Accueil RED&S


Language theory, semiotics: critism of legal positivism?

Anthony-Patrice Chamboredon
June 1995

  1. "Speech act" theory and legal positivism
  2. Semiotics and legal positivism



I- "Speech act" theory and legal positivism

Origins of the speech act theory

Legal theory is necessarily involved with question of language, meaning and interpretation. Language has increasingly become a major object of inquiry within divers range of legal theory. This is partly due to the work of J.L. Austin and his speech language theory, between 1962 and 1973. At that time discussions were very passionated, then, after a less popular period, it refinds a great interest at the begining of the eighties. In France, it has penetrated philosophy thanks to the works of Ducrot and Ricoeur. The first publication on this subject dates from 1974. In Europe the situation is quite similar with one remarquable exception in 1962 with the Swedish K. Olivecrona and the Finish G.H. von Wright. But in 1972, Alf Ross published an article with a significating title:"The Rise and Fall of the Doctrine of Performatives ", in one hand he rejected following Austin's first distinction between performatives and constatives, or more generally, between "saying something" and "doing something", but on the other hand, he didn't accept the second version of the theory, with the distinction between the locucionnary, illocucionnary, and perlocucionnary act. Alf Ross distinguished only two main types of speech act: indicative and directive, in his book "Directives and Norms". Each of these acts had one or several norms behind to determine its normative consequences. These norms were called "norms of competence".

Contribution to the question of the legal language specificity: a frontal attack against basis of the pure theory of law.

The theorical reason of the language theory attraction exerciced on jurists is that in the sixties legal theoricians and philosophers went to a saturation point caused by kelsenien theory. Thus performative theory allowed a frontal attack against the bases of pure theory by supressing gründ norm "hypothesis". In admitting existence of performative speech acts, which didn't ask existence of superior norms, we could denounce what Mc Cormick called "the imperativist fallacy". But it is also important to stress that speech act theory allowed to oppose oneself to the epistemological rupture, inherited from Hume and Kant applied in our field by Kelsen between Sein and Sollen, Is and Ought . Some jurists had this deep intuition that this radical scission which closed them in the "sollen world", the one of signification of will act, cut them from reality, and deprive legal reasonning from perception fondations sining by its irrationalism. Though, doctrine of performatives by its theory of presuppositions reinaugurate the link between speech act and its factual presuppositions. It is the reason why the speech act theory has been used in legal language theory and logic and law, with a peculiar accent on norms theory, thus as derivation of norms from indicative propositions. And it has also been applied to analyse what we call legal act, particuliarily judicial decision thus as explanation of judicial facts and institutions.

But in fact, what jurists expect from philosophers is the key of the solution of their old problem to truly know what specificity of the legal language in relation to other types of language consist in. And in this very question lay a difficulty, or may be a paradoxe: philosophers analyse speech acts in the very structured tradition of the common language theory while jurists are not totally persuaded that the legal language was precisely "common". We are not talking here about some lexicographical specificities of every language, as professional or techniqual languages inevitably contain; to this extent the legal language is in the same situation that what ever science or pratical discipline is. The problem is neither at the semantic nor syntaxic level, because under this aspect, the legal language is totally shaped as common language, even if some of its terms or expressions have a very different sens. In other words, jurists knows that technicity, see even the oddness of some expressions composing the legal language (cf. picture, p4) could be replaced by a terminology coming only from the common language, whithout that, specificity of the former relating to the second disappear automatically. What jurists seek for, is the specificity of the relation that their language maintain with reality, because they are convinced that this relation is different with the one that relates to common language with the corresponding reality. The question to philosophers is nothing less than the relation of law with the world of extra legal realities. It is the specificity of the phenomenum "law" which is at stake. But speech act theory, like the first version of Austin on performatives, can be inoperating for jurists if we only consider general intuition according to propositions of law have as distinctive quality "to do something with words" because this simple truth is well known by jurists. We know that in law, words like "do" bind marriage or transfer or share goods, condemn and put in jail, sometimes kill, create things and facts or make them disappear. At least, they can be confirmed in their idea following wich, the relation to legal language to its corresponding reality is not passive reproductive but active productive and constituing as some expressions of the common language used by Austin in his first version theory, like the promise.

