LEGAL POSITIVISM AS A REALIST THEORY OF LAW

Brian Leiter
bleiter@uchicago.edu
to appear in P. Mindus & T. Spaak (eds.), The Cambridge Companion to Legal Positivism
7/21/2019

By “legal positivism,” I mean the view about the nature of law that H.L.A. Hart

articulated most powerfully in 1961 (while learning from and modifying the work of Hans

Kelsen), and that Joseph Raz developed further in the 1970s and 1980s, according to which (1)

where a legal system exists, there exists a “rule of recognition” specifying the criteria in virtue of

which norms are valid law; and (2) a rule of recognition is nothing more than a complex psycho-

social artifact constituted by the practice of officials, in particular, the criteria of legal validity

they converge upon and which they treat as obligatory (in Hartian lingo: that they accept from

“an internal point of view”).1 That means laws and legal systems rest at bottom on the

conventional practices of officials.

By realist theories of law, I mean theories which: (1) describe without sentimental or

moralizing illusions what law is and how it actually operates in human societies (descriptive

adequacy takes priority over moralizing sermons); (2) recognize that law is never adequate to

explain how courts adjudicate all cases that come before them; and (3) account for law and

adjudication within the constraints of a naturalistic theory of the world, i.e., one that eschews

appeal to any entities or properties that do not find a place in successful empirical scientific

accounts of natural and social phenomena. Both the American and Scandinavian (self-

1
I take Hart’s view to be an improvement on Kelsen’s kind of positivism in several crucial respects, most
importantly, in naturalizing the notion of validity via appeal to certain kinds of social practices (Hart, in this regard,
has more in common with the Scandinavian Realist Alf Ross than he acknowledges, a point to which I return,
below). I do not view other self-proclaimed positivist views as serious contenders.

Electronic copy available at: https://ssrn.com/abstract=3304243

identified) “legal realists” were proponents of realist theories of law in this sense, albeit in very

different ways, a point to which I return. Hart was a critic of both American and Scandinavian

legal realisms, though in both cases he missed his mark. The irony is that Hart’s legal positivism

is also a realist theory of law (more in the American than Scandinavian sense), and once we sort

out the misunderstandings and confusions, it will be clear that legal positivists and realists form a

unified theoretical front against the moralizing and ideological obfuscators about law, from Lon

Fuller to Ronald Dworkin.2 It will also turn out that one of Raz’s additions to Hart’s theory,

namely, the idea that law necessarily claims authority in Raz’s “Service Conception” sense,

betrays the realist ambitions of Hart’s theory.

A. The Realism of Hart’s Legal Positivism

Hart’s theory of law as developed in The Concept of Law (1961) is a realist theory in four

important respects.

First, Hart recognizes that law in modern societies operates primarily outside the courts,

namely, as a way of guiding ordinary people (sometimes with the help of lawyers) as to what

they can and cannot do and, in particular, how they can acquire various legally created powers,

such as the power to get married (and accrue the associated legal and tax benefits) or the power

to dispose of their property after their death. Any general theory of law that focuses primarily on

courts is not realistic. It is not simply that courts decide only a miniscule fraction of all legal

disputes (massive agreement about what the law requires, after all, is the norm in functioning

legal systems: see Leiter 2009: 1226-1228), it is that the primary function of law in modern

societies is to help people organize their affairs to avoid disputes, either with other citizens or

with the state: in other words, to avoid adjudication. Law tells people what they need to do to

2
This is not true of every view that self-identifies as either legal positivism or legal realism, but it is true for
the core cases discussed in the text.

Electronic copy available at: https://ssrn.com/abstract=3304243

buy a house, to get divorced, to avoid tax penalties, to enter into business relationships, and it

also sets conditions under which their freedom can be restricted or their property seized.

Second, adjudication of some legal disputes inevitably involves judgments that are not

constrained by law but instead require the exercise of discretionary judgment, informed by non-

legal considerations: the law is sometimes rationally indeterminate (i.e., the class of legitimate

legal reasons underdetermines the decision the adjudicator must reach [Leiter 2007: 9-12]). Any

theory that presents adjudication as a matter of discovering pre-existing answers in a legal

heaven is not realistic. Rules determinate in their application in most instances can never be

without uncertainty, unless they adopt patently absurd approaches, such as, “When the rule is

unclear, the plaintiff wins.” Hart saw, correctly, that indeterminacy in law was inevitable—

though he was wrong in thinking the “open texture” of language was its primary cause (a point to

which I return)—and he also saw correctly that this was not a feature of law anyone should try to

remedy (Hart 1961: 128-130): we are far better off with officials who can produce authoritative

resolution of disputes that no one anticipated than we would be with a system that tried to settle

every matter well in advance.

Third, the social phenomenon of law is explicable in wholly naturalistic terms, without

recourse to objective moral values, or Kantian transcendental conditions on judgment: law and

legal systems exist entirely because of psychological and sociological facts about persons. On

one reading, the claim that law is part of the natural order is bound to seem trivial (theists and

others in the grips of religious illusions to one side): is not everything part of the “natural order”

of the world? But positivists and realists understand this claim in a more demanding way:

everything real is part of the causal account of nature developed by successful empirical

sciences, whether physics or biology or psychology. There are no facts outside the causal order

Electronic copy available at: https://ssrn.com/abstract=3304243

of nature as described by the various sciences. Legal systems have important normative

dimensions—officials and sometimes subjects take the system’s demands to be obligation-

imposing, for example—but this is a complicated psychological and sociological fact about

persons, not a fact about the world apart from those psycho-social facts about persons. Any

theory that posits transcendental conditions on legal judgment (e.g., Kelsen) or mind-

independent moral values (e.g., Dworkin) is not realistic.

Fourth, Hart recognizes that law is not necessarily a good thing, that governance by law has

costs, and as a result we should never presuppose that law is morally good or obligatory. This

has been the central idea of the positivist tradition in legal theory: we can say what the

normative system called “law” requires while recognizing that what the law demands may not be

what we should do. Talk of law not being “morally binding” must of course be given a non-

cognitivist gloss, consistent with Hart’s actual way of marking normative practices.3 Any

theory, like Dworkin’s, that builds the justification of coercive power into the nature of law is not

realistic.

B. Legal Realisms, American and Scandinavian

Hart contributed to the perception that legal positivism was not a realist theory of law by

his famous, but ultimately misguided, criticisms of the American and Scandinavian legal realists.

