exactly what we mean. Thus, the categories that shape the
objective, real world from what is in fact given should be called
the constitutive categories of reality.62
The methodological forms to which Rickert alludes are

peculiar, respectively, to the various standing disciplines.
Referring in his treatise to Cartesian dualism, Rickert writes:
This other species of dualism, according to which the
world is supposed to consist of two types of reality, each
excluding the other – the world of extensio and the world
of cogitatio – is created by physics and by psychology, each
with its respective methodological form.63
Physics has its own methodological form, and so does
psychology. Legal science, too, has its own methodological
form, namely, imputation or, as Kelsen sometimes puts it,
61 See Rickert, Der Gegenstand der Erkenntnis, 6th edn. (n. 60), at 409-10.
62 Ibid. at 406-7 (quotation marks and emphasis in original), compare Rickert,
Der Gegenstand der Erkenntnis, 2nd edn. (n. 60), at 211.
63 Rickert, Der Gegenstand der Erkenntnis, 6th edn. (n. 60), 424 (emphasis in
original), see also at 404, 410, 411, 424, 426, et passim, and compare Rickert,
Der Gegenstand der Erkenntnis, 2nd edn. (n. 60), at 208, 210, 217, 221, et passim.
the ‘law of normativity’.64 As he explains, looking back on

the theses he defended in Hauptprobleme:
[T]he core problem becomes the reconstructed legal norm,
understood as the expression of the specific lawfulness, the
autonomy, of the law, as the legal counterpart to the law of
nature (Naturgesetz) – the ‘law of the law’, so to speak, the law
of normativity (Rechtsgesetz) What is obviously of importance
in Main Problems is securing the objectivity of validity, without
which there can be no lawfulness whatever, let alone the specific
lawfulness, the autonomy, of the law. But without the expression
of that autonomy, without the law of normativity, there can be no
legal knowledge, no legal science. Therefore: objective judgment,
not subjective imperative. ‘The law of normativity is – outwardly
– like the law of nature, in that it is directed to no one and valid
without regard to whether it is known or recognized.’ If the
analogy between the law of normativity and the law of nature is
still fairly limited here, this is in order to prevent the confusion
of the two, indeed not to lose sight – because of the analogy – of
the specific lawfulness, the autonomy, of the law as against the
causal lawfulness of nature.65
The normative or non-naturalistic import of Kelsen’s

enterprise, the force of his law of normativity, plays itself
out in the context of nomological legal science, understood
as Kelsen’s alternative to psychologism and naturalism in
legal science. Specifically, the focus is on the methodological
form of legal science – the relation of peripheral imputation.
Where the antecedent condition obtains, this marks the
imputation of liability to the act, a necessary relation. Where
the ascription of liability to a person is made, this marks a
64 See the quotation immediately below.

65 Kelsen, ‘“Foreword” to the Second Printing of Main Problems in the Theory
of Public Law’ (n. 55), 5-6 (emphasis in original). Kelsen’s quotation within
the quotation is from the Hauptprobleme (n. 46), at 395, repr. in Hans Kelsen
Werke, vol. 2 (n. 46), at 529 (‘outwardly’ appears in italics in the Hauptprob-
leme but not in the ‘Foreword’ quoted here). See also Kelsen, LT (n. 49),
§ 11 (b) (at p. 23-5).
change in that person’s legal position. The change, Kelsen

insists, is a normative change, not a causal change.
4. Concluding remark
Coming full circle, back to the Introduction, I should like
once again to allude to the distinction between inclusive and
exclusive legal positivism, comparing it with the distinction
between legal positivism qua naturalism and legal positivism
without naturalism. The first distinction has inclusive legal
positivism riding piggyback on exclusive legal positivism.
That is, in all those legal systems correctly characterized by
means of the ‘exclusive’ variant, no distinction whatever is
marked by inclusive legal positivism; the two views come
to the same thing. The second distinction, however, that
between legal positivism qua naturalism and legal positivism
without naturalism, marks a difference that is constant.
That is, a characterization of a given legal system by appeal
to legal positivism qua naturalism is always different from
a characterization of the same system by appeal to legal
positivism without naturalism.
Kelsen, our proponent of legal positivism without
naturalism, wages battle on two fronts, against natural law
theory and against naturalism. And he responds on both
fronts with doctrines that count as independent doctrines in
his legal philosophy – ‘independent’ in that neither doctrine
is derived from the other. He responds to natural law theory
with the separation principle and to naturalism with the
nomological normativity thesis. The import of these two
doctrines is to be sharply distinguished from legal positivism
qua naturalism, where the separation principle is simply
a corollary of naturalism and where there is of course no
nomological normativity thesis.

The Very Idea of Legal Positivism: Stanley L. Paulson

Uploaded by

Copyright:

Available Formats

The Very Idea of Legal Positivism: Stanley L. Paulson

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Very Idea of Legal Positivism: Stanley L. Paulson

Uploaded by

Copyright:

Available Formats

The very idea of legal positivism

denial of the morality principle comes to can be refined, we

stood in a great variety of ways, as illustrated by prominent examples in

there is necessary overlap between the law and morality. It

Specifically, in Part One of the paper, two theses are of

nineteenth century.11 By contrast, ‘positivism’ is a term of

in Königsberg 1855), in Helmholtz, Vorträge und Reden, 4th printing, 2 vols.

of naturalism, and David Hume’s view counts as another. I

Relegating to a footnote what we, influenced by

And, lest the significance of the appeal to habit be

law. And if this reduction of sovereignty to fact is indeed

remainder to matters physiological or psychological. And

Back to Austin, as promised. I have spoken of tracing

Fear, a brute fact, is the operative notion here, and the

‘that which is not feared is not apprehended as an evil’); Austin, Province

2. The substitution of naturalism for positivism

These notions – feeling, sentiment, instinct, the

a hundred years ago and in an extraordinary treatise on

3. Positivism without naturalism. The case of

Normativity,39 Kelsen tells us, is his alternative to other

Raz, whose statement is in some respects the most extreme

figures are asking or answering our questions’, we ‘run

45 Paul W. Franks, All or Nothing. Systematicity, Transcendental Arguments, and

thesis of justified normativity, however, Kelsen’s normativity

46 As Kelsen writes: ‘It would be a serious misunderstanding if one wanted

peripheral imputation, is peculiar to Kelsen. Both doctrines

tion doctrines – central and peripheral – is found in a lengthy essay: Kelsen,

and ‘legal consequence’, more precisely, in terms of the

50 Kelsen, LT (n. 49), § 11 (b) (p. 23) (emphasis added).

science of his day into something scientifically respectable.52

52 See generally Kelsen, Über Grenzen zwischen juristischer und soziologischer

Its counterintuitive character aside, a point to which

This is close to being right as a statement of Kelsen’s

55 Hans Kelsen, ‘“Foreword” to the Second Printing of Main Problems in the

Still, formulation II seems counterintuitive in imputing

the empowerment of a legal organ to follow through with

59 See Kelsen, LT (n. 49), § 11 (b) (at p. 23).

(Gesetzlichkeit) as an example of a methodological form in

The methodological forms to which Rickert alludes are

the ‘law of normativity’.64 As he explains, looking back on

The normative or non-naturalistic import of Kelsen’s

64 See the quotation immediately below.

change in that person’s legal position. The change, Kelsen

You might also like