The Ndebele Language Corpus: A Review of Some Factors Influencing the Content of the Corpus

The Ndebele language corpus described here is that compiled by the ALLEX Project (now ALRI) at the University of Zimbabwe. It is intended to reflect as much as possible the Ndebele language as spoken in Zimbabwe. The Ndebele language corpus was built in order to provide much-needed material for the study of the Ndebele language with a special focus on dictionarymaking and research. Like most corpora, the Ndebele language corpus may in future be used for other purposes not thought of at the time of its inception. It has been designed to meet generally acceptable standards so that it can be adaptable to various possible uses by various researchers. The article wants to outline the building process of the Ndebele language corpus with special emphasis on the challenges that faced compilers, and possible solutions. It is assumed that some of these challenges might not be peculiar to Ndebele alone but could also affect related African languages in a more or less similar situation. The main focus of the discussion will be the composition of the Ndebele language corpus, i.e. the type of texts that constitute the corpus. The corpus is composed of published texts, unpublished texts and oral material gathered from Ndebele-speaking districts of Zimbabwe. It will be argued that the use of the corpus and its reliability for research depends among other factors on its contents. It will also be shown that the contents of a corpus depend on a number of factors, some of which include sociolinguistic, political and economic considerations. These considerations have implications on both the content and quality of published and oral texts that constitute the Ndebele language corpus. Keywords: corpus, oral materials, code-mixing, code-switching, mother- tongue, Ndebel