This paper examines the present situation facing a linguist studying South-Bauchi West Chadic languages (SBW) spoken in Nigeria. Among those 27 or so idioms, some are dead, others are quite robust, while the majority are severely endangered. The question is: what kind of polylectal grammar can we propose for those languages? In order to address this question, the linguist has to take into account various factors: the socio-linguistic situation of SBW languages; the economic and political situation of Nigeria and Europe ; the theoretical issues debated in the linguistic community. These factors will be briefly sketched before coming to the conclusion that, as was indicated by the organizers of the conference in the title of this paper, in a situation where development issues (production of a standard language, teaching material, etc.) would require human and financial resources which, if they existed, could be put to a better usage, we have to concentrate on the documentation of SBW languages before they disappear. For that, while being aware that the concept of a language equated to a people and a territory, is a fiction belied by the current and historical realities of population migrations and language contacts that bring about multilingualism and linguistic continuum, the linguist will still favour the classical tools of structural linguistics that derive from that conception, and aim at finding the structures underlying the language uses of SBW speakers, however partial, local and overlapping they may be. The paper will finish with a synopsis of the classical comparative grammar of SBW languages that has been in the process of elaboration for the past 10 years or so. This grammar will take the form of a core grammar, which coud be regarded as an abstract, non-historical proto-SBW working as a stable term of comparison for the other languages. Against this proto-SBW, the actual present-day languages or sub-group of languages will be contrasted, and trends of evolution will be analysed. The best documented languages are as of now are: Zaar, Geji, Chaari, Dott, Sigidi. More data will be collected on Zakshi, Kal, Polci and Pelu. This data collection process should be over by the end of 2011. Then the process of analysing and writing the grammar proper will start.