Matrix-Language Approaches to Classic Code-Switching: the MLF and 4-M models under scrutiny

Abstract

Code switching is a phenomenon present in contexts where speakers use more than one variety. It is interesting for sociolinguistic studies to inquire into this behaviour and see whether it is an accidental behaviour or it follows certain rules and constraints. This paper aims to offer some of these grammatical aspects of code-switching observed in an Algerian context, mainly in the speech of bilingual speakers at university. Our prime objective is to explain the different manifestations of the formal properties of two typologically different codes; Algerian Arabic and French and their implications in shaping the syntactic/pragmatic structures in mixed codes. Then, our aim is not to attribute each switch a function since our frame of reference is not functional but we try to explain the ‘how’ and ‘why’ questions differently. The centre of our interest is the socio-pragmatic reality of syntactic constituents in bilingual speech. The analysis of our AA/Fr CS and MSA/AA code-switched data is based on Myers-Scotton’s MLF and its supportive models. In the MLF approach, there is always an asymmetry between the languages involved in CS. One of the major shortcomings of the MLF model discussed in the literature is the inadequacy of the notion of the CP as a unit of analysis. Reducing the matrix language to a property of a CP restricts the constraints on code-switching into purely structural limitations and therefore ignores other determining factors in shaping code- switched sentence patterns, be they socio-linguistic or psycho-linguistic.