A Linguistic Geography of Africa (review)

There are around two thousand languages spoken in Africa, almost a third of the languages of the world, belonging (aside from a handful of as yet unclassified isolates) to four phyla or superfamilies. The variety of language typologies to be found in Africa is considerable, yet it has long been proposed that there is a recognizable African language “type.” Until Joseph Greenberg’s seminal study (1963), attempts to classify the languages of Africa tended to confuse typological with genetic considerations. The expansion or, indeed, explosion in African linguistics since Greenberg’s work has provided an ever-increasing body of data that have not only consolidated the larger part of Greenberg’s proposed genetic groupings, but have also hinted at the complex threads of interaction between genetic and contact-induced factors that underlie the relations between African languages. Indeed, genetic relationship is clearly not the only parameter for understanding many of the processes involved in the development of African languages and the relationships between them. As the editors say in their introduction, more detailed information on areal relationship patterns is a sine qua non for a better understanding of African language history. The same, of course, can be said for languages in the rest of the world. Given the growing appreciation of the importance of contact-induced change, and, since Dixon’s work on Australian languages, of the almost certainly equal relevance of convergence alongside divergence in the history of languages generally, the present volume is a major contribution not only to African linguistics, but to linguistics in general.

The book brings together fourteen scholars of African linguistics, many of whom are amongst the leading names in the field today, to examine the contact situations across the continent more closely than has been done hitherto and to answer three fundamental questions identified by the series editor in his foreword: Why are genetic and typological classifications of African languages not coextensive? Why are there numerous typological features that cut across the established genetic classifications? And finally, how significant is the role of areal diffusion and thus of language contact in the various affinities observable between so many African languages?

In order to present a general perspective on areal relationship in Africa, the book comprises nine chapters. Following the introduction, which lays out some general discussion and an explanation of the raison d’être for the volume, Bernd Heine and Zelealem Leyew’s chapter (pp. 15–35) attempts to address the basic question that underlies the entire book: Is Africa a linguistic area? This may be something of a simplistic question, but the answer is far from straightforward. For one thing, defining what should be understood by the term “linguistic area” has itself not always elicited unanimity amongst linguists. Generally, a linguistic area is a region where a number of contiguous languages share a number of linguistic features “whose presence can be explained with reference neither to genetic relationship, drift, universal constraints on language structure or language development, nor to chance” and where such a “set of features is not found in languages outside the area” (p. 16). Adopting this purely linguistic definition without recourse to extralinguistic factors such as historical or geographical forces and criteria, the authors conclude, unsurprisingly, that it is not possible to consider the whole of the African continent as a single linguistic area in terms of a set of linguistic properties that are found in Africa and nowhere else. Nonetheless, they go on to propose that there is a range of typological features that are found in Africa more extensively than in the rest of the world, such that it is, after all, possible to predict that any given language is more likely to be an African language than not. This general statement does need to be tempered with the finer tuning arising from the ensuing discussions in the book. The chapters that follow therefore further hone the editors’ findings, first by looking at evidence from two specific...