This special issue of the International Journal of Bilingualism represents an important contribution to the field of code-switching (CS) not only because it is a joint collaboration between scholars based at the University of Ghana and New York University, but also because it presents new CS data sets. The contributions in this exciting volume constitute a particularly welcome addition to the still growing but already extensive body of literature on CS, for various reasons. Firstly, it brings a new group of scholars into the discursive arena. Secondly, very interesting new data sets are made available, stepping in some cases outside the mould of non-European matrix language/ European embedded language, such as the papers on Kabiye and Logba interacting with Ewe. Thirdly, a broad range of methodological approaches is presented, with techniques ranging from media studies to fine-grained phonological analysis.The language pairs that we learn about in this volume are spoken in Togo (Kabiye-Ewe) in Ghana (Akan-English, Ewe-English, Logba-Ewe) and by Ghanaians in Italy (English-Akan). The structural and sociopragmatic accounts of these data, for the most part, confirm current explanatory models such as Myers-Scotton's (1993a, 2002) matrix language frame (MLF) model, Muysken's (2000) three-way typology of CS patterns involving alternation, congruent lexicalization and insertion, as well as Gumperz (1982:83) we-they codes and Auer's (1984, 1995) language negotia- tion sequences based on conversation-analytic exchange structures.The present commentary reflects on sociolinguistic, pragmatic and structural/psycholinguis- tic issues relevant to the West African context and the specific CS data analysed in this volume. There is wide recognition in the field that the social, historical, political and economic context of language contact is crucial for understanding the way CS gets used and hence the pragmatic, conversational and structural shape of the data. This consensus is expressed by Bullock and Toribio (2009, p. 14), who point out that there are three major strands in the study of CS: the structural, the psycholinguistic and the sociopragmatic 'although, in principle, a full account of CS cannot be achieved without the integration of findings from each of these strands ...'. Each explanatory paradigm (sociopragmatic, structural and psycholinguistic) responds to concrete research interests and each involves quite different epistemologies about what counts as the relevant research questions, but also one must take into account different ontological considera- tions regarding what counts as reliable data. In the case of this commentary, we seek to bring together ideas that will help to make some of the links between sociopragmatic and structural CS in West Africa.Given our separate areas of expertise, the three authors have decided to elaborate separate but interrelated sets of commentaries. Melissa Moyer begins with insights brought out in the papers about the sociolinguistic and sociopolitical context of CS in Ghana and Togo. Peter Auer focuses on aspects of pragmatics and conversational structure and, finally, Pieter Muysken comments more specifically on ideas regarding grammatical processing and language typology.Some sociolinguistic insightsA sociolinguistic account of CS in West Africa must start by taking into account the sociopolitical conditions of contact between European colonizers and the various majority and minority ethnic groups that have existed historically and that have shaped language contact and the CS pairs that are discussed in the present volume. Each one of the articles gives contextual information on the situation of language contact. Dorvlo, in his contribution, provides a rich background account of language contact as relevant to speakers of Logba, an endangered language from the Na branch of Ghana-Togo Mountain (GTM) languages. (The classification of the 14 GTM languages, formerly called the Togo Remnant languages, remains unresolved. …
Read full abstract