ABSTRACT
As we have already noted, there are fundamental problems about trying to find out about the spoken word of the past. The recordings and corpora on which analyses of contemporary spoken French depend have to give way to the fundamentally unsatisfactory data of the written word, which can never be more than a partial reflection of direct speech. In the case of French, the problems are compounded by the tradition of standardization which is especially characteristic of the seventeenth century and by the nature of the orthographic system which falls lamentably short of the phonemic ideal of having each sound unambiguously represented by only one symbol. In order to find out about the speech of the Grand Siecle, scholars have to piece together clues offered by diverse texts, including the comments of grammarians and lexicographers, and the representation of direct speech in popular literature, texts in patois and comic theatre, satire and burlesque. In addition, comparative data, allowing reconstruction in a way similar to the reconstruction of Vulgar Latin forms, may be of value; comparison may be made for instance between 'standard' French and the French taken overseas by colonizers and settlers in the seventeenth century. If, for example, 'French' creoles that are widely separated geographically share a common feature which is no longer a feature of French spoken in France, one may hypothesize that it was perhaps present in the spoken and popular usages which constitute the source of those creoles.