The fault is not entirely his, to be sure: the American legal realists were rhetorically exuberant,

rather than analytically precise, and an unfortunate translation of Alf Ross’s Of Law and Justice

probably exacerbated the misunderstandings of the Scandinavian position.4

3
Recall that for Hart the ascription of the “internal point of view” to agents is based entirely on behavioral
criteria, i.e., what actors do and say. This no doubt betrays the influence of Gilbert Ryle’s behaviorism in
philosophy of mind.
4
See the discussion in Leiter & Etchemendy (2017: section 5.1), drawing on the work of Jakob
Holtermann.

Electronic copy available at: https://ssrn.com/abstract=3304243

 Hart also solidified the unfortunate assimilation of the two realisms, American and

Scandinavian, when, in reality, they have almost nothing in common: calling them both

instances of “realism” is mainly an accident of intellectual history and a confusion about words.

While the Scandinavians adopted the label “realism” to signal their opposition to metaphysical

idealists, who thought the nature of reality was dependent on the human mind and its categories, 5

the Americans invoked a different sense of “realism”—partly colloquial and partly

corresponding to an intellectual tradition stretching from Thucydides to Machiavelli and beyond-

-as being candid about what really happens (especially when courts decide cases), without

sentimental or moralistic illusions (cf. Leiter 2012). The latter is the sense of realism I am

mainly concerned with here.

The difference should hardly be surprising given the very different professional and

intellectual backgrounds of the two groups. The major American Realists—Karl Llewellyn,

Jerome Frank, Herman Oliphant, Max Radin, Underhill Moore, William O. Douglas, among

others—were law professors, lawyers and jurists, with no philosophical training and, rather

obviously, no aptitude for or interest in philosophical questions.6 What they were all interested

in was how courts actually decide cases, and what lawyers and law students needed to know if

they were to reliably advise their clients about how they would fare in the courts. They advanced

no theses about the nature of law apart from these practical litigation-oriented concerns, though it

is striking that their tacit assumptions about the nature of law were all positivist ones, a point to

which we will return (Leiter 2001).

5
Hägerstrom is a case in point: see Mindus (2009: 48). Mindus also notes other writers skeptical about
the lumping together of Scandinavian and American Realism, id. at 137 n.2.
6
Felix Cohen was the exception—although his philosophy PhD served him badly. See, e.g., (Cohen 1935),
which arguably adopts an indefensible version of rule-skepticism. For discussion, see Leiter (2007: 68-73).

Electronic copy available at: https://ssrn.com/abstract=3304243

 The Scandinavians were different: they were either primarily philosophers or had some

formal training in philosophy, and all were primarily academics and scholars who conceived of

their jurisprudential questions in philosophical terms. In contemporary terminology (Jackson

1998), they were occupied with variations on what philosophers call the “location problem,” that

is, the problem of how to locate some phenomenon—e.g., consciousness or moral value—in a

world naturalistically conceived, that is, conceived in the causal terms described by the empirical

sciences. So, for example, philosophers interested in the location problem might ask how we

make sense of subjective conscious experience, or of the idea that something is morally wrong,

in a world that is exhaustively explained by physics, chemistry and biology? Or, how can we

locate the experience of Beethoven’s 9th Symphony as beautiful, or the sense that genocide is a

moral abomination, in a world thought to be composed of nothing other than physical, chemical,

biological and/or psychological facts?

For the Scandinavian Realists, the precise issue was how to locate a legal system of

norms in a world so conceived. The Scandinavians were philosophically diverse on some issues,

so generalizing about them will no doubt run roughshod over some individual differences. Yet

there are two basic issues at stake: metaphysical (do these norms really exist?) and semantic

(what is the meaning of legal statements mentioning such norms?). On the metaphysical

question, there are elements of eliminativism in the views of Olivecrona (1971) and Ross (1959).

Just as the eliminativist about religious discourse thinks that the entities that religious discourse

refers to (gods, souls, spirits) do not exist, and so should be eliminated from any realistic picture

of what the world is like, the legal eliminativist thinks that legal obligations, legal rights, and

legal duties do not really exist either, since such concepts play no role in a scientific explanation

of the world. Because the Scandinavians sometimes seem to be eliminativists at the

Electronic copy available at: https://ssrn.com/abstract=3304243

metaphysical level, they (generally) opt for “non-cognitivist” views about the semantics, that is,

views according to which statements about legal obligations, rights and duties do not really try to

refer to anything, but are simply the expression of the “attitudes” of legal actors.7 So, for

example, judgments about legal obligations and duties are interpreted as expressing certain non-

cognitive attitudes or feelings: a court’s judgment of the form, “Mr. Smith has a contractual

obligation to pay Mr. Jones $5,000 for those widgets” really means something like, “I, the judge,

feel the force of the legal demand that Mr. Smith pay under these circumstances, and if he

doesn’t, I will sanction him!” The psychologism of some of the Scandinavians held that legal

norms, both their existence and their effects on behavior, were to be explained in terms of the

psychological states of persons who accepted these norms, rather than the reality of the norms:

that X believes or feels he has a duty or obligation is what is explanatory of behavior, not the

existence of the duty (cf. Mindus 2009: 142 n. 31). A similar kind of psychologism is, I think,

implicit in Hart, though he sometimes cannot quite acknowledge it, perhaps because of the

influence of J.L. Austin’s ordinary language philosophy on his thinking. This requires further

explication.

Hart was a metaphysical anti-realist about moral norms: such norms do not really exist,

since the best explanation of the world would not include facts about what is morally right and

wrong. This might suggest that the normativity of morals—the ability of what is morally right

and wrong to give people efficacious reasons to act—should be understood in exclusively

psychological terms, i.e., as a psychological fact about what particular people believe and feel

when they are told that, “It is morally wrong to do X” or “It is morally right to Y.” The most

charitable reading of Hart should understand him as accepting something like this view, since,

7
See the more detailed discussion in Leiter & Etchemendy (2017: section 5.1). See also Spaak
(2011: 163-165; 2009: 493-494). As Patricia Mindus reminds me, Hägerstrom is an unusual case,
opting for an error theory about “rights,” but a kind of emotivism about judgments of moral value.

