Skip to navigation – Site map
Letter to the Editors

For a truly realistic theory of law

Mauro Barberis
Translated by Paolo Sandro
p. 7–14


A truly realistic theory of law – the possibility of which is examined in this work – should have the following characteristics. It should be more general than current theories, that is, be applicable to both common law and civil law; in consequence, it should invert the relationship, commonly instituted by positivist theories, between legislation and adjudication. Both in historiography and in legal comparison, of which legal theory is an extension to a higher level of abstraction, adjudication is the central moment of the legal process, while legislation can be seen as one of the many attempts, and perhaps not even the most successful, to limit and control adjudication. An example of the explanatory capacity of a truly realistic theory is the so-called crisis of sources: legislation institutes judges, but they are, ultimately, those who set the hierarchy of sources and the hierarchical position of legislation itself.

Top of page

Full text

A realistic doctrine of the sources builds on experience,
but recognises that not all law is positive,
in the sense of “formally established”.
(Ross 1958: 101)

  • 1 See Barberis 2015 and Barberis 2016. I thank Andrej for this observation, for which I take full res (...)

1Andrej Kristan pointed out to me that, in some of my recent work, a reversal of the commonly established relationship between legislation and adjudication emerges.1

  • 2 A similar famous reversal is found in Gray 1909: 78-79: “judges’ decisions are the law, legislation (...)

2 Indeed, mainstream theory – a mix of positivism and legal realism – tends to reduce law to legislation, and adjudication to an interpretation of law. Instead, in my recent work, this relationship is reversed: the main function of law is adjudication, while legislation becomes mainly a way to control and limit adjudication.2

3 Andrej asked me to outline briefly the contours of this inversion, which I shall label, self-ironically, a truly realistic theory of law. However, readers should expect neither a sketched nor a fully-fledged theory; in what follows, I shall strive only to show whether it is possibile.

1 General jurisprudence

4First, one has to ask what general jurisprudence (allgemeine Rechtslehre) is, or rather: what is it that legal theorists really do? They essentially do two things which are not mutually exclusive; in any single theory, both are present, and their extent is primarily determined by the university environment in which theoreticians work.

5 Those who cultivate jurisprudence as the general part of private or criminal or constitutional law tend to conceive it as a continuation of legal dogmatics on a higher abstraction level, that is, as an analysis of the language used by legal scholars to study and interpret law. This first type of theory has obvious advantages, particularly if compared with a philosophy of law of a more speculative kind. But it tends to overestimate the role of legal dogmatics, central in continental legal culture, and to adopt its cognitive horizon. For example, it adopts ethnocentrism: the tendency to consider the concepts handled by continental legal dogmatics to be universal.

6 On the other hand, those who cultivate general jurisprudence as a subject in its own right, however, tend to conceive it as a historical and comparative study of both law and those phenomena that have performed the same functions in different ages and cultures. To the tools of logic and language, indispensable in any research, this second type of theory adds those proper to the social sciences, such as history, sociology, political science, economics, including their respective epistemologies. Compared to the first type of approach, this latter is more general, because it aims to give an account of more phenomena, and less abstract, because it distinguishes more carefully between their historical and cultural specificities.

  • 3 See Raz (1979: 105): ʻthe existence of norm-creating institutions, though characteristic of moder (...)

7 A truly realistic theory of law is primarily a theory of the latter type. It is more general precisely because it focuses upon adjudication: the function of settlement of disputes which is more general than legislation.3 But it is less abstract because it distinguishes between different shapes and aspects of adjudication, of which statutory interpretation is only one. As we shall see in the next section, this looks like a further extension, in a realistic and evolutionary sense, of legal positivism.

2 Positivism, realism, evolutionism

  • 4 Cf. paradigmatically Leiter 2007. A tradition of research connects theories and doctrines linked by (...)

