The aim of this study is to contribute to the methodology for determining whether a given characteristic of a language is a product of language contact or of language-internal grammaticalization. We have taken as a test problem a formal structure that is relatively rare across languages but that occurs in a few geographically proximate languages belonging to different families. The presence of a typologically rare phenomenon in neighboring but unrelated languages raises the question of whether the structure may be a product of cross-linguistic contact.The structures that we examine involve the split coding of person and number of the subject, in which a pronoun preceding the verb codes person only. Plurality of the subject is coded by a suffix to the verb, usually the same suffix for all persons. In some languages the split coding of person and number operates for all persons, while in others the split coding is limited to some persons only. This structure has been observed in several languages spoken in a small area of Northern Cameroon. Three of these languages, Gidar, Giziga, and Mofu-Gudur, belong to the Central branch of the Chadic family, while two other languages, Mundang and Tupuri, belong to the Adamawa branch of the Niger-Congo family. Outside of this geographical area, this structure has been observed in Egyptian, some Cushitic languages, and in some languages of North America.Since every linguistic phenomenon must have been grammaticalized in some language at some point, we must consider first whether there are language-internal prerequisites for such grammaticalization. For each language of the study, we show that the split coding of person and number may represent a product of language-internal development. The presence of the phenomenon in a language that does not have language-internal prerequisites can then be safely considered to be a product of language contact.
If the inline PDF is not rendering correctly, you can download the PDF file here.