The work with datum and database: considerations through and enunciative theory of language

Abstract: The objective is to discuss the notions of datum and analysis when the theory in question is an enunciative one, bearing in mind two central questions: if the several enunciative theories question language through a semantic point of view, aiming at studying what is singular in the speech of each speaker, how can an enunciative study be based on a corpus and, even more, constitute a database? Are enunciative studies not refractory, in a certain way, to the notion of linguistic study based on corpora? In order to answer these two questions, we present the example of a database of language disorders, ENUNSIL (Banco de Dados Enunciação e SIntoma na Linguagem), considering that, if the notion of datum is problematic to enunciative studies, such difficulties augment exponentially when symptomatic speech is at stake.