8The current dominant jurisprudential approach, as has been mentioned, is already an extension, in a realistic sense, of the tradition of theoretical studies called legal positivism.4 A truly realistic theory is a further extension of this dominant approach in a realist and even evolutionary sense. But in what sense positivist, realist and evolutionist?

9 By legal positivism I mean the theory of positive, social, and actual law, as such distinct from critical morality as mainly practiced by the millennial natural law tradition. For legal positivists – who started working after the continental codification of law – positive law and positive morality, but also custom and religion, are distinct phenomena and yet connected in a variety of ways depending on the period and the culture. In ancient times, and in non-Western cultures, they often had no names to distinguish between them; in the West and after the codification, however, distinguishing between them becomes essential.

10 Legal realism is a form of positivism which favours adjudication over legislation. Today, there is no theory of law that does not recognise the judicial production of law, especially in the form of interpretation, often regarded as judicial legislation. A radically realistic theory of law considers adjudication – the resolution of disputes, bringing justice to the parties – to be a function which is more general and essential than legislation, which becomes mainly a way to control the former.

11 Legal evolutionism is, finally, a further extension of the positivist and realist traditions which criticises creationism, that is, the tendency to attribute phenomena to the will and planning of one or more creators. Biological evolution may, after all, also depend on human acts. Think, for example, about the crossbreeds hybridised by farmers, or about the projects of genetic engineers. However, no one would say that breeders or genetic engineers create life, while it has been said that the world was created by God, or law by the legislator or the judges themselves.

  • 5 The idea, although not the term, dates back to the System of Logic (1843) by John Stuart Mill. For (...)

12 Of course, compared to biology, law depends more on acts of human production. But you cannot reduce law to the mere sum, or to the individual effects, of each of them. Neither legislators nor judges can claim to create law, but only to take part in its production, participating in what North American theorists call the legal process. Law itself is not a mere set of rules, but their system or order. Like all systems, law also has emergent properties (see §5), properties additional to those of its individual parts, and determined by the evolution of the relations between them.5

3 The doctrine of sources

13While theorists of the second type identified in §1 deal with the social sources of law, albeit in a sense broader than that theorised by Joseph Raz, the first type deal with the formal sources, those studied by legal dogmatics, especially of the continental kind. After the French Revolution, in particular, this type of theory embraced the narrative of the origins of law, and the doctrine of sources typical of continental legal dogmatics.

14 According to this doctrine, every legal order has a political origin: it is born in a creative act of a constituent power, attributed to the people, who establish the legislative power and, through it, all the other constituted powers. All law is, then, produced by formal sources of a legislative kind – the Constitution, statutes, bylaws – already hierarchically ordered according to the power that produces it. Other sources of law – custom, dogmatics, case-law in civil law systems – are not formal but “cultural”, that is, useful only for the purpose of supplementing, commenting, or applying the formal sources.

15 This doctrine also aims at generality: even common law is, in fact, depicted in this way by English legal positivists. The Norman conquerors of England, as some sort of constituent power avant la lettre, instead of legislating directly, would set up royal courts, which have since produced common law. This, in turn, was conceived by English positivists, from Thomas Hobbes onwards, as some sort of judicial legislation, although it, in fact, took place before parliamentary legislation.

16 A truly realistic theory of law tells a different story and draws different consequences from it. There have always been, in the West, dispute settlement bodies, called to do justice between parties, called referees, judges or courts. The English monarchs, much like the continental ones, only set up new ones, which have prevailed over the other courts in a process of institutional selection. They were, in fact, more efficient, supported by the central government and, in having assigned a decisive role to the jury, even mimicked, in the English case, local justice.

17 Within this other narrative, legislation has prevailed over other sources of law because it was applied by royal courts, because it was codified and finally because it was legitimised by the democratic ideology of the constituent power. An authentic (cognitive) theory of sources, however, does not have to necessarily reproduce the continental (normative) doctrine. Such a theory is obliged to explain the pervasive role played so far by legislation even in common law systems, but also to give an account of the other legal sources actually used by courts. The best approximation to this theory of sources is found in Alf Ross.

