ABSTRACT
We would have liked to present here the first comprehensive non-lexically based historical classification of the Bantu languages. For us that would be a statement that reflected, not so much the origins and whole evolution of the languages spoken by Bantu communities over the 5,000 or so years that have elapsed since late Proto-Bantu (PB), but rather the development of their early ancestors out of PB during the first two or three millennia of that period. We have been unable to do this satisfactorily in the time available due to the huge mass of data it would be necessary to collect, examine and interpret: this is discussed further in §4. Instead, in §4 we set out our original intended methodology and our perforce modified method, and in §5, we present an initial set of proposals for a nonlexically based historical classification. These should be treated as a base for future work. We hope that even if parts of this attempt are judged unsuccessful, that will be progress, since it should then be possible to state the progress achieved and the reason(s) for failure.
2 MODELS