( The "specificity" of legal language ...)

— "La Cour, vidant le délibéré, et adjugeant le profit au défaut, met l'appellation, et ce dont est appel au néant, émendant quant à ce, corrigeant et réformant la sentence des premiers juges, décharge l'appelant, condamne l'intimé aux dépens de l'incident, dont distraction au profit de Me Bizotin avoué, qui la requiert pour le surplus des fins de la demande, met les parties hors de cause et les renvoie dos à dos, dépens compensés"...
Saperlotte quel jugement! ... mon avoué va me demander au moins soixante quinze francs pour m'expliquer la chose!...
(Daumier, Les gens de Justice)

If the speech act theory was limited to this simple constatation, it would'nt be so important and even misleading if the conclusion was to say that the legal language is only a under-sector of the common language, the one where we do things in speaking. Plus, it would be twice wrong, on the one hand, because in law we also express propositions to describe and to note, and on the other hand, because doing something in speaking in the common language is totally different to do these things in speaking the legal language. It is thus the case with promise, promise in the common language is far different from legal language, in the first, the act of promise just creates a promise while in law, a promise has a second effect, irreductible to the first, the creation of a fact or a legal act with all the bundle of rights and legal obligations following from. In other words, in law we do not speak to do speech acts, but to carry out, through them, legal acts.

A radical difference exists between "doing something in speaking" and "doing something in law in speaking". Legal speech acts are very closed to the common language called "performatives" but they are not identical. The solution saying that the difference consists in the pragmatic context is not sufficiant. Because these contexts are in fact for exemple the civil or penal code with difficulties of interpretation and application, thus law itself. It is a vicious circle where Law constitutes the criterium (pragmatico linguistic) deciding what is legal and what is not. Where can we find the criterium of distinction. Two differences seem to be relevant. The first difference is at the level of what Austin call "the conventionnal procedure" whose the observation determine success or failing of performatives which seems to be very poor in comparison with the judicial proceeding, the one for exemple that determine validity of judicial acts. Formalisation of those last is such that we are in presence of a difference in the quality of the object. They do not establish conditions for success of legal performatives as speech acts, but as legal acts. There are thus, two different procedures: the linguistic determining the success of some enounciations as such type of speech act, the other, legal, stipulating conditions of the succes of this speech act as legal act. A speech act as performative at the common language level may succeed and fail as a legal act for exemple, because of a legal incapacity of the person who has contracted. The second difference is more fondamental. The aim of legal performatives is to create facts not only intersubjective acts, valuable for the parties in the conversational context, but also for the other, even for those who haven't any knowledge of the relation that binds them. The effects of the legal performatives are "erga omnes". More over these effects may escape to their own authors by the requalification of a juge, independantly to the will of the parties concerning the signification they have wanted to give to their act.

In conclusion, performativity of a speech act cannot be considered systematically as a synonyme of its juridicity. But it doesn't mean that we cannot take any advantage of the Austin' speech acts theory . This theory has evoluated since 1962, with new analysis very creating for our discipline. An example with the distinction of Searle' two categories of rules, the normatives and the constitutives that we can refind in a another form with Hart who distinguished primary and secondary rules, the first prescribing to carry out or to refrain from carying out some behaviour, the second conferring powers, or more generally, concerning the creation and modification process of duties and obligations. The main thesis of Hart, following which, law is the union of these primary and secondary rules, owe for a great part its formulation to the analysis of the speech act theory that the oxonien theoricien new very well for having participate to its elaboration. But it is commonly recognised that his thesis oppose oneself to the normativist analysis of the judicial reality, and more generally to the vision of law as constraint order composed by primary rules only with sanctions. Finally in legal logic, the affirmation of Searle according to "inside of some constitutive rules system, it is possible to derive "ought" from "is". This slits the wall of the transcendantal distinction between "sein" and "sollen" and open a new and immense philosophical possibillities for the doctrine of legal rationality as a form of practical reasoning, which is finally a return to the antiques and medieval sources, with wich law has never breached.


As we have seen, speech act theory can constitute a mean to find a breach in the kelsenien positivism theory. May semiotics, as a new form of language theory, breach also the wall of positivism?

