When English meets isiXhosa in the clause: An exploration into the grammar of code-switching1

The phenomenon of code-switching accords linguists the opportunity to study the interaction between two distinct grammatical systems. The contact languages may have different specifications for corresponding lexical items; and a likely outcome of utilising two linguistic systems in the same speech event is the creation of a ‘new grammar’ in which some lexemes exhibit loss of certain abstract features. Using Myers-Scotton’s (1993b, 2002, inter alia) Matrix Language Frame and the 4-M model, this study examines naturally-occurring  isiXhosa-English data, collected from Rhodes University students in Grahamstown. The study shows that in bilingual speech some English transitive verbs lose their typical subcategorisation features such that they have to be ‘retransitivised’ by isiXhosa affixes. The paper argues that code-switching data  can provide insights into the nature of lexical entries across languages as well as the patterns of grammatical convergence which result from language contact. Southern African Linguistics and Applied Language Studies 2011, 29(2): 127–13