Legal Positivism: Critical Assessment and Epistemological Reflexions

From IVR encyclopedie

Jump to: navigation, search

by Sebastián Urbina

To make a critical assessment of legal positivism requires stating, more or less clearly, what legal positivism is—but this is not an easy task. Legal Positivism is like a dragon with many heads. In this respect, it is important to remember that Norberto Bobbio (1) claimed that Legal Positivism could be understood: first, as an approach to the study of law; second, as a legal theory, (as a determined theory or conception of law) and third, as an ideology of justice.

On the other hand, it is usually stated that legal positivism can be seenfirst, as ‘legal formalism’; second, as ‘ideological positivism’, and third, as ‘methodological positivism’, also called ‘conceptual positivism’. ‘Legal formalism’ asserts that the legal order is a self-sufficient system able to give a univocal solution to any legal conflict; according to ‘ideological positivism’, we—citizens and officials—ought to obey the law no matter its content; while ‘methodological positivism’ maintains that the concept of law ought not be defined according to evaluative properties but only on the basis of descriptive ones.

In any case, this critical assessment will be undertaken on the basis that legal positivism is an account of the concept of law that denies that there are necessary constraints—be it on the content of the law or in the interpretative activities of judges and, furthermore, that both the existence and the content of law can be fully explained by contingent social activities of persons who serve as officials in a given legal system.

Legal theories, generally speaking, can usually be characterized as either descriptive or normative. This distinction is useful but we must be cautious when we use such terms. On the one hand, legal descriptive theories try to describe the structure of legal orders. Let us take two examples to see what ‘descriptive legal theories’ may mean. First, according to Harris (2), Kelsen’s legal theory says that a legal system is to be equated with a collection of pure norms interpreted by legal scientists as a non-contradictory field of meaning.Second, Olivecrona’s legal theory says that a legal system is to be equated with the mass of rule-ideas, intermittently revived in the minds of the members of a political community, who owe their psychological effectiveness to the echo of certain ‘law words’.

If we claim that these two theories are descriptive, does it mean that Kelsen and Olivecrona observed the same legal object, saw different things and, consequently, described legal systems in different ways? If description tells us ‘what it is’ as something different from ‘what ought to be’, we would have to deal with the problem of different descriptions of the same legal object. How do we know that one description is the true one? We should remind ourselves here that description (in legal fields or not) requires selection of what is considered relevant. Otherwise, legal systems seem to be something more than factual regularities, social or psychological. We require, inter alia, justificatory reasons. Why? Because we (legal participants) want to understand the legal object better and not only undertake, for example, an exercise in descriptive sociology.

If this is correct, we have to recall certain classics. In a similar (but not equal) way we have to criticise some epistemological basis of positivism. Like Dilthey, Simmel, Collingwood, Windelband, Rickert, Weber, Gadamer, Habermas, von Wright and others, we oppose: first, ‘methodological monism’, that is, the idea that there is only one method to arrive at knowledge, no matter the features of the object; second, the method of natural sciences would be the true method that we should incorporate into legal methodology, and third, causal-explanation as the unique and true way to know the object.

We can say that any legal theory, any legal ontological commitment, is elaborated from a given view of reality that is not placed in the realm of necessity. In this respect, to observe (a legal object or any other thing) we require previous criteria for observation. Hence, any legal theory or any legal ontological commitment shows how a given jurist understands the world and the place that the legal reality occupies in it. Next step is whether legal theorists should assume the same ontological commitments as natural scientists. I believe not. In my opinion, the most acceptable legal theory would include a realist alternative in relation to legal facts (in the sense that they cannot go against the laws of nature) and an idealist alternative in relation to legal meaning contents, in so far as they cannot be touched, measured, etcetera.

Hence, the legal ‘is’ is a two faced expression: on the one hand, it includes what we (legal participants) have selected as the relevant physical entities, at least for our legal purposes; on the other hand, it includes what we have selected as the relevant normative behaviour. This ‘selection’ is not irrational because the legal community elaborates justifications for it. An alternative could be to believe that some privileged beings, using a privileged method, would be able to reach ‘the legal truth’ or ‘the legal essence’.

