The first originality of this project is that it proposes to focus primarily on those languages which
can be rightfully suspected to be isolates, according to a majority consensus among specialists (cf. list
in Task 1). We thus intend to collect and collate all published and unpublished linguistic data on those
language isolates, by consulting published sources, old and recent, and by contacting scholars having
done first-hand research on them. We also intend to proceed ourselves with collecting first-hand
material on some of those languages, in our areas of expertise (e.g. on the Irimba language, spoken by
a group of Pygmies in Southern Gabon). Our project then, aims explicitly at achieving a new
classification of African languages, which could supersede Greenberg's, now in use for the last 40
years.
The second original aspect is the application of new methodologies to these corpora. For instance the
method of virtual reconstructions (i.e. the reconstruction of the shape of the source item for lexical
items putatively derived from a proto-language) can be a powerful tool for the identification of loan-
words and the stratification of the lexical material found in the language(s) under examination - this, it
must be said, applies first and foremost to Bantu languages and secondarily to other well-
reconstructed groups, like Eastern and Southern Nilotic, Eastern Cushitic, and a few others. Virtual
reconstructions have been developed and applied by members of our laboratory working on cultural
vocabularies of Gabonese languages (with the pioneering studies of Mouguiama-Daouda on
ichthyonyms (1995), and, jointly with Hombert, on mammal names).
A third novel angle will be the taking into consideration of ecological issues in the dynamics of
language evolution. In conformity with Nettle's approach, we expect to find maximum diversity of
languages in the wet equatorial areas. The GARP software (Genetic Algorithm for Rule-Set
Prediction) originally developed for determining the ecological niches of plant and animal species, can
be applied to archaeology (cf. Banks et al. 2008), by correlating geographical coordinates of known
archaeological sites with their estimated ages. Their paleo-environment can thus be determined with
considerable precision, which should have momentous consequences for our understanding of past
population movements in sub-Saharan Africa. In particular a precise knowledge of ancient forest-
savanna ecotones will help us better to assess the dynamics of hunting-gathering groups - and later on
of agriculturalists also, since it has been shown that most cultivating groups in Africa also made
extensive use of wild products at their disposal, and this at the forest-savanna ecotone in Central
Africa as well as on the slopes of the East African highlands.
Consequently, the scientific contribution made by this project will be a more precise reconstruction
of the history of African peoples, and a new classification of their languages in and around the
equatorial forest through the use of recently developed tools for historical linguistics and language
ecology.
The possible obstacles that we have to face are obvious : the segmentation of the field between
various disciplines (history, archaeology, anthropology, population genetics and of course linguistics).
However, we shall benefit from the very serious progress achieved over the last five years by the
Eurocores Program "Origins of Man, Language and Languages" whose very aim was to enable
practitioners of the various sciences to find a common ground and learn to communicate with one another. As former participants in the program, we feel assured that considerable progress has indeed
be made in this direction - witness, as far as Bantu Africa is concerned, the most interesting results of
the "Gabon languages and genes" project, conducted at DDL under the direction of Professor Lolke
van der Veen (mentioned above under 1.1). |