Last week, I attended an interesting and well-organized conference in Paris entitled Towards Proto-Niger-Congo: Comparison and reconstruction. The talks represented most of the major branches of Niger-Congo, as well as stock-level surveys, and some of the highlights from my own perspective were the following:

• Gregory Anderson’s discussion of S/TAM/P (pronounced as “stamp”) morphemes in Niger-Congo where one often finds “portmanteau morphs that encode the referent properties of subjects in addition to various TAM and polarity categories”. The presence of such morphemes is particularly concentrated in Niger-Congo languages within Africa (though does also go somewhat beyond the stock). The significance of this work is that it suggests a possible idiosyncratic grammatical feature beyond noun class systems (and, possibly, verb extensions) that could be used to help more conclusively establish the relatedness of the proposed Niger-Congo sub-branches, even those, like Mande, where noun class systems are not attested.
• Ulrich Kleinewillinghöfer’s paper which was primarily focused on the noun class systems of a group of Adamawa languages but also showed how these languages help strengthen the case for a close relationship between Adamawa languages and Gur languages within Niger-Congo. These groups are otherwise separated geographically, making such a connection of clear interest for reconstructing the history of the stock.
• Guillaume Segerer’s paper on subgrouping within Atlantic which was noteworthy for using traditional sound correspondences to establish subgroups rather than the more usual lexicostatistics which have dominated comparative Niger-Congo classification. This is especially important for Atlantic where lexicostatistics yield surprisingly low figures of cognacy (perhaps because historical language taboos have caused frequent lexical replacement, as someone suggested orally at the conference—I believe Guillaume himself).
• John Watters’ paper on reconstructing the tense-aspect system of Bantoid. Even though some Bantoid have highly developed tense systems (including, for instance, several degrees of remoteness in the past), he concludes that these are most likely to have developed from a system that did not mark tense morphologically but, rather, was organized around something like a perfective/imperfective opposition.

Of course, there were many other nice papers, and a complete list of abstracts is available online.

I have two other observations coming out of the conference discussions as a whole. First, while perhaps not surprising, the discussion was on the whole skewed towards issues relevant to specific subgroups, rather than the family as a whole. This reflects the fact that, in a family so large and diverse, there really are not many people who attempt studies of the entire family. Rather, one has Atlanticists, Mandeists, Bantuists, etc. It made me wonder if the situation is similar for other large and spread out families, such as Austronesian or Afro-Asiatic. (Indo-European is, perhaps, a special case where my impression is that there are both lots of subgroup specialists and whole-family specialists.)

The second relates to the special status of the Mande group. There is some controversy regarding whether it belongs within Niger-Congo at all (see Dimmendaal 2011:323), but even setting that aside, Mande is noteworthy in being a relatively large and dispersed subgroup with, apparently, very clear-cut boundaries. That is, it is always clear if one is dealing with a Mande language or not. This is not true for other larger subgroups of Niger-Congo. For instance, no clear criteria have ever been established to separate Bantoid from the rest of Benue-Congo, the unity of Atlantic is disputed, and even geographic offshoots in Kordofanian are not clearly a proper subgroup of Niger-Congo (Dimmendaal 2011:324). Mande is not completely unique in this regard as groups like Kru and Ijoid also, to the best of my knowledge, have fairly clear boundaries, but it does pose an interesting historical question, that does not seem to have received much explicit attention: Why is the membership of Mande so clear when the membership of, say, Benue-Congo is so much harder to establish?

Reference

Dimmendaal, Geerit J. 2011. Historical linguistics and the comparative study of African languages. Amsterdam: Benjamins.

jeffgood

More Posts