Let us consider the example of a ‘soft’ legal positivist, like Hart. In my view, it is not the case that he (or any legal positivist) could observe from the outside or state ‘detached statements’. What really happened, in my view, is that Hart developed certain epistemic virtues, like creativity, intellectual honesty, objectivity, tenacity, coherence, etcetera. But he cannot adopt a detached view, as a Taliban, for example, can. Hart was enmeshed in the Rule of Law and our form of life and he couldn’t observe from the outside. In sum, descriptions are always selective (not pure and neutral) and diverse legal theories, like Austin’s, Kelsen’s, Hart’s, Dworkin’s, etcetera, are not the result of different descriptions because of different observations. They are the result of different ideas about which are the most relevant features of the object to better represent it, granted certain premises and purposes.

What has been commented above means, that the distinction between descriptive and normative (at least in legal normative fields) is rather problematic in practice, perhaps not in theory. This would be so because, in these fields, there is an intertwinement of the descriptive and the normative. This is not so in physics, for example. In sum, there is an intertwinement between facts and values or descriptive language and normative language. It means, inter alia, that descriptive sentences may influence or modify evaluations but also evaluations (or normative language) may influence or modify descriptive language. It also means that we should talk of descriptive and normative language only within a given context because it, through time, can modify the meaning of words or sentences and their evaluations. In addition, terms that can be descriptive in one instance, for example ‘Jew’ (or others like ‘old people’) can be evaluative in another.

Granted that there is not a unique and true description of legal normative texts, the diverse descriptions may carry different levels of approbation or disapprobation. Finally, in so far as social and moral conventions change, what could be understood in a given sense, at time (t-1) could be understood (the same piece of language) at time (t-2), in a different sense. This makes the permanent task of legal adaptation so important. But not any legal adaptation goes, but the best it can be, here and now. This would be so because we are not neutral observers but committed participants, and because our social practices presuppose values.

According to the present conception of legal positivism there are the following criteria of legality: theSeparability Thesis, according to which there are no conceptually necessary moral criteria of legality; theSocial Fact Thesis, according to which the status of any norm as law is determined entirely by social facts or processes; and the Conventionality Thesis, according to which the criteria of legality, in any possible legal system, are determined by the conventional practices of those persons who serve as officials of the system in a legislative, executive, or judicial capacity.

The Separability Thesis

Another way to put this thesis is to claim that it is possible for a legal system to have criteria of legality that do not include any moral norm. Even if there is a legal system lacking criteria of legality that do not include any moral norm, it doesn’t show that there is not a relation between this legal system and certain moral criteria. This would be so because there is not any legal system lacking a certain conception of justice on the basis of this legal system. Hence, even if there are no criteria of legality—including, explicitly, moral norms—the underlying view of justice (or morality) of a given legal system will permeate the whole legal enterprise.

The relevant point is that any legal system, treating in a certain way different social relations, shows how they ought to be regulated. In this respect, no legal system tries to regulate human behaviour in an unjust way but in a just way. That is why there is always some underlying view of justice in the basis of any legal system. In this respect, it is a trivial truth that legal systems are not conceptually linked to morality, understood as a unique morality. Otherwise, some legal positivist thinkers have said that legality is one thing and legitimacy another. In this vein, I would say that one thing is formal validity and another legal content, but no legal system could dispense with them. Similar considerations apply to legality and legitimacy, as I have commented before.  In relation to distinctions, like formal validity–legal content, or legality–legitimacy, one thing is to make analytical distinctions for a better understanding of legal systems and another to believe that they are, really, radically separated and self-sufficient worlds.

Legal positivists tend to believe that in the real world there are impenetrable walls that clearly separate the different domains of social life. The Separability Thesis is an example of this new kind of metaphysics. But there are intertwinements and porous membranes that do not, at least always, allow for clear-cut distinctions. To deny it is important for legal positivists because they require ‘clear-cut observable entities’ to successfully engage in the supposedly sharp distinctions between descriptive language’ and normative language, facts and values, context of justification and context of discovery, and similar ones.

The Social Fact Thesis.

More concretely, it says that the existence and content of laws and legal institutions are constructed, or determined, by social facts about how persons, who serve as officials in the legal system, behave in the exercise of their official duties. There is a sound claim in this thesis. Laws and legal institutions do not come from heaven but are the result of the task of human beings. It is difficult to reject it. What seems problematic is the idea that we can understand these social facts through empirical descriptive claims that refer to beliefs—more concretely, the legislators’ and judges’ beliefs about what they must do in their legal task. Does this mean that the Social Fact Thesis is committed to a psychological or sociological enquiry?