Electronic copy available at: https://ssrn.com/abstract=3304243

unlike Hans Kelsen, his was an “impure” theory of law, in which anti-realism about norms was

conjoined with non-cognitivism about the semantics of normative judgment: to judge that doing

X is morally (or legally) wrong is just to express a certain kind of attitude or feeling, presumably

one tied—psychologically—to motivation and action (cf. Toh 2005). Hart did, of course, reject

the view he attributed to Alf Ross, namely, that legal validity consists in “a verifiable hypothesis

about future judicial behavior and its special motivational feeling” (Hart 1959: 165). (Hart

failed to realize that this was a stipulative account in Ross, not an ordinary language account:

Ross wanted to show how we could “locate” doctrinal or “external” statements about what the

law is in a world naturalistically conceived, a point to which I return.) For Hart, to say that a

legal rule is valid is just to say that it satisfies the criteria of legal validity in the Rule of

Recognition. But neither I nor Ross is denying Hart’s conceptual point about judgments by legal

actors about legal validity, namely, that its meaning is not reducible to predictions about

behavior: even Ross’s treatment of such judgments was non-cognitive, like Hart’s (Ross 1959:

6-9). Rather, the point at issue here concerns what an anti-realism about norms, which Hart

accepts, entails about motivation and action. On the realist/naturalist view, to the extent an

agent’s judgment that X is morally right (or legally obligatory) has normative force for the agent,

that normative force is explicable in terms of certain psychological facts about the agent. Hart

sometimes seems to deny this, e.g., when he says that “the internal character of [normative legal

judgments, like, ‘You must stop at the red light’] is not a mere matter of the speaker having

certain ‘feelings of compulsion’; for though these may indeed often accompany the making of

such statements they are neither necessary nor sufficient conditions of their normative use in

criticizing conduct, making claims and justifying hostile reactions by reference to the accepted

standard” (Hart 1961: 167) But Hart is best understood here as simply revealing his

Electronic copy available at: https://ssrn.com/abstract=3304243

commitment to a kind of Rylean behaviorism about attitude ascriptions: agents manifest the

‘internal point of view’ simply in virtue of using the right kind of normative language and

making the right kinds of justificatory or critical comments about, respectively, conforming or

deviant behavior vis-à-vis some rule. Contemporary naturalists have no reason to accept the

behaviorism, given the explanatory fruitfulness of psychological explanations in so many

domains.

In sum, the Scandinavians have deeper affinities with the great Austrian legal positivist

Hans Kelsen and the great English legal positivist H.L.A. Hart than they do with the American

Realists: the latter want to figure out why courts decide as they do, and then let lawyers in on the

real causes accordingly; while all the former—that is, Kelsen, Hart, and the Scandinavians--are

interested in traditional philosophical questions about the metaphysical nature of law and (in the

case of Hart and the Scandinavians) how it can be located within a naturalistic worldview, that is,

one that takes for granted that the world is, more or less, as the sciences describe it. Hart and the

Scandinavians differ from Kelsen in their willingness to think the answer can involve

“naturalizing” certain normative concepts in law, that is, understanding them as complicated

psychological and sociological artifacts, and thus as part of what Kelsen would have understood

to be an “impure” theory of law.8

Hart’s Mistaken Critique of Legal Realism

Hart attributes to legal realism a view he calls “rule-skepticism,” a view he deems correct

in certain instances (consistent with his own realism about adjudication). Hart considers two

different kinds of rule-skepticism. One, which I will call “Conceptual Rule-Skepticism,” holds

that “talk of rules is a myth, cloaking the truth that law consists simply of the decisions of courts

and the predictions of them” (Hart 1961: 133). The second is the only important kind for his
8
I’ve explored that possibility as well: see the essays in Part II of Leiter (2007).

Electronic copy available at: https://ssrn.com/abstract=3304243

putative dispute with legal realism, which I will call “Empirical Rule-Skepticism.” According to

this view, “Rule-scepticism has a serious claim on our attention, but only as a theory of the

function of rules in judicial decision” (Hart 1961: 135); more precisely, this realist/skeptic claims

“that it is false, if not senseless, to regard judges as themselves subject to rules or ‘bound’ to

decide cases as they do” (Hart 1961: 135).

Conceptual Rule-Skepticism grows out of Hart’s conflation of the American and

Scandinavian Realists. Conceptual Rule-Skepticism is a claim about the nature of law, i.e., that

it consists only in what courts say or in predictions of what courts will do. Hart’s critique of this

as an account of the nature of law as reflected in ordinary language is wholly persuasive, in my

view (cf. Leiter 2007: 69-73): such a theory of the ordinary concept of law cannot account for

how judges think about what the law “obligates” them to do, or about the way in which lawyers

and citizens criticize judicial conduct by reference to the law. But did either the Scandinavians

or the Americans endorse Conceptual Rule-Skepticism?

Alf Ross, as we have seen, did proffer something like this, but not because he thought it

was adequate to the “ordinary” usage of the concept of law, or because he was insensitive to the

points Hart raised. He proffered such an analysis as part of a philosophical reconstruction of

statements about what the law is by scholars and observers, one that would make sense of its

normative concepts in a world naturalistically conceived: by reinterpreting normative claims

about “duties” and “obligations” as empirical predictive claims, Ross aimed to show that law

was not polluted with metaphysical nonsense. Ross was not doing ordinary language

philosophy about how things looked to legal insiders; Hart was. Hart criticized Ross as offering

an implausible analysis of the “ordinary” concept; Ross would have agreed. Hart offered no

direct response to the actual “location problem” Ross was concerned with.

10

Electronic copy available at: https://ssrn.com/abstract=3304243

 In the case of the American Realists, Hart was also guilty of misreading, although in

fairness to him, at least one Realist, Felix Cohen, was, arguably, a Conceptual Rule-Skeptic,

although Cohen is nowhere cited by Hart: Hart’s American Legal Realism relies on an

amalgamation mainly of Frank, Holmes, and Llewellyn. Plainly these writers do talk about the

importance of “predicting” what courts will do. The question is whether, in so talking, they are

fairly read as offering an analysis of the concept of law. Only Hart’s grossly anachronistic

reading suggests an affirmative answer.

