CogniTextes (Dec 2010)
Compositionnalité gestaltiste et construction du sens par instructions dynamiques
Abstract
The pervasiveness of polysemy in languages compels us to consider the interpretation of utterances as a dynamic process in which the meaning of each linguistic unit is determined as the overall meaning of the utterance is constructed. This interactive process, that can be called gestalt compositionality, is radically opposed to the classical mechanism of compositionality advocated by linguistic formalisms based on the primacy of syntax. By contrast, our approach amount to considering the syntactic structure of an utterance as a byproduct of meaning construction, and no longer as its starting point. The interpretation of an utterance thus directly hinges on a process of interaction between the various basic components of that utterance: lexical units, grammatical markers, order relations between units or, more generally, basic constructions in the sense of construction grammar. Modeling this dynamic process is a major issue for all linguistic trends that reject the classical principle of compositionality, like the cognitive grammars and the theories of énonciation. In this article, we present a theoretical framework in which the interaction dynamics follow a principle of convocation-evocation: the various basic components of an utterance contribute to the construction of a global representation in an intersubjective space that we call verbal scene. Each component is used to evoke a new element of the verbal scene under construction, but to do so it must first convoke other elements present in the same scene or in the interlocutory situation. The principle of convocation-evocation is what defines the identity of a linguistic unit. We express this identity in the form of a dynamic instruction of meaning construction and we consider that every linguistic unit yields such an instruction. Thus, meaning is indeed the result of a gestalt composition process, since the contribution of each basic component depends on the contribution of other components in the utterance. With examples from French and English, we demonstrate a first attempt at modeling this process. We will also discuss the very special cognitive status of the verbal scenes, and defend the idea that the specificities of human language follow from this function of representation in an intersubjective space, the term “representation” having to be taken not in its usual sense in the cognitive sciences, but in the original sense of its Latin etymon: repraesentare – make present.
Keywords