Otherwise, one thing is that legal institutions are constructed, and another one is that they are determined by these social facts. If the latter, we should suppose that we are not free, at all, to shape our legal institutions. That is, given certain social facts we would be determined to have certain legal institutions. Another problem is which level of determination we are talking about. If we talk of concrete social facts (about how persons who serve as officials in the legal system behave in the exercise of their official duties) it is hard to understand how the officials’ legal decisions can be determined. By the way, what does concretely legal behaviour mean? Can we compare it with the behaviour of a dog when it comes into rut? The dog’s posterior behaviour is biologically determined. Do legal positivists claim that? If not, we are led to a constructive approach.

In constructive approaches we are not in the realm of ‘necessity’. In relation to legal knowledge we can use the two known metaphors: the pyramid, built on sure foundations, and the net that has no foundations but interlocking parts, emphasizing the value of coherence. I think that in legal fields we participate in both. On the one hand, our knowledge is foundational, not because we have secure foundations, but because we have must sources as foundations and we cannot dispense with them in legally valid decisions. Granted this, our legal knowledge is also coherentist because we have to join coherently several legal items, within given procedures, when taking legal decisions.

Otherwise, if the task of legal positivists is to observe the (let us say) legal behaviour of legislators and judges and scrutinize what they believe they must do legally, what will they understand? Are they interested in understanding or only in causal explanation? I believe that a legal theory should be neither sociology nor psychology. In this respect, to make empirically descriptive claims is not the proper way to understand the rule-governed behaviour of legal participants.

There are two basic versions of legal positivism: Exclusive legal positivism and Inclusive legal positivism. The first (represented by Raz and others) says that law is necessarily independent on morality. The second (represented by Hart and others) says that, in some cases, there is a connection between law and morality. This is so when the criteria of validity, of a given legal system, include moral standards. But there is only a contingent connection between law and morality.

In spite of the differences between these two versions, they have a common problem. That is, how we can arrive at normativity (obligatoriness) starting from social facts. In other words, they should justify how the confirmation of certain social facts is a sufficient justificatory basis to infer a duty to obey the law. If we accept that any legal system is conceptually linked to some morality (or theory of justice) we can solve this problem. But, then, we would be out of legal positivism because this doctrine claims that we find the obligatoriness of law in the law itself, without resort to some morality (or theory of justice). We should remind us here that legal positivism also claims that law is a content and independent normative system.

Otherwise, in which sense is judges’ behaviour observable? Is it the same as observing the rain? Is observation of physical behaviour observation of human action? If observation in casual terms is not enough to understand human action, what should we observe? Furthermore, legal theorists cannot predict the future legal behaviour of judges. Neither can a judge. Hence, we are not like physicists and our method should not be empirical but hermeneutic, as even Hart has accepted.

The Conventionality Thesis.

Most contemporary legal positivists follow H.L.A. Hart in believing that the content of the criteria of legality are fixed or determined by a conventional rule of recognition. On this view, the criteria that determine what counts as law, in any given society, are defined by the norms that officials collectively accept and obey in performing their official duties as legislators, executives and judges.

We should differentiate, again, two aspects. On the one hand, the claim that the content of the criteria of legality are fixed or determined by a conventional rule of recognition. On the other hand, the claim that the criteria that determine what counts as law, in any given society, are defined by the norms that officials collectively accept and obey in performing their official duties as legislators, executives and judges. First, even if the criteria of legality were determined by a rule of recognition, this ‘determination’ is relatively open.Second, what the criteria of legality determine is what counts as law, understood as legal normative texts (valid because duly enacted), not their legal meaning contents that require interpretation.

Let us take an example of a mother and her little daughter. The mother criterion would say: ‘when going on a trip, be good’. Granted that this hypothetical legal content is too vague to be operative, the teachers who go with the children will settle what can be concretely defined as good and bad behaviour, on a given trip. If this is correct, valid enacted law is insufficient to let us know what the content of the law is. In this respect, the rule of recognition can identify the valid legal rules but not their legal meaning contents because the law is not a static enterprise. We can see again the important and permanent task of adaptation undertaken by judges who, at large, share an internal point of view.