4 The theory of sources

18The theory of sources built on the constituent power requires that any rule, to be called juridical, must be produced by a body authorised to do so by a superior rule. This requirement can obviously not hold for the same constituent power without triggering an infinite regress: as such the constituent power must be conceived of as an original fact, or as a revolutionary act, or as an extra ordinem source, or the like. But such a requirement does not by itself explain the judicial application of additional sources: custom, doctrine, case law, implicit norms, foreign norms not expressly incorporated, etc.

  • 6 See Ross 1958: 100-101. Immediately before this, one can read: ʻThe term “positivism” is ambiguous. (...)

19 As a matter of fact, the theory sketched by Ross in chapter three of On Law and Justice (1952/58) lists four main sources of law, of which at least three are not produced by normative authorities authorised by a superior rule and are as such different from statutory law. As Ross himself puts it, ʻ[a] realistic doctrine of the sources builds on experience, but recognises that not all law is positive, in the sense of “formally established”.6

  • 7 See Ross 1958: 76-77: ʻMetaphorically speaking, we can perhaps say that legislation delivers a fini (...)

20 Here I report Ross’s list in reverse order – i.e., from reason or tradition to legislation, through custom and precedent – emphasising the realistic traits of the list, while pointing out its creationist residues. The main realistic feature is the characterisation of sources as simple materials used by judges in adjudication. The creationist residue, however, consists of metaphors applied to such materials by Ross: reason and tradition are treated as raw materials, custom and precedent as semi-finished products, and legislation as the finished product.7

21 The first type of source is reason, or rather tradition: a “free” source or simple raw material that the judge can mould as she wishes. It is questionable whether reason or tradition can be labelled as sources as such, given that they were often merely required to prove the legality of other sources, such as custom. It is nonetheless true that, for many centuries, courts or juries, and especially English ones, have produced “free” law of this kind considered to be reasonable, traditional and non-arbitrary because it has been produced by a few royal judges united by a shared culture.

22 The second type of source is legal custom, distinguished with difficulty from non-legal custom. The main distinguishing criteria used are substantive, including compliance with reason and tradition, or formal, such as application by courts. Ross combines the two types of criteria considering customs that emerged in the areas regulated by law, in particular judge-made, to be legal.

23 The third type of source, which originated from the first two, is judicial precedent, which stood out as such when the House of Lords declared itself bound to its own precedents, much like the way in which continental courts were subordinate to codes. Today, even in common law countries, there may be no matter which is not regulated by statutes, but this does not diminish the role of precedent as a source of law. Even the application of law, in fact, produces precedents that are binding for future judges.

24 The fourth and final type of source is legislation itself, a form of law so paradigmatic by now that it almost makes any other form inconceivable, and still so widespread that it must be, in turn, divided at least into constitutional, parliamentary and administrative (or regulatory) legislation. For a truly realistic theory of sources of law, however, it is essential to make a correction, level criticism at and update Ross’s 1958 text.

25 The correction stresses that even legislation is never a finished product. As shown by Ross in chapter four on interpretation of On Law and Justice, legislation is a product which is somewhat more finished than custom and precedent, but still subject to judicial interpretation. Also, the application of legislation is not necessarily always more predictable than the application of precedent. In fact, its predictability depends on many factors besides legal drafting: the organisation of the judiciary, the political context, the degree of social pluralism, etc.

26 The criticism notes that if the primacy of legislation over adjudication depends on the democratic “dignity” of the first, then it runs the risk of growing dim. After all, already in the nineteenth century, the main type of legislation was not the parliamentary statute, but the code, i.e., an “aristocratic” source, since it is produced by legal scholars. Today, moreover, legislation through parliamentary initiative is a minority compared to the legislation enacted on government´s initiative, producing a type of law which is more technocratic and autocratic than democratic. Not surprisingly then, checks over this type of law are guaranteed not as much by democratic parliaments (which are nowadays often controlled by the executive power itself), but rather by supreme or constitutional courts, which are legitimised only by the enforcement of rights.

