Abstract

Creoles are notorious for the extreme variation they display within a single variety and across varieties. Such variation has been interpreted as symptomatic of alternate grammars punctuating a creole continuum consisting of a basilect, acrolect and mesolect (Le Page and Tabouret-Keller, 1985). However, I have shown in Baptista (2015) that distinct creole varieties (Santiago and São Vicente) of Cape Verdean Creole that have been traditionally described as basilectal and acrolectal do not uphold such characterization upon closer examination of their grammatical systems and of the speech patterns of individual speakers. In Baptista (2015), a careful analysis of the Santiago and São Vicente varieties of Cape Verdean Creole showed that Santiago for instance which has been typically described as basilectal actually displayed acrolectal features in core domains such as gender agreement and Tense, Mood and Aspect markers (TMA). The study of individual speakers also demonstrated that the same speaker may display speech patterns that swing like a pendulum between the acrolect and basilect. This study brought in plain view the descriptive inadequacy of the labels basilect, acrolect and mesolect and challenged well-established claims regarding the status of the Santiago and São Vicente on the Cape Verdean Creole continuum.
In light of the inadequacy of the creole continuum, I propose in this paper an alternative approach to the examination of variation observed in creoles. The main research questions driving this study are as follows: Is the variation observed in creoles better accounted for in terms of competing I-grammars that emerge over time? Would this new lens help us attain a higher level of descriptive and explanatory adequacy than the creole continuum provides?
In order to fully explore this topic, I carefully examine synchronic and diachronic data as well as acquisition scenarios by targeting the TMA system of Cape Verdean Creole (CVC) and by comparing the degrees of variation that this particular creole displays at these three levels.
This paper is organized as follows. Following the introduction, in the second section, I lay out the traditional definition of the creole continuum and briefly summarize the arguments I made in Baptista (2015), pointing to its limitations and descriptive inadequacy. In the third section, I turn to the framework of competing I-grammars which I argue is a more descriptively and explanatorily adequate approach to analyzing the observed variation. I provide a definition of I-language in light of competing I-grammars, drawing from Lightfoot (2006), Yang (2002), Obata & Epstein (2015) and Epstein (2016) and show how the approach of competing I-grammars applies to the synchronic study of TMA markers in the creole under investigation. In the fourth section, I elaborate on a possible scenario of acquisition of TMA markers in CVC using diachronic texts in support of such scenario and in the fifth section, I propose a syntactic analysis accounting for their synchronic distribution. The sixth section summarizes the key findings of this study.