0 - 521 - 87611 - 7 - A Linguistic Geography of Africa

More than forty years ago, Joseph Greenberg (1963) demonstrated that the African continent can be divided into four distinct genetic phyla, or families as he called them, namely Niger-Congo (or Kongo-Kordofanian), Nilo-Saharan, Afroasiatic, and Khoisan. For subsequent generations of Africanists, this classification has served as a reference system to describe the relationship patterns among African languages. In this tradition, scholars doing comparative work on African languages were preoccupied to quite some extent with reconstructing and understanding similarities across languages with reference to genetic parameters. One effect this line of research had was that an interest in other kinds of linguistic relationship was never really pronounced. Especially the question of whether, or to what extent, structural similarities and dissimilarities among African languages are the result of areal, that is contactinduced relationship, has never attracted any major research activities beyond individual studies dealing with lexical borrowing and related subjects. Whether the African continent constitutes an areally defined unit, or whether it can be subclassified into linguistic areas (or sprachbunds, or convergence areas) remained issues that were the subject of casual observations or conjecture, or both, but not really of more detailed research. Still, once more it was Greenberg who drew attention to the importance of areal relationship in Africa. Not only did he venture to point out major linguistic areas (1959), but he also was the first to come up with important findings on the areal distribution of phonological and morphosyntactic properties across Africa, and with hypotheses on the areal distribution of these properties (1983). The title of chapter 2 of this book echoes that of a paper by Joseph Greenberg (1959), and this choice is deliberate: with this book we wish to build on the foundations laid by Greenberg, demonstrating that in the course of the last decades some headway has been made in areal classification since his paper appeared nearly half a century ago. A common thread to all the contributions of this volume is that genetic relationship is far from being a parameter for understanding many of the processes characterizing the history of and typological relationship among African languages, and the message implicit in these contributions is that for a