The idea that philosophy involves "conceptual analysis” via the analysis of language (or

appeals to intuitions about how concepts or terms apply) is an artifact of Anglo-American

analytic philosophy of the twentieth-century; indeed, as practiced by Hart, it really reflects the

influence of fashionable views in philosophy of language current at Oxford in the 1940s and

1950s. The American Realists were not philosophers, let alone analytic philosophers, let alone

students of G.E. Moore, Russell, and Wittgenstein, let alone colleagues of J.L. Austin. The idea

that what demands understanding about law is the “concept” of law as manifest in ordinary

language and ordinary intuitions would have struck them as bizarre. While the Realists had

much to say about adjudication and how legal rules work in practice, they had nothing explicit to

say about the “ordinary” concept of law or our intuitions about it.9

So what did the American Realists mean in talking about ‘prediction’? In Law and the

Modern Mind, Jerome Frank cautions the reader early on that he “is primarily concerned with

‘law’ as it affects the work of the practicing lawyer and the needs of the clients who retain him”

9
Raz and his followers often obscure this point. Raz, for example, claims that the argument for his “hard”
or “exclusive” positivism “is not an argument from the ordinary sense of ‘law’ or any other term. It relies on
fundamental features of our understanding of a certain social institution, the primary examples of which are
contemporary municipal legal systems....” (1979: 50). But all his examples are examples of ordinary or familiar
distinctions people draw between law and other normative systems, between legal and other kinds of normative
competence. The giveaway is the appeal to “our understanding of a certain social institution”: what could that
possibly be other than an armchair report of what Raz takes people to say and believe about law

11

Electronic copy available at: https://ssrn.com/abstract=3304243

(Frank 1930: 47 note). Holmes begins “The Path of the Law” by emphasizing that he is talking

about the meaning of law to lawyers who will “appear before judges, or...advise people in such a

way as to keep them out of court” (1897: 457). Against this background, infamous statements

like Llewellyn’s--”What these officials do about disputes is, to my mind, the law itself” (1930:

3) --make perfect sense. Theirs is not a claim about the “concept” of law, but rather a claim

about a practical heuristic, how it is helpful to think about law for attorneys who must advise

clients what to do. For your client the franchisee in Connecticut doesn’t simply want to know

what the rule on the books in Connecticut says; he wants to know what will happen when he

takes the franchisor to court. So from the practical perspective of the franchisee, what one wants

to know about the “law” is what, in fact, the courts will do when confronted with the franchisee’s

grievance. That is all the law that matters to the client, all the law that matters to the lawyer

advising that client. And that is all the American Realists wanted to emphasize.

Ironically, it turns out the American Realists could not have been Conceptual Rule-

Skeptics, because their arguments for the indeterminacy of law actually presuppose the

positivist’s concept of law. The central claim of legal indeterminacy is the claim of rational

indeterminacy: the claim that the “class of legal reasons” fails to justify a unique outcome in

some or all cases. The “class of legal reasons” is the class of reasons that are taken to properly

justify a legal conclusion (and thus “compel” it insofar as legal actors are responsive to valid

legal reasons). So, for example, appeal to a statutory provision or a valid precedent are parts of

the class of legal reasons in the United States, while appeal to the authority of Plato’s Republic is

not: a judge is not obliged to decide one way rather than another because Plato says so. Any

argument for indeterminacy, then, presupposes some view about the boundaries of the class of

legal reasons. When Herman Oliphant (1928) argues, for example, that the promise-not-to-

12

Electronic copy available at: https://ssrn.com/abstract=3304243

compete cases are decided not by reference to law, but by reference to uncodified norms

prevalent in the commercial culture in which the disputes arose, this only shows that the law is

indeterminate on the assumption that the normative reasons the courts are actually relying upon

are not themselves legal reasons.10 So, too, when Holmes chalks up judicial decisions not to

legal reasoning but to “a concealed, half-conscious battle on the [background] question of

legislative policy” (1897: 467) he is plainly presupposing that these policy concerns are not

themselves legal reasons. The famous American Realist arguments for indeterminacy which

focus on the conflicting, but equally legitimate ways, lawyers have of interpreting statutes and

precedents11 only show that the law is indeterminate on the assumption either that statutes and

precedents largely exhaust the authoritative sources of law or that any additional authoritative

norms not derived from these sources conflict. It is the former assumption that seems to

motivate the Realist arguments. Thus, Llewellyn says that judges take rules “in the main from

authoritative sources (which in the case of law are largely statutes and the decisions of the

courts)” (1931: 13). In short, the American Realists presuppose something like the positivist

idea of a rule of recognition whose criteria of legality are exclusively ones of pedigree: a rule (or

canon of construction) is part of the law in virtue of having a source in a legislative enactment or

a prior court decision.12

Matters are more interesting with respect to what I am calling Empirical Rule-

Skepticism. Here is Hart’s characterization of the view in question:

10
According to Oliphant, the courts found ways to enforce promises by sellers of business not to compete
with purchachers, because this made good economic sense: there would be no point to the transaction otherwise.
By contrast, courts found ways to invalidate promises not to compete by employees by employers, since such
arrangements were disfavored by trade guilds and took unfair advantage of the bargaining position of employees.
11
The classic discussions are Llewellyn (1930: 72-76); Radin (1930); Llewellyn (1950: 395-406).
12
One difference is that most positivists recognize customary practices as sources of valid law, whereas it is
not clear what the Realists think about this.

13

Electronic copy available at: https://ssrn.com/abstract=3304243

 [I]t amounts to the contention that, so far as the courts are concerned, there is nothing to

circumscribe the open texture [of language in which rules are framed]: so that it is

false, if not senseless, to regard judges as themselves subject to rules or “bound” to

decide cases as they do. They may act with sufficient predictable regularity and

uniformity to enable others, over long periods, to live by courts’ decisions as rules.

Judges may even experience feelings of compulsion when they decide as they do, and

these feelings may be predictable too; but beyond this there is nothing which can be

characterized as a rule which they observe. (Hart 1961: 135)

Empirical Rule-Skepticism on this rendering, then, involves two claims: (1) legal rules are

indeterminate; and, as a result, (2) legal rules do not determine or constrain decisions.

Obviously (1) is not an empirical claim, but a philosophical or jurisprudential one. But

(2) could be true even if (1) were false (that would be pure Empirical Rule-Skepticism, we might

say). Yet Hart is surely correct that most American Realists endorse both (1) and (2).

Unfortunately, Hart is mistaken about the Realist argument for (1), and thus underestimates the

amount of indeterminacy in law. And the real dispute, if there is one, between the American

Realists and Hart turns in the end on (2).