Legal positivists accept that the internal point of view does not require that officials believe that the rule of recognition is morally legitimate because they could accept this rule for prudential reasons. In this respect, it is common to distinguish between internal and external point of view.

The later would refer to those who either observe (let us say, from outside) the legal order at stake, or just obey for prudential reasons. Although there is not any logical necessity, officials, generally speaking, adopt and internal point of view and it means that the legal system or, at least, their most important norms and values are, on their view, morally correct. But to claim that a given legal system is morally correct means that there is an ultimate justification, and ultimate justifications are moral ones. These ultimate justifications are the basic values of any legal system, or some conception of justice that is the justificatory basis of any legal system.

Otherwise, the inevitable (pragmatically speaking) task of legal adaptation could not be properly done only on the basis of prudential reasons. To search for the best legal solution, here and now (specially in hard cases) requires, to do a good job, a compromise with, at least, the basic values and goals of the legal system at stake.

Final Comments

According to S. Toulmin there are three classes of approaches to philosophy. ‘For one group of philosophers, the thing that makes rational discussion and agreement possible, at all, is the Nature of the World; for a second group, it is the Structure of Human Thought; for a third group, it is the conditions- and interactions- that ensure and adaptation or harmony between Thought and the World … For the first group, the rational task depends ultimately on our capacity to recognize certain permanent and necessary (or ‘objective’) features and relations … One group sees philosophy as a technical discipline, whose formal quasi-mathematical method and concerns overlap those of geometry and logic’.(3)

In so far as legal positivist tendencies are based on formal validity and/or social efficacy we could say that the cited approach to philosophy, is close to legal positivism. Another feature, common to legal positivist tendencies would confirm this opinion. I refer to the common idea that what matters is not content (that would be contingent) but form (that would be permanent). It is the known legal positivist view that law may have any content. At the same time, legal realist tendencies (that emphasize social efficacy) would be placed in the same group, in so far as legal science would be a science of facts, social, psychological or both. Furthermore, some legal positivist tendencies rely on theoretical reason but not in practical reason, that would be a self-contradictory concept, according to outstanding legal positivists, like Kelsen or Ross.

Should legal theorists assume the same ontological commitments as natural scientists?. I believe not. In my view, legal theorists should accept both, realist and idealist ontological commitments. It means that the most acceptable legal theory would include a realist alternative in relation to facts and an idealist alternative in relation to legal meaning contents. Nevertheless, when we talk of facts we are not dealing with facts tout court. As M. Taruffo says: 'When we talk of ‘construction’, ‘definition’ or ‘identification’ … it seems clear that we are not making reference to the fact as an occurrence of the empirical reality, but to sentences, of different nature, that refer to occurrences that supposedly happen in the world of empirical reality … what is constructed or defined according to concepts, values or norms are sentences relative to facts of the real world … Accordingly, in the process ‘the fact’ is in reality what it is said of a fact: it is the enunciation of a fact, not the empirical object that is stated’.(4)

If we accept this view, we would have a language that refer to occurrences of the empirical world, and another language that refers to the meaning contents of legal positive rules. The first kind of language would refer to objects that pertain to the empirical world. These types of views would be closer to a realistic ontology. The second language would refer to human constructs, in a stronger sense than the first one. I don’t propose an idealist view (in ontology) that claims that the empirical world does not exist apart from the thinking subject. I propose this idealist ontology only for the meaning contents of legal positive rules. This idealist ontology does not support the idea of a Cartesian self-enclosed mind. Language, in general, and legal language, in particular, organizes and categorizes social reality but this task is not made by a solipsist mind but, in our case, by legal communities.

Finally, J. Mosterín says: ‘Since forty years ago everyone attacks the invisible and silent positivist windmills … But nobody confess to be a positivist and nobody defends positivism … There is not a unified scientific language. There is not a unique scientific method. There is not a unique and true description of the world … It is clear that science is not a set of verified statements about our sense impressions.’(5)

We could apply these considerations to Legal Theory. First, there is not a unified legal language. Nevertheless, some legal scholars claim to use the only right legal language. Second, there is not a unique legal method. Nevertheless, some legal scholars claim to use the only right legal method. Third, there is not a unique and true description of the legal system. Nevertheless, some legal scholars claim that they can do it, granted that they are using, both, the right legal language and the right legal method. Fourth, legal science is not a set of verified legal statements about our sense impressions. Very few legal scholars, if any, would support this super-realist view. Most believe that legal scholars build legal models from which to scrutinize the legal system. Nevertheless, some legal scholars claim that their constructed legal models are the only right ones because they would show ‘the essence’ of Law, or of legal concepts.

