Nowadays, European and other big languages have multimillion annotated text corpora, and the role of corpus methods in language studies constantly grows. African linguistics lags far behind in this respect. However, several mid- and small-size text corpora for Mande languages have been already build: Bambara Reference Corpus (> 11 million words), Maninka Reference Corpus (about 3.5 million words), Eastern Dan corpus (almost 500 thousand words), Mwan Corpus (about 50 thousand words); further corpora are planned. There is already a number of important linguistic studies for Mande languages performed on the base of these corpora.
Language corpora can be also used by non-linguists. The language corpora have certain time depth (e.g., about 50 years for the Bambara corpus, and some texts are even older), and they include texts of different genres. For example, a sociologist or politologist can perform a corpus study for the key terms of his/her interest, and to discover the evolution of the notions in the language in the time.
In my communication, some case studies will be presented to exemplify this approach.