Remember that Hart acknowledges that legal rules are indeterminate, but claims that is a

phenomenon at “the margin of rules” (1961: 135); the realist skeptic, on Hart’s rendering, is

simply “a disappointed absolutist” (1961: 135), someone who set the bar for legal determinacy

too high. But the American Realists, with some exceptions (notably Jerome Frank), recognized

that law was mostly rationally determinate; they pointed out that it tended to be indeterminate,

however, in those cases that reached the stage of appellate review. And they argued for that

proposition based on the empirical evidence of lines of appellate cases that could not be

14

Electronic copy available at: https://ssrn.com/abstract=3304243

rationalized in terms of the official doctrine. Their argument for indeterminacy was different

from Hart’s, which may help explain the apparent disagreement.

According to Hart, legal rules are indeterminate because “there is a limit, inherent in the

nature of language, to the guidance which general language can provide” (1961: 123).

Language is, in Hart’s famous phrase, “open-textured”:

There will indeed be plain cases constantly recurring in similar contexts to which

general expressions are clearly applicable (“If anything is a vehicle a motor-car is one”)

but there will also be cases where it is not clear whether they apply or not. (“Does

‘vehicle’ used here include bicycles, airplanes, roller skates?”). The latter are fact-

situations, continually thrown up by nature or human invention, which possess only

some of the features of the plain cases but others which they lack. (1961: 123)

Language is “open-textured” in that while words have “core” instances--aspects of the world that

clearly fall within the extension of the word’s meaning--they also have “penumbras,” cases

where it is unclear whether the extension includes the aspect of the world at issue. In cases in

which the facts fall within the penumbra of the key words in the applicable legal rule, a court

“must exercise a discretion, [since] there is no possibility of treating the question raised...as if

there were one uniquely correct answer to be found, as distinct from an answer which is a

reasonable compromise between many conflicting interests” (1961: 128).

While some writers have challenged the implicit semantics that informs this argument for

the indeterminacy of language (a fortiori of law),13 I think we should agree with Hart that legal

rules must necessarily be indeterminate in some measure given the indeterminacy of language

itself, and that this type of indeterminacy resides “at the margin of rules” (1961: 132). What

13
For criticism of such approaches, see Leiter & Etchemendy (2017: section 5.3).

15

Electronic copy available at: https://ssrn.com/abstract=3304243

bears noting here is that the American Realist arguments for indeterminacy did not rely on these

claims about language.

The American Realists located the indeterminacy of law not in general features of

language itself, but in the existence of equally legitimate, but conflicting, canons of interpretation

that courts could employ to extract differing rules from the same statutory text or the same

precedent.14 Indeterminacy resides not in the rules themselves, but in the ways we have of

characterizing what rules statutes and precedents contain. Llewellyn, for example, argues that

any precedent can be read “strictly” or “loosely,” and that either reading is “recognized,

legitimate, honorable” (1930: 74). The strict interpretation characterizes the rule of the case as

specific to the facts of the case; the loose interpretation abstracts (in varying degrees) from the

specific facts in order to treat the case as standing for some general norm. But if “each precedent

has not one value [that is, stands for not just one rule], but two, and...the two are wide apart,

and...whichever value a later court assigns to it, such assignment will be respectable, traditionally

sound, dogmatically correct” (1930: 76), then precedent, as a source of law, cannot provide

reasons for a unique outcome, because more than one rule can be extracted from the same

precedent.

Llewellyn thinks the same is true of statutory interpretation, where one can find

“‘correct,’ unchallengeable rules of ‘how to read’ which lead in happily variant directions”

(1950: 399). By mining the cases, Llewellyn shows that courts have endorsed contradictory

canons of construction, like “A statute cannot go beyond its text,” but also “To effect its purpose

a statute must be implemented beyond its text” (1930: 90). But if a statute can properly be

construed in contradictory ways to stand for different rules, then reasoning from the statute will

not justify a unique outcome in the case at hand.

14
I take it something similar is true of the Italian Legal Realists of the Genoa School.

16

Electronic copy available at: https://ssrn.com/abstract=3304243

 One difficulty with these Realist arguments is that they rely on a tacit conception of

legitimate legal argument. The assumption is that if lawyers and courts employ some form of

argument--a “strict” construal of precedent, a particular canon of statutory construction--then that

form of argument is legitimate in any and all cases. Put in this strong form, the assumption

cannot be right: not every strict construal of precedent will be legally proper in every case.

Even Llewellyn must recognize this, as suggested by his famous--but clearly facetious--example

of the strict reading that yields, “This rule holds only of redheaded Walpoles in pale magenta

Buick cars” (1930: 72). But that would never be a legitimate construal of a precedent (barring

some bizarre scenario in which all these facts turned out to be legally relevant), and Llewellyn

surely knows as much. The claim cannot be, then, that any strict or loose construal of precedent

is always valid. It must only be that lawyers and judges have this interpretive latitude often

enough to inject a considerable degree of indeterminacy in to law. 15

Notice that the Realist argument for Empirical Rule-Skepticism is actually conceptually

severable from the claim that the law is indeterminate. For centrally, it is based on the

observation that the decisions courts reach do not fall in to patterns that correlate with the rules

they invoke; rather the decisions reflect judges’ response to the underlying facts of the cases.

This was the point of Oliphant’s example of the promise-not-to-compete cases mentioned earlier;

it is Llewellyn’s point too in his discussion of the New York sales cases.16 What causes judges

15
In an unpublished paper, Leslie Green has objected that there appears no principled explanation why
the indeterminacy should not go “all the way down,” as it were. But the key is that the argument depends on a
“sociological” notion of which interpretations of sources are “legitimate,” meaning accepted by most lawyers and
judges. Lawyers and judges may, of course, treat interpretive moves as acceptable even without a principled basis
for doing so.
16
Llewellyn (1960: 122-124). According to Llewellyn, a series of New York cases applied the rule that a
buyer who rejects the seller’s shipment by formally stating his objections thereby waives all other objections.
Llewellyn observed that the rule seemed to have been harshly applied in some cases where the buyers simply may
not have known at the time of rejection of other defects or where the seller could not have cured in any case. Upon
looking at the underlying facts of the cases, however, Llewellyn discovered that in each case where the rule seems

17

Electronic copy available at: https://ssrn.com/abstract=3304243

to decide as they do, according to the Realist, is not legal rules,17 but a sense of what would be

fair on the facts of the case at hand. All American Realists agreed on this point. Judge

Hutcheson claimed that “the vital, motivating impulse for the decision is an intuitive sense of

what is right or wrong for that cause” (1929: 285). Frank invoked the example of “a great

American judge,” Chancellor Kent, who had confessed that, “He first made himself ‘master of

the facts.’ Then (he wrote) ‘I saw where justice lay, and the moral sense decided the court half

the time; I then sat down to search the authorities....but I almost always found principles suited to

my view of the case....;” (1930: 104 note). Llewellyn noted that to understand what appellate

courts are really doing, one must appreciate “how far the proposition which seems so abstract has

roots in what seems to be the due thing on the facts [before the court]” (1930: 29). As a result,

Llewellyn’s advice for the lawyer was “on the facts...[to] persuade the court your case is sound,”

even if the lawyer must also provide a “technical ladder” to the “sound” result (1930: 76).

