Contenu du sommaire : Quoi de neuf en morphologie? sous la direction de Bernard Fradin et Françoise Kerleroux

Though inflection and derivation have been distinguished from each other ever since there has been work on grammar, we ask whether this distinction is problematized anew in contemporaneous morphological approaches, characterized as lexeme-based, and not morpheme-based. We shall present the "Split Morphology Hypothesis" as it is