Abstract
This article addresses the issue of historical interpretation of documents produced in semiotic economies that do not correspond to the dominant ones. It tries to show how attention to linguistic ideologies may assist historical interpretation. Rather than truth or factuality, voice should be central to this interpretive project. Voice can be identified by a close ethnographic inspection of textual practices, placed against the ackground of local hierarchies of signs and sign values from which individual signs derive meaning and function. An analysis of a handwritten document called “History of Zaire” (Congo) and written by Tshibumba shows how forms of genre work offer spaces for Tshibumba to define himself as a historian, that is, as the producer of prestigious types of ordered and organized knowledge. The act of writing in generically regimented ways is what allows Tshibumba to call his document an instance of historiography.