Frank pointed out that the very same advice had been given by a former President of the

American Bar Association (1930: 102 note). In short, the core of the Realist defense of

Empirical Rule-Skepticism is, in fact, empirical: they looked at what the courts (primarily the

appellate courts) really did, and found that legal rules were clearly not the determining factors in

a large number of cases.

to have been harsly applied what had really happened was that the market had gone sour, and the buyer was
looking to escape the contract. The court, being “sensitive to commerce or to decency” ((1960: 124) applied the
unrelated rule about rejection to frustrate the buyer’s attempted violation of a basic norm of commercial conduct:
namely, parties should honor their commitments even under changed market conditions. It is, in short, the
“background facts, those of mercantile practice, those of the situation-type” (1960: 126) that determine the court’s
decision.
17
Some legal rules, of course, structure the presentation of facts to the court, and thus “the facts of the
case at hand” are unintelligible without legal rules. Conceding this point does not change the fact, however, that
there still may be a disjunction (as the Realists claim) between the substantive rules of law that judges invoke in
their decisions and the real bases for their decisions.

18

Electronic copy available at: https://ssrn.com/abstract=3304243

 At the same time, the Realists--unlike many of the later Critical Legal Studies writers--

did not overstate the irrelevance of rules. For one thing, the American Realists were (generally)

clear that their focus was indeterminacy at the stage of appellate review, where one ought to

expect a higher degree of uncertainty in the law. Cases that have determinate legal answers are,

after all, less likely to be litigated to the stage of appellate review. Most of the time, the Realists

seemed to understand this point. Thus, Llewellyn explicitly qualified his indeterminacy claim by

saying that, “[I]n any case doubtful enough to make litigation respectable the available

authoritative premises...are at least two, and...the two are mutually contradictory as applied to the

case at hand.” (1931: 1239). And Max Radin noted that judicial “decisions will consequently be

called for chiefly in what may be called marginal cases, in which prognosis is difficult and

uncertain. It is this fact that makes the entire body of legal judgments seem less stable than it

really is” (1942: 1271). Empirical Rule-Skepticism is surely more plausible when it is not

advanced as a global claim about adjudication and the law.

Notice now that even if we agreed with Hart that the open-texture of language affects

rules only “at the margins” (whatever that means, a point to which we return), the Realists have

now given us an additional reason (beyond Hart’s) to expect indeterminacy in law. If the

Realists are right, then not only do legal rules suffer from the open-texture that Hart describes,

but statutes and precedents will frequently admit of “manipulation”—legally legitimate

manipulation, of course--and thus be indeterminate in this additional respect as well. The

combination of sources of interdeterminacy (the open-texture of language, and the conflicting

canons of interpretation) seems sufficient to move indeterminacy from the margins to the center

of cases actually litigated through the stage of appellate review.

19

Electronic copy available at: https://ssrn.com/abstract=3304243

 Hart is not wholly insensitive to the Realist arguments, though he treats them cursorily.

In response to Llewellyn’s point, for example, that a court can interpret a precedent both

“loosely” and “strictly” and thus extract two different rules from the same prior decision, Hart

says simply this: “in the vast majority of decided cases there is very little doubt [as to the rule of

the case]. The head-note is usually correct enough” (1961: 131). Alas, a first-year litigation

associate knows that this approach to precedent would be a recipe for disaster. To extract

“holdings” without regard to the facts of the case--which is all a head-note typically provides--is

mediocre lawyering. Skillful lawyers know exactly what Llewellyn describes: that the “rule” of

a prior case can be stated at differing degrees of specificity, and so made to do very different

rhetorical work depending on the needs of the case at hand.

So both Hart and the American Realists acknowledge indeterminacy in law, and both

acknowledge, accordingly, that rules do not determine decisions in some range of cases. Hart

seems to think they disagree over the range of cases about which these claims hold true, but that

is not entirely clear. Consider Hart’s ultimate rejoinder to Realist skepticism about the influence

of legal rules on court decisions:

[I]t is surely evident that for the most part decisions...are reached either by genuine effort

to conform to rules consciously taken as guiding standards of decision or, if intuitively

reached, are justified by rules which the judge was antecedently disposed to observe and

whose relevance to the case in hand would generally be acknowledged. (1961: 137)

There is, alas, no argument here: “it is surely evident” is simply a denial of what the Realist

affirms, yet offers no evidence for the denial other than Hart’s armchair confidence in the

correctness of his own view.

20

Electronic copy available at: https://ssrn.com/abstract=3304243

 But perhaps Hart and the American Realists are not really disagreeing? That may seem

surprising, but I now think it the most plausible interpretation, given the clear Realist elements in

Hart’s own views, and the clear positivist elements in the views of the American Realists. After

all, when Hart limits indeterminacy to the “margin of rules,” this is against the backdrop of one

of his central insights, namely, that law exists mainly outside the courts as a way for individuals

to plan their affairs and guide their conduct. The American Realists sometimes make the

“American mistake” (think of Dworkin) of focusing far too much on courts, but if we put that

parochial prejudice aside, they have no theoretical reason to deny Hart’s insight. And that

insight is compatible with recognition that cases that reach the stage of appellate review are

especially likely to be ones where the law is indeterminate, precisely for “selection effect”

reasons: parties have more incentive to pursue further litigation when the law is unclear and the

result could go either way. Hart has no theoretical reason to deny this latter point.

Remember also that the American Realists explicitly did not discount the role of rules in

adjudication. Llewellyn is representative when he asks, “Do I suggest that...the ‘accepted rules,’

the rules the judges say that they apply, are without influence upon their actual behavior?” and

answers, “I do not” (1930b: 444). The Realist approach, says Llewellyn, “admits...some relation

between any accepted rule and judicial behavior” but then demands that what that relation is

requires empirical investigation, since it is not always the relation suggested by the “logic” (or

content) of the rule (1930b: 444). As he puts the point elsewhere: Realists deny that

“traditional...rule-formulations are the heavily operative factor in producing court decisions”

(1931: 1237). But to deny only this claim is to admit that rules play some causal role in

decisions.

