On the Supposed "Gradualness" of Creole Development

Two recent works by Carden & Stewart (1988) and Arends (1989) have tried to prove a gradual rather than a single-generational origin for Haitian and Sranan respectively. Both arguments, however, are severely flawed. The Carden-Stewart argument from Haitian reflexivization is shown to depend on misinterpretations of both bioprogram theory and generative principles. Further, their claim that early Haitian was not a full language would entail that Middle English (among others) was also not a full language. Arends' claims of radical diachronic change in Sranan involve treating as an early creole sample a fragmentary text which, given the social and historical context of seventeenth-century Suriname, was most probably produced by a second-language learner of the creole. Reanalysis of Arends' data shows that he exaggerates the significance of marginal forms and mistakenly treats the inherent variability characteristic of all languages as evidence for ongoing change. In fact, none of the data reviewed in these works is inconsistent with the emergence of Haitian and Sranan as full languages in a single generation.