Even if the essence is understood, not as a ‘real essence’ but as a ‘nominal essence’ (for example, asking for the functions of sentences containing a given legal concept) it is not self-supporting. That is, legal concepts (or legal systems) are understandable when placed in larger theories and practices, not when they are isolated in their internal structure. Nevertheless, not all legal analytic thinkers see their task as separated from (and not overlapping with) other disciplines. Not all of them see legal philosophy as a second order field of study isolated from first order truths about the world.


1. Bobbio, N; Contribución a la Teoría del Derecho, edited by A. Ruiz Miguel, F. Torres , Valencia, 1980, page 31.

2. Harris,J.W; Law and Legal Science, Clarendon Press, Oxford, 1982, page 24.

3.Toulmin, S; Knowing and Acting, MacMillan Publishing Co., New York, 1976, pages 56,57.

4. Taruffo, M; La Prueba de los Hechos, edit. Trotta, Madrid, 2002, pages 113,114.

5. Mosterín J; Ciencia Viva, Espasa, Madrid, 2001, page 102.


Alexy, R; El Concepto y la Validez del Derecho, Gedisa, Barcelona, 1994.

Brix, B; ‘’On the Dividing Line Between Natural Law Theory and Legal Positivism’’, 75 Notre Dame Law Review1613 (2000).

Coleman J. and Leiter B; ‘Legal Positivism’, in A Companion to Philosophy of Law and Legal Theory, D. Patterson, editor, Oxford, Blackwell Publishing, 1996

Dworkin, R; Taking Rights Seriously, Cambridge MA, Harvard University Press, 1977.

Dworkin, R; Law’s Empire, Cambridge MA, Harvard University Press, 1986.

Finnis, J; Natural Law and Natural Rights, Oxford, Clarendon Press, 1980.

Fuller, L; The Morality of Law, rev.ed., New Haven: Yale University Press, 1964.

Garzón Valdés, E; ‘Algo más acerca de la relación entre Derecho y Moral’, en Derecho, Ética y Política, Centro de Estudios Constitucionales, Madrid, 1993, pages 317-335.

Hart, H.L.A; The Concept of Law, Revised Edition, Oxford, Oxford University Press, 1994.

Hart, H.L.A; Essays on Jurisprudence, Oxford, Clarendon Press, 1983.

Himma, K. E; ‘’Inclusive Legal Positivism’’, in Jules Coleman, Scott Shapiro (eds.) and Kennet Einar Himma (assoc. ed.), The Oxford Handbook of Jurisprudence and Philosophy of Law (Oxford: Oxford University Press, 2001).

Himma, K. E; ‘’Law’s Claim to Legitimate Authority’’, in Jules Coleman (ed.) Hart’s Postcript (Oxford: Oxford University Press, 2001).

Himma, K. E; ‘’Judicial Discretion and the Concept of Law’’, 19 Oxford Journal of Legal Studies 71 (1999).

Kelsen, H; Teoría Pura del Derecho, UNAM, Mexico, 1983.

Lyons, D; Ethics and the Rule of Law, Cambridge, Cambridge University Press, 1984.

MacCormick, N; ‘Natural Law and the separation of Law and Morals, in Natural Law TheoryContemporary Essays, Oxford, Oxford University Press, 1992.

Nino, C; Derecho, Moral y Política, Ariel, Barcelona, 1994.

Peczenik A. and Urbina S; ‘Why Officials?: On Legal Positivism, Old and New’, Rechtstheorie 26. Band, 1995, Heft 2. pages 139-162.

Raz, J; The Authority of Law, Oxford, Oxford University Press, 1979.

Ross, A; Sobre el Derecho y la Justicia, Eudeba, Buenos Aires, 1963.

Urbina, S; Legal Method and the Rule of Law, Kluwer Law International, The Hague, 2002.

Waluchow, W; Inclusive Legal Positivism, Oxford, Clarendon Press, 1994.

Personal tools