AAVE/creole copula absence: A critique of the imperfect learning hypothesis

This study confirms the robustness of the finding in the literature on African American Vernacular English (AAVE) and creole English (especially in the Caribbean) that omission of copular and auxiliary be varies systematically according to predicate type. Verbal predicates are associated with the highest rates of copula absence and following NPs with the lowest rates; following adjectives or locatives show intermediate rates (see Rickford 1998:190). Although this pattern is highly consistent, convincing explanations for it remain elusive. A recurrent suggestion (McWhorter 2000; Winford 1998, 2004; Wolfram 2000) is that the AAVE and creole English pattern is inherited independently from general processes of imperfect second language learning (simplification, generalization) that operated as the African ancestors of today’s speakers acquired English. In this paper, we pursue this possibility, but discover that the grammatical conditioning of copula absence in AAVE and creole varieties is distinct from the patterns found in second language learning data. We examine four sets of data on English acquired as a second language (Indian English, South African Indian English, Singaporean English, Spanish English) and show, using two statistical measures, that conditioning of copula absence in the second language data does not resemble the AAVE and creole pattern. (One possible exception is the high rates of omitted be with verbal predicates, for which we explore possible explanations.) We show further that typological diversity in copula systems also militates against a universal markedness-based pattern. The findings reduce the possibility that the overall AAVE/creole pattern derives from a general tendency in second language acquisition and increase the possibility that the pattern reflects a shared substrate influence from West African languages or other historical contact factors.