II- Legal positivism and semiotics

Origins and definition of semiotics

The term semiotic stems from Lockes who wanted to design a project of a science of sciences. But more generaly speaking, semiotics may be defined as the studying of signs and system of signs used in communication. It is also a level of language analysis whose the aim is the constitution of a data language theory whith syntaxic, semantic and pragmatic rules. The first great precursor in the 19th century was Ch. S. Pierce. The project of signification theory went on production of models and methodological processes for an empirical analysis of discourse and signifying socials behaviour.

Applyed to law there are two mains types of "legal semiotics", one is oriented towards logical formalisation of legal propositions, the other is attached to the construction of a legal grammar as a set of rules governing the production and interpretation of speeches and social practices with a "legal" value.

The first drafts in semiotic of law were drawn by Felix E. Oppenheim in 1944, it has been constitued by an analogy with semiotic of logic and mathematic language systems. Semiotic such understood, found its theoricien in Ch. W. Morris and R. Carnap who divided semiotic in syntaxic, semantic and pragmatic. This theory was an application of mathematic methodology to the science of natural languages and artificial languages. But since Br.Wrobleski we may divide the legal semiotic between semiotic of legal language and semiotic of the language of the jurists. But semiotics as project of grammar of law comes from a distinct tradition, essentially from linguistic of F. de Saussure, L. Hjelmslev, E. Benveniste and anthropology with C. Levy-Strauss. This approach globaly qualified as structural is original regarding the choice of its subject matter and its level of analysis. First, in that sens semiotic is not a science of signs, but much more a science of the signification. But signification is every where, speeches, behaviour, every phenomena in the world has some signification. It is the case for law. And semiotic intends to approach the legal phenomenum in its globality, that is to say as the main social signifying dimension. This very ambitious project implies a choice of level of analysis closed to those which has been adopted in legal theory and sociology of law. If the legal semiotic may bring to the jurists some elements of description of what law is, it enrichers the understanding allowing to compare methodologically some different types of normative speeches to other social practices. Thus, very numerous are the problems traditionnally studied by legal theory whose legal semiotic may given some responses or at least some complementary lights.

Here our concern is about knowing wether semiotic may be a radical criticism to legal positivism or complementary?

For Bernard S. Jackson "semiotics is a radical criticism to legal positivism". Though, for the application of semiotic to law, he explains that three elements must be combined. The first is a positivist methodology in that sens, it needs to be used in the presence of an empirical subject matter well defined, texte or other signifying object; the second is naturalistic epistemology, in the extent that the "structures élémentaires de la signification" get an extreme cararacter of generality; the third is a realistic ontology in that sens that there is no for the semioticien, in the analisys, to consider an other reality than the significations belonging to the studied object.

But the first element, the positivistic methodology is mainly influenced by Saussure and Greimass theories. But the question posed to them was wether our understanding of language could provide some correspondance between our capacity of communication and the real world to wich it refers. In other words, how to provide an objective and universal grammar while our situation is such that we cannot be beyond the historical/cultural specificity of the language which we have made and through which we relate? - This division between objectivism and relativism of language has found its expression in feature of structuralist semiotics which attempt to restrain language as an object of knowledge and thus adopts a positivistic approach.. So, the criticism which has been made to the structuralism is that it tends towards an ahistorical or essentialist view of language and meaning. And it true that a variant approach has been the discourse theory so labelled post structuralist when the form is shifted to the historical/cultural specificity which requires a change for language to discourse, or in term of Saussure opposition, from langue to parole. The most proposition of his theory is "meaning is not desirable from signs, and thus cannot be a closed system".

But even if Greimas' semiotics mist upon the dual semioticity of language, that is the relation between the structure sign and their inter relationship and the narrative with which they operate, a critic still remain, it still employs a priviledge or essentialist view of language having a core. Thus still there is a close intellectual proximity betweeen positivists semiotics and dominant tradition of legal positivism. It is the reasonwhy, in adopting not only the same methodology but also the same epistemological tenents of greimasian positivist semiotic, A. Hunt considers that seriously limits the Jackson's critical capacity on legal positivism.

