
Full text loading...
Bickerton and others have proposed models of creolization in which a creole with a bioprogram-unmarked grammar appears with the first generation of native speakers. When we construct the history of reflexives and anti-reflexives in Haitian Creole, we find instead a gradual development over more than 200 years, starting from a typologically unusual system that seems an unlikely candidate for the unmarked setting of the bioprogram, and passing through one or two intermediate stages to the typologically unmarked present-day system. A comparison with the limited available data on first and second language acquisition suggests that a model of creolization based on functional considerations and inheritance from a preceding pidgin will account for this history at least as well as a model based on first language acquisition. The history of Haitian Creole Binding Theory thus shows a classical "deep creole" acting much like Sankoffs analysis of Tok Pisin, and quite unlike the predictions of Bickerton's model or any model that predicts that a stable creole will develop in a single generation. This Haitian Creole data therefore implies a gradualist model of creolization, in which "creolization" is seen as a process extending over a number of generations of native speakers.