27 The update, finally, notes that parliamentary legislation is no longer the supreme source of law. In nation states, legislation is subject to the constitution, which is considered to be true law and not just positive moral, as John Austin thinks, precisely because it too is applied by constitutional and ordinary judges. In many countries, and especially in the European Union, domestic legislation is also subject to the constraints of international law. And the legality of this latter also depends on formal criteria less than it does on its effective enforcement by courts, both domestic and international.

5 The “crisis” of sources

28The reversal of the roles of legislation and adjudication that characterises our truly realistic theory is documented by one last phenomenon: the so-called “crisis” (of the doctrine) of sources. Produced by the French Revolution and legitimised by the democratic ideology of the constituent power, this doctrine establishes which sources are to be applied by courts and which not. After a failure of experiments, such as the référé législatif, it was impossible to prevent judges from interpreting or supplementing these sources. But it was decided that they should apply recognised sources, from which case law itself was excluded.

  • 8 This is the gist of the thesis defended in Pino 2011.

29 The mainstream theory of law is still based on this compromise: judges may interpret only sources established by the constituent power or by the legislator. The crisis (of the doctrine) of sources, however, forces us to reconsider this compromise. Great constitutional and international supreme courts, in fact, also bestow on themselves the power to choose their sources. In particular, they promote to the rank of supreme sources normative materials that were not previously considered to be formally legal, and redefine the hierarchical relationships between them and formal sources.8

30 This has happened many times since Marbury vs. Madison (1804), and in cases far more doubtful than the US Constitution. European treaties have been interpreted as foundational documents of the European Community; the preambles of some constitutions have been assimilated into the French bloc de constitutionnalité; fundamental laws were used as Israel’s constitution, etc. The effects of these decisions – an example of the emergent properties of legal systems (cf. §2) – are sometimes still labelled as the crisis of sources, although, overall, constitutional dogmatics has by now taken note of the phenomenon.

  • 9 See Troper 2005.

31 However, the most radical legal realism itself, represented by such continental law theorists as Michel Troper and Riccardo Guastini, manifests a curious ambivalence towards the crisis (of the doctrine) of sources. On the one hand, they were amongst the first to have highlighted the phenomenon and often draw the very same conclusions that a truly realistic theory could draw, as we are about to see. On the other hand, they argue that the aforementioned decisions are often little or not motivated at all, and when they are, Troper accuses them of committing a constitutionalist fallacy.9

  • 10 See Guastini 2008: 124-125.

32 There is, in this resistance, a residue of continental formalism; perhaps they both confuse the continental doctrine of sources with an authentic general theory of sources because they are legal theorists of the first type and as such in the wake of Hans Kelsen. Yet Guastini draws from the so-called crisis of sources – in fact, from the mere evolution of their relationships – the same conclusion that a truly realistic theory would draw. Evidently, supreme sources and their hierarchy are fixed in the final instance by constitutional adjudication.10

33 A few more words, in closing, about the fear that a truly realistic theory of law, in spite of its cognitive character, might obliquely legitimate the so-called government of courts.

34 Firstly, “government of courts” is a contradiction in terms. Adjudication is structurally distinct from government, because adjudication can be exercised only negatively, on someone else’s initiative. Accordingly, adjudication can, at best, control someone else’s government, and not exercise it in person.

35 But then, and most importantly, the system of precedents by courts works a bit like Willard Quine’s theory of knowledge. On the fringe, or if you want on the high levels of constitutional and international law, there are sometimes spectacular changes caused possibly by a single “big decision”. But in the great body of judicial practice, or on the lower floors of civil, criminal and administrative law, changes of a given rule require more decisions to set a precedent and are much slower, gradual and interstitial.

