Puisqu'il n'existe pas de mots qui ne soient à personne
(Bakhtin 1979 in Todorov 1981: 83)
Le discours, c'est-à-dire le langage dans sa totalité concrète et vivante.
(Bakhtin 1963 in Todorov 1981: 44)

European travel accounts are widely used as sources to write the history of societies that did not themselves produce a large amount of textual documentation. On the Coast of Guinea, and more particularly on the area formerly known as Gold Coast (approximately the littoral of modern Ghana), many of the written documents providing historians with information on coastal societies prior to nineteenth century were produced by Europeans travelers. Most of these individuals were merchants, craftsmen, pastors or soldiers who had settled several years in the numerous forts and trading posts erected along the seashore to protect and enhance the trade of chartered companies. Others were seamen and merchants who only spent a few weeks at a stretch plying the African Coast aboard men-of-war or trading ships, exchanging manufactured goods for gold, slaves, and ivory. Professional writers, who had not traveled to Africa, but had obtained data from various written sources or from travelers, also composed some of the accounts. Therefore, these documents show an extraordinary diversity in form and content, which historians and archeologists need to investigate before using them as primary sources for reconstructing the past.

In the first part of the present paper, which focuses on a specific genre, the seventeenth-century travel accounts, I recall the historiography of the recent critique of these sources. I also point out that despite decisive methodological breakthroughs, some heuristic dimensions or attributes of these texts are yet to be recognized and assessed. I then re-examine these sources from the dual perspective of historic and linguistic anthropological methods and introduce an exploratory approach centered on the Bakhtinian approach to discourse.