Les desinences personnelles en Indo-Europeen

This dissertation, written in French, is a study of the Indo-European personal endings. It consists of two main sections. The first consists of a complete collection of all data available, covering all endings, for all persons and all numbers, found in the oldest-attested Indo-European languages, together with an attempt to account for the internal and mutual influences within each language or language group. In addition, in the case of endings that have remained obscure and unexplained so far, an original analysis has been suggested whenever possible. An effort has also been made to determine the elaboration, within each branch, of any in-between stages leading to the systems that are attested. For each language or language grouping, this section provides an account of every ending, from its supposed original shape in the parent language, to its actual attestations, either as the regular outcome of regular sound change, or as resulting from remodellings themselves caused by the internal influence of other endings. The second section deals with the reconstruction of the system of the parent Proto-Indo-European language. Moreover, that system, itself, requires further analysis, and so a set of internal reconstructions is here proposed as well, which point to a clear-cut, fully regular and symmetric system, where each function was always covered by a single marker. An attempt has further been made to account for how, from such a state, the system evolved to the one (less clear and asymmetric) that is usually reconstructed and that must have prevailed at the very last stages of the common language. By their very nature, such internal reconstructions (or, as the case may be, at least some of their details) are more or less speculative although each one seems to have at least some structural motivation, which is explained.