Top of page


Mauro BARBERIS, 2015: Contro il creazionismo giuridico. Il precedente giudiziale fra storia e teoria. Quaderni fiorentini per la storia del pensiero giuridico moderno (2015) 44: 67-101.

Mauro BARBERIS, 2016: Pragmatics of Adjudication. In the Footsteps of Alf Ross, in A. Capone and F. Poggi (eds.), Pragmatics and the Law. Vol. 2. Berlin: Springer, 2016.

Maurizio FERRERA, 2016: Rotta di collisione: euro contro welfare? Roma-Bari: Laterza.

John GRAY, 1997: The Nature and the Sources of the Law (1909/21). Aldershot: Ashgate/Darthmouth.

Riccardo GUASTINI, 2008: Teoria e dogmatica delle fonti. Milano: Giuffrè.

Ricardo GUASTINI, 2015: A Realistic View on Law and Legal Cognition. Revus. Journal for Constitutional Theory and Philosophy of Law (2015) 27: 45-54. URL : DOI: 10.4000/revus.3304.

Brian LEITER, 2007: Naturalizing Jurisprudence. Essays on American Legal Realism and Naturalism in Legal Philosophy. Oxford: Oxford University Press.

Giorgio PINO, 2011: Costruzione, decostruzione, ricostruzione. Ars Interpretandi, 16: 19-56.

Joseph RAZ, 1979: The Authority of Law. Oxford: Clarendon.

Alf ROSS, 1958: On Law and Justice. London: Stevens & Sons.

Michel TROPER, 2005: Marshall, Kelsen, Barak and the Constitutionalist Fallacy. I•Con. International Journal of Constitutional Law 3 (2005) 1: 24-38.

Top of page


1 See Barberis 2015 and Barberis 2016. I thank Andrej for this observation, for which I take full responsibility, even more than for the invitation to contribute to Revus.

2 A similar famous reversal is found in Gray 1909: 78-79: “judges’ decisions are the law, legislation is only the main legal source”. Cf. also Guastini 2015: 45: “the law is a set of norms in force [...] applied [...] by law-applying agencies”.

3 See Raz (1979: 105): ʻthe existence of norm-creating institutions, though characteristic of modern legal systems, is not a necessary feature of all legal systemsʼ, while ʻthe existence of certain types of norm-applying institutions is.ʼ

4 Cf. paradigmatically Leiter 2007. A tradition of research connects theories and doctrines linked by historical rather than conceptual associations.

5 The idea, although not the term, dates back to the System of Logic (1843) by John Stuart Mill. For an application of the concept to European law, cf. Ferrera 2016: 94.

6 See Ross 1958: 100-101. Immediately before this, one can read: ʻThe term “positivism” is ambiguous. It can mean both “which builds on experience” and “what is formally established”.ʼ

7 See Ross 1958: 76-77: ʻMetaphorically speaking, we can perhaps say that legislation delivers a finished product, immediately ready for use, while precedent and custom deliver only semi-finished product, which have to be finished by the judge himself, and “reasonproduces only certain raw materials from which the judge himself has to fashion the rule he needs.ʼ

8 This is the gist of the thesis defended in Pino 2011.

9 See Troper 2005.

10 See Guastini 2008: 124-125.

Top of page


Bibliographical reference

Mauro Barberis, « For a truly realistic theory of law », Revus, 29 | 2016, 7–14.

Electronic reference

Mauro Barberis, « For a truly realistic theory of law », Revus [Online], 29 | 2016, Online since 15 October 2016, connection on 23 March 2017. URL : ; DOI : 10.4000/revus.3624

Top of page

About the author

Mauro Barberis

Professor of Legal Philosophy at the University of Trieste (Italy)

Address: Università degli studi di Trieste, Dipartimento di Scienze Giuridiche, del Linguaggio, dell'Interpretazione e della Traduzione, Piazzale Europa 1, 34127 Trieste, Italy.


By this author

Top of page


All rights reserved

Top of page