When synchrony meets diachrony : The negation system of Fɛʔfɛʔ in the light of the Jespersen Cycle

Abstract

This dissertation is concerned with synchronic variation as the reflection and residue of language change. The empirical focus is the negation system of Fɛʔfɛʔ. The leading question is whether there exists a roadmap leading from the wild variation observed in Fɛʔfɛʔ to languages which like Modern English make use of a dedicated marker of negation. The thesis advocated is that language change is the means by which the Fɛʔfɛʔ type of languages can be related to the Modern English type. The argumentation builds on cross-generational variation within Fɛʔfɛʔ, on data from related languages, and on data from languages with historical records. From a Saussurean perspective, synchronic and diachronic phenomena do not in any way relate to each other. However, diverging from Saussure, Labow (1965) launches the apparent-time approach that builds on synchronic variation as a screen into diachronic change. Hence, it becomes possible to disclose the dynamic tendencies embedded in the synchronic system, thus relegating ‘the antinomy between synchrony and diachrony to the periphery as seen from the English, French, Cairese, and Welsh data explored in this thesis. From a Fɛʔfɛʔ internal perspective, I offer a comparative analysis taking into consideration the variation attested between teenagers and elderly speakers. In this regard, we observe that sì and si̍… bα̍ are invading the negation system, thus forcing the exit of older markers. Moreover, when the present adult generation is interposed between teenagers and elderly speakers, we observe between the three generations a subtle continuum bringing together archaic and innovating patterns peacefully co-existing within the overall speech community. Building on the parallelism between the Fɛʔfɛʔ data and the data from languages with historical record, I propose for Fɛʔfɛʔ an analysis of the same nature as the one advocated to account for similar kinds of variation across language families, namely, a Jespersen Cycle (JC) analysis. However, I do not embrace the standard version of the JC which implicitly postulates that the items under analysis are discrete entities stemming from discontinuous periods of time, with each time frame bringing its own marker as it goes. Rather, I perceive in cross-generational variation a miniature reflection of different periods of time imbricated into each other in the form of an undisrupted continuum. Hence, an item may span two or more generations and overlap with other items innovated in later generations. Moreover, this corroborates the facts observed in the present and past states of languages with historical records discussed, thus raisings for the parametric enterprise the challenge of addressing the situation of language internal variation that pervades across language families against the background of the standardly well accepted view that variation equals parameterization. Immediately related to this is the need to capture co-existing variants originating from different periods on a single clausal structure. I address this question after proposing a relative chronology for the co-existing markers attested in today’s Fɛʔfɛʔ. The broadly comparative approach adopted to sort out the Fɛʔfɛʔ data constitutes an inspiration for the analysis of languages that don’t have written records from previous generations.