21

Electronic copy available at: https://ssrn.com/abstract=3304243

 We thus arrive at a possible rapprochement between two schools of Anglophone realist

jurisprudence, that of Hart’s positivism and the Legal Realists. Both Hart and the American

Realists agree that valid legal rules (i.e., those with the relevant pedigree) matter, that they are

often determinate in their application, and that at the higher levels of adjudication, indeterminacy

obtains and judges must exercise discretion. In addition, the Scandinavian Realists can agree

that Hart has a satisfactory account of the “ordinary” concept of law, while disagreeing whether

the best attempt to solve the location problem for law in a world conceived naturalistically

requires the theoretical artifice of recasting normative talk in law in predictive terms or whether a

non-cognitivist approach, that arguably Hart favors, will suffice for statements about the law by

both legal actors and observers. And all the realists and the positivists agree that law does not

always suffice to explain what happens in adjudication and all reject the objectivity of moral

values. Even if there can be reasonable doubts about whether the muddle-headed American

Realists understood all these points, or whether the Scandinavian Realists, in the grips

(generally) of logical positivism, did, we can affirm that sensible realist jurisprudence can and

must affirm them.

Realism, Positivism and Raz’s Authority Argument

In the recent positivist tradition, the most influential contribution to the theory comes

from Hart’s student Raz, who argued, against Hart, that no rule of recognition at the foundation

of a legal system could incorporate moral, or content-based, criteria of legal validity (Raz 1985).

Although a rule of recognition is constituted by the practice of officials—those criteria of legal

validity that they converge upon and treat as obligatory—it cannot, on Raz’s view, rely on

criteria of legal validity that require judges to assess the content of the norms to determine their

legal validity. According to Raz, law necessarily claims authority, even if it ultimately lacks a

22

Electronic copy available at: https://ssrn.com/abstract=3304243

justification for that claim, and only source-based criteria of legal validity are compatible with

the possibility of law claiming authority. A legal system can only claim authority if it is possible

to identify its directives without reference to the underlying (“dependent”) reasons for that

directive. This is a “prerequisite” for claiming authority because what distinguishes a (practical)

authority in the first place is that its directives preempt consideration of the underlying reasons

(including, e.g., moral reasons) for what we ought to do, and in so doing actually makes it more

likely that we will do what we really ought to do. But Hart’s “soft positivism,” as well as

Dworkin’s view, makes the identification of law depend on the very reasons that authoritative

directives are supposed to preempt, and thus makes it impossible for law to fulfill its function of

providing authoritative guidance. Thus, to the extent judges purport to treat norms as legally

blinding based on their appealing moral or other normative content, they are simply exercising

discretion under the mask of doing what the law requires.

Soft positivists, defenders of Hart’s own official view in his “Postscript,” have lodged

various objections to Raz’s argument, but from a realist point of view, the real worry about Raz’s

argument is that it presupposes a highly moralized and unrealistic conception of what it is for law

to claim “authority.” Raz understands authority in terms of what he calls “the Service

Conception,” in which a claim to authority is a claim to provide a subject of authority better

“reasons” for acting in accordance with what Raz calls “right reason” than the subject would

arrive at without the intermediation of the authority (Raz 1985). (We can gloss, for purposes

here, “right reason” as what one really ought to do.) We can bracket for a moment the

extravagant metaphysics of reasons such a view requires—though it is hard to see how to make

sense of the objective existence of such reasons from a naturalistic point of view (Leiter 2015)--

and make a more modest objection: why think that when law claims authority it is claiming to

23

Electronic copy available at: https://ssrn.com/abstract=3304243

help its subjects do what they really ought to do? One might look, even cursorily, at the history

of the world to think this is implausible: of course law claims authority in the sense of claiming

the right to tell those subject to its jurisdiction what they must do, but the idea that this claim to

authority is a claim to Raz’s Service Conception looks like moralizing nonsense.

In defense of Raz’s view, one might note that legal systems do claim that what they

demand of their subjects is what they really ought to do. No legal system, after all, adopts as an

explicit principle: “Do this, even though it is wicked and unreasonable.” Most legal systems, in

fact, adopt the posture, “Do this,” without much further qualification at all, but let us concede the

point: it is still not sufficient to motivate Raz’s Service Conception even on its own terms.

Legal systems may often claim that they are requiring people to do what is morally right, or

God’s will, or in the public interest, and so on; but they do not claim that their subjects are more

likely to do what is morally right or God’s will or in the public interest by following the law’s

demands. Legal systems claim authority to direct their subjects’ behavior, but they do not do so

on the basis that they are thereby performing a “service” for their subjects.

This response exposes Raz’s repudiation of realism in legal theory: for, against Hart’s

realist aim of a “descriptive jurisprudence,” Raz asserts “there is an interdependence between

conceptual and normative argument” (1986: 63) such that the account “does not necessarily

conform to everyone’s notion of authority in every detail. It does claim to be an explanatory

account in singling out important features of people’s conception of authority” (1986: 65). Of

course, no theorist thinks their descriptive account conforms “to everyone’s notion…in every

detail,” so this is just a distraction. But where is the evidence that Raz’s account even captures

some “important features of people’s [which people?] conception of authority”? Raz, I fear, has

24

Electronic copy available at: https://ssrn.com/abstract=3304243

simply smuggled in a moralized conception of authority18—as providing a service for its

subjects—without even the pretense that it satisfies even lax evidential demands of a “descriptive

sociology” as Hart, the realist, proposed.

