This dissertation explores a typological puzzle within the Romance language family: How is it that French could have evolved so divergently from the rest of the family that it has apparently lost two of the most characteristic features of Romance languages, i.e., (1) a highly inflected verb paradigm and (2) optional nominal and pronominal subjects? Using a ±60,000-word corpus of Everyday Spoken Swiss French (ESSF), comprising 14 educated, middle-class speakers conversing with family and friends, I examine the subject clitics and subject NPs from the theoretical framework of grammaticalization, since the data are gradient and represent change in progress. This type of data collection is crucial given the increasingly diglossic situation that exists between Spoken and Written French. The corpus shows that the subject clitics have fully morphologized 100% into prefixes in all but a few, predictably trailing, environments of the 3rd person. Thus the former clitics are no longer subject pronouns but person/number inflectional morphemes on the verb. To fill the void in the subject pronoun paradigm, a set of true, personal pronouns (moi, toi, lui, etc.) has moved in. But they, like other full NP subjects, are highly optional. Just as in Spanish and Italian, these overt subjects are used only sparingly, i.e., for contrast, emphasis, and disambiguation. When they do appear, they are most likely to occur pre-verbally, thus maintaining the traditional SVO word order of French, although alternate word orders also occur. The redevelopment of a highly inflected verb paradigm, albeit prefixal, not suffixal, and the consequent restructuring of the personal pronoun paradigm has allowed for the reemergence of null pronominal and nominal subjects in ESSF. The corpus also reveals a new, markedness-driven, morphological change starting in certain 3Sg. impersonal verbs, i.e., the development of a zero-morpheme. Based on these findings, two classic tests from generative grammar, Montalbetti's (1984) Overt Pronoun Constraint and Chomsky's (1981, 1982) Pro-drop Parameter are applied and show that ESSF behaves like other Romance pro-drop languages. This reanalysis of ESSF is discussed from the perspective of Romance typology, French Creoles, and first and second language acquisition. Finally, pedagogical implications are proposed.
展开▼
