The Roots of Negative Concord in French and French Based Creoles

2 Introduction Within the Principle and Parameters model of Universal Grammar (UG), double negatives, which I will henceforth refer to as Negative Concord (NC), have been argued to invoke a fundamental syntactic relation between the head and the specifier of a constituent. More precisely, it has been proposed that sentential negation universally projects a syntactic constituent conform to X' theory composed of a negative head and of an optional dependent specifier (Laka 1991, Zanuttini 1991, Pollock 1989), either base generated in Spec NegP or moved to it. Within this structural approach to negation, Negative Concord is taken to illustrate the concurrent overt manifestation of both a negative head and a negative specifier which combine through overt or covert movement to form a single negative constituent (Zanuttini (1991)). This analysis has desirable consequences. In contrast to a long standing grammatical tradition which condemns double negatives as logically faulty, the structural approach to negation fully legitimizes this construction as one of the possible option provided by UG for the expression of negative meaning. It also establishes a parallel with the by now classic approach to overt questions, multiple questions and relative clauses in terms of Spec-Head relations. On this view, parametric variations are expected between languages in which negation is manifested with an overt head alone, with an overt specifier alone or with both an overt head and an overt specifier and Negative Concord could be conceived as an unmarked UG option if overt realization of both the Spec and the head of the Neg constituent could be shown to be a default option for constituents of this nature. This is an advantage in view of the fact that Negative Concord may well be the most " unmarked " expression of negation. As is well known, it appears as a possible historical state in many languages (Cf. the " Jespersen cycle "). It has also been observed to be a common feature to most Creole languages, irrespective of their origins or the nature of their substrate and superstrate. If as Bickerton (1981) has argued , Creole grammars instanciate the least marked options permitted by UG, then the use of double negatives, far from being accidental or illogical is probably better thought of as " the norm " rather than the exception. In this paper, the structural analysis of Negative Concord is put to the test in light of comparative data from French …