Does that mean Hart was correct in the “Postscript” that rules of recognition can

incorporate content-based criteria of legal validity?19 Not quite, though not for the reasons I

have given previously. Hart himself, recall, agrees that a satisfactory account of law must

explain how the law fulfills an epistemic function, i.e., enabling at least officials to recognize

what the law is in many if not all cases. (Hart himself in “The Postscript” thought soft

positivism was compatible with such an epistemic function.) But I now reject the argument I

defended in Leiter (2007: 126-127), about what I called the “Extreme Scenario.” In this

scenario, it would be part of the conventional practice of officials to treat “Natural Law” as the

obligatory criterion of legal validity. Of course, the whole history of the world shows the

requirements of the putative “natural law” are a subject of massive controversy (because there is

no such thing), which might appear to mean that if a rule of recognition treated “Natural Law” as

the sole criterion of legal validity, it could not discharge any epistemic function. The problem

with this kind of objection—apart from its trading on conceptual possibilities not nomic

18
Moralized in the sense that its main aim is to explain why deference to authority by rational subjects can
be justified.
19
Shapiro (1998) offers a slightly different argument against soft positivism, one that relies not on a claim
about authority but on a claim about what it is to be “guided” by a rule (as he notes [1998: 487) “Hart...has
always claimed that the essence of law is the guidance of conduct.”) According to Shapiro, one is “motivationally”
guided by a legal rule when one’s “conformity is motivated by the fact that the rule regulates the conduct in
question” (1998: 490). Shapiro argues that on Hart’s view, the rule of recognition must provide motivational
guidance to officials (1998: 492) but that a rule of recognition incorporating content-based criteria cannot make a
“practical difference” to the judges: “The rule makes a practical difference to one’s practical reasoning [reasoning
about what one ought to do] only if, in this counterfactual circumstance, the agent might not conform to the rule”
(1998: 495). But, argues Shapiro, a rule of recognition incorporating moral criteria could provide no motivational
guidance, since judges are already motivated to do what morality requires. It should be obvious why this
argument fails, even if its conceptual claim about “providing guidance” is correct: first, it is unrealistic, indeed
preposterous, to think that judges are already motivated to do what morality requires; and second, even if
individual judges were motivated to act morally, there is no evidence that their understanding of morality
coincides with what the rule of recognition demands of them.

25

Electronic copy available at: https://ssrn.com/abstract=3304243

actualities--is that under concrete historical circumstances, agents often do converge on parts of

what they take to be “natural law,” and thus appeals to natural law can serve an epistemic

function. Hart’s own view of meaning--in terms of what competent speakers of a language

would take to be extensions of particular terms--should have led him to realize as much: “due

process” and “equal protection” may have bigger penumbras than “vehicle,” but they still also

have a core set of applications which speakers of the same language—and thus adherents to the

same moral ideology manifest in that language—will agree upon. And that means that content-

based tests of legal validity are not all the same: some are really ideological covers for exercises

of discretion (as hard positivists have always correctly emphasized), but others may simply

involve exactly the same kind of interpretive approach as with respect to any rule with the right

pedigree, i.e., appeal to a competent speaker’s understanding of the term (except, of course,

where pedigreed rules of interpretation dictate otherwise). We may still, from a positivist/realist

point of view, reject the idea of genuine content-based criteria of legal validity but that would be

on the grounds of the explanatory fruitfulness of that way of demarcating the legal from the non-

legal (Leiter 2007: __-__) and not from claims about the concept of authority. Law claims

authority, but not in Raz’s sense; judges sometimes treat moral considerations as criteria of legal

validity, but in the same way they treat the ordinary understanding of “vehicle” as a criterion of

applied legal validity in a particular case; and the best explanation of law does not need to

countenance as “law” those cases where judges treat moral considerations as criteria of legal

validity where that judgment depends entirely on substantive moral judgment and argument.20

References

Cohen, Felix. 1935. “Transcendental Nonsense and the Functional Approach,” Columbia Law

20
Thanks to participants in the conference in Stockholm in November 2018 for useful discussion and
comments—especially Patricia Mindus, Torben Spaak, and Kevin Toh. Thanks also to Michael Sevel for helpful
discussion of Raz’s views.

26

Electronic copy available at: https://ssrn.com/abstract=3304243

 Review 35: 809-___.

Frank, Jerome. 1930. Law and the Modern Mind.

Hart, H.L.A. 1959. “Scandinavian Realism,” reprinted in his Essays in Jurisprudence and

Philosophy (Oxford: Clarendon Press, 1983).

-----. 1961. The Concept of Law (Oxford: Clarendon Press). Citations are to the 3rd

edition [add details]

Jackson, Frank. 1998. From Metaphysics to Ethics

Leiter, Brian. 2001. “Legal Realism and Legal Positivism Reconsidered,” Ethics . Reprinted

in Leiter (2007)

-----. 2007. Naturalizing Jurisprudence (Oxford: Oxford University Press).

-----. 2009. “Explaining Theoretical Disagreement,” University of Chicago Law Review

-----. 2011. “The Demarcation Problem in Jurisprudence: A New Case for Skepticism,”

Oxford Journal of Legal Studies

-----. 2012. “In Praise of Realism (and Against ‘Nonsense’ Jurisprudence),” Georgetown Law

Journal

----. 2015. “Normativity for Naturalists,” Nous Supplement

----and Matthew Etchemendy. 2017. “Naturalism in Legal Philosophy,” Stanford Encyclopedia

of Philosophy

Llewellyn, Karl. 1930. The Bramble Bush

----. 1930b. “A Realistic Jurisprudence—The Next Step,” Columbia Law Review 30: 431-465.

-----. 1931. “Some Realism about Realism—Responding to Dean Pound,”

-----. 1950. “Remarks on the Theory of Statutory Interpretation,” Vanderbilt Law Review

-----. 1960. The Common Law Tradition (Boston: Little, Brown).

27

Electronic copy available at: https://ssrn.com/abstract=3304243

Marmor, Andrei. ___. Farewell to conceptual analysis…

Mindus, Patricia. 2009. A Real Mind: The Life and Work of Axel Hägerström ( ).

Oliphant, Herman. 1928. “A Return to Stare Decisis,”

Olivecrona, Karl. 1971. Law as Fact, 2nd edition (London: Stevens & Sons).

Radin, Max. 1930. “Statutory Interpretation,” Harvard Law Review 43: 863-885.

-----. 1942.

Raz, Joseph. 1985. “Authority, Law, and Morality,” The Monist

-----. 1986. The Morality of Freedom (Oxford: Oxford University Press).

Ross, Alf. 1959. On Law and Justice (Berkeley: University of California Press).

Shapiro, Scott. 1998. “On Hart’s Way Out,” Legal Theory

Spaak, Torben. 2009. “Karl Olivecrona on Judicial Lawmaking,” Ratio Juris 22: 483-__.

-----. “Karl Olivecrona’s Legal Philosophy: A Critical Appraisal,” Ratio Juris 24: 156-__.

28

Electronic copy available at: https://ssrn.com/abstract=3304243