Some Morphosyntactic Aspects of French/English-Bantu Code-Mixing: Evidence for Universal Constraints

Abstract

An examination of one of the current issues in code-mixing (CM) research: determining whether universal constraints on CM exist. This issue is addressed in light of data drawn from CM in English /French with some selected African (Bantu) languages, including Ciluba, Lingala, & Swahili. It is suggested that in the African context CM is governed by a general principle, the matrix code principle. This is the idea that in CM involving languages, L1 (eg, an African language) & L2 (eg, a Western language), where L1 is identified as the host (or matrix) code & L2 as the guest (or embedded) code, the grammar of L2 must conform to the morphosyntactic structure rules of L1, the language of the discourse. Four pieces of evidence are presented that provide support for the proposed principle. It is shown that (1) in accordance with the morphosyntactic structure of the African language involved, a French/English article is omitted before a French/English noun used in French/English-Bantu code-mixed variety; (2) following Bantu syntax, French/English verbs used in this variety are inflected with Bantu rather than French/English inflectional morphology; (3) French/English adjectives are postposed rather than preposed to the nouns they modify, as required in Bantu languages; & (4) French/English infinitives used in the variety under consideration undergo "reinfinitivization," a process whereby they take on an additional infinitive marker, the Bantu infinitive marker, so as to fit Bantu verbal morphology. AA.