For instance, one of the most aspects in Jackson's criticism to positivist thesis is about the unity of legal system. His most general contention wich is advanced stem from Greimas' insistance that law is a dual semiotic system. It implies an analytic distinction between the units of language (words) and text. On one hand a system of meaning words and, at the same time, a system of meaning of the text which is not reducible to the words from which it is constituted. The Greimasian analysis of the semiotic proposes of its basic units (words) provides the key to the discourse as a totality (text). It is a contrast with the discourse theory wich does not privilege the semantic level, rather it gives priority to the social hierarchies and the exercice of power. Whilst legal semiotics identifies law as a dual semiotic system, Hart's application of ordinary language approach law as a unitary system which employs a natural language.

And Jackson's substantive break with legal positivism lies precisely in this denial of the myth of unity of legal system wich involves a brief application of discourse analysis. He makes the point that the myth is a message, that is that the legitimating claim of state legal orders attach considerable importance to the presumption of unity. But it seems to hold back from comitting himself to the view that there exists a plurality of autonomous but interacting legal systems. The very myth of unity appears to ressort itself in his formulation. "The law consist not in a single semiotic system, but in several. This feeling is reinforced when he distinguished between three contexts which law distinct semiotic system, (the meaning of the statute, its doctrinal interpretation and its judicial application). And the other "contexts" to which he refers, namely lawyers-client interaction, in and out of the court settlement, and also by his observation that in this different context the semiotic system have different audiences. He still make a less radical claim than the rejection of the presupposition of unitarity legal system so charasteristic of legal positivism with which he shares such common ground. It still unproven that the application of semiotic theory, at least in the form developped by Jackson can provide a critical thrust to the tradition of legal positivism.

Jackson replies that the denouncing of semiotics as a positivist science and a suppplementary analysis level to the traditional exegetic studies under the pretex that semiotics prileges only the textes is a misunderstanding. His two principal arguments are the following: The gremassian theory is constituted from an hypothesis which concerned the "structures élémentaires de la signification". It is a logical prevaillance accorded to the deep level. These structures exert constraints with effects at the level surface. But these constraints are only formal and general, and they are not able to devaluated a critical approach of the expression forms or of the own contents of manifestation at the surface.Secondly, semiotic is not only an analysis of textes introducing a different more abstract meta language. It does not privilege texte regarding other signififying manifestations (gestures, iconic, proxemic, spatial, musical, and so on...). Landowski writes that "our goal over take limits of the only textual analysis: what we call the legal, it is not only a vast corpus of linguistical expressions, it is also a bundle of institutions and actors, of situations and decisions, of facts and judicial acts whose its understanding as a globally signifying system needs the construction of semiotics models which obviously, wouldnt be only textual or linguistic . And effectively, semiotics want to be a positive science regarding its methodology, but there is no contradiction in the fact that this science attemps to over take the limits of the positivism to englobe the whole of the critical analysis of the law. It is why semiotics does not belong to the legal positivism but to the realistical tradition of legal philosophy. For the special reason that semiotic move away all form of reification of its subject matter, and then every form of reification of law. As C.A.D.Husson , to think about law as a language, is to treat him as a medium, is to show that it is an instrument of action, an operator, not a thing.

Then the two following positions may be concede: One, semiotics is supplementary" to positivism, in that it accepts the positivist definition of when a law exists (in terms of sources" theory of law - in Hart's terms, " Rules of pedigree", which require normally the doing of some act by an official source), with semiotics merely asking how the sens that those criteria have been fulfilled is conveyed (or, what is the evidence that the law has been passed through the official sources); and two, semiotics, while positivist in the sens of being empiricist, does not accept the positivist definition of when a law exist, but seeks to explain the sens that a law exists, whether or not the relevant" rules of pedigree" have been fulfilled.

But Jackson maintains the second position. And he adds that "a sense that law exist "may be quite differently constructed whithin different speech communities- the courts (who are the most likely to take the positivist" position in sense (a) above, practitionners (who may often rely upon professional interaction plus unofficial written sources); the public (via the media, etc).

Then, if we understand semiotics as a criticism rather against normativism than legal positivism in general, we are quite agree specifying that it is a radical criticism on normativist approach only .


© Réseau Européen Droit et Société