Contenu du sommaire : Langue, praxis et production de sens, sous la direction de Paul Siblot

Should linguistics encompass the real ? Some think it shouldn't, but can we conceive of a meaning disconnected from its reference ? We only have access to the world such as we perceive it, not to the objective world, and what matters is that a stable intersubjective community can agree on its representation. Referential semantics presupposes a language which is oriented towards « the outside » . The fact that this exstant outside is firmly structured by language does not mean one should consider it as a mere discursive object. Referential conceptions postulate a homogeneous meaning. Our contention, however, is that meaning is heterogeneous and that according to the type of expression used it pertains either to a descriptive model, or to an instructional model, or on occasion to both.

From times immemorial, grammarians and linguists have assigned precedence to the noun over the other parts of discourse. The essentialisation of the nominal archetype, abundantly illustrated in the great mythical texts, leads numerous semantic analyses to belie the very theoretical principles which they defend. Our starting point is a reassessment of the definition of nomination, in which the noun is not viewed as designating the object named « in se », but « for us », thus nominating our relation to that object. A model of nominal significance is offered, in which praxis is viewed as bearing on meaning and in which we attempt to integrate the dialectic relationships between language and the real, whether those be inscribed in speech or mobilised through discourse.

Our analysis uses an oral corpus made up of sequences in which informants are asked a specific question i.e. to indicate the boundaries of their neighbourhoods. The answers obtained are utterances of the following type, « The neighbourhood runs from Street X as far as to Street Y » . How are we to define the roles and inter-relations of these « fictive motion » sentences in the production of meaning ? How can praxis — that concerning motion here — be inscribed in the forms of language ? In order to try and answer that question one is led to posit hypotheses concerning the meeting points of both linguistic representation of praxis and the praxis of linguistic output itself, i.e. of both represented praxis and representing praxis.

Any praxis is inscribed in time. To what extent does man's perception of time leave an inprint on Saussurian "langue" ? Revisiting a well beaten track in linguistic studies, we propose to give this farrcaching question a partial answer concerning the opposition between French imparfait and passé simple. We first capture the opposition between those two tenses in a passive way, in which man as a subject (subjected to...) bears witness to the flow of time from future to past ; we then capture the opposition in an active way, in which man as a project bears witness to the flow of time from past to future. This two-fold representation bears on the opposition between imperfect and simple past. It is shown next that linguistic theories view tenses as markers of localisation in time whereas verb forms also help mark how the time implied by the process is viewed, what orientation is given to it. Our contention is that implied time is represented in a downward fashion (particularisation) with imperfect and in an upward fashion (generalisation) with simple past.

When verbal transitivity becomes « invisible », one is confronted with a surface void. An interpretative calculus (calculation) then takes over which either concludes that this absence marks indeterminedness (absolute use), or reduces this void through environment scanning, whether it be context (Ø anaphora) or global situation (Ø deixis). In the last two cases the principle remains stable though their differ in illustration, which leads one to admit that for everything pertaining to things linguistic, one must take into account extra-linguistic reality and integrate it not only in the study of text phenomena that make up meaning but also maybe in the description of the system itself.

Grammarians describe a grammatical language which doesn't really enable one to predict what actual language use is like. The latter is a far more complex hyperlanguage which takes into account human discourse both as an empirical reality and as a practice inscribed in an environmental, communicational and social background. To learn a language is to learn how to navigate within a hyperlanguage. And the outside world itself partakes of meaning inasmuch as its perception, within the hyperlanguage, gives the reference whose externality has to be reckoned with. As for grammar books and dictionaries, they also partake of the hyperlanguage as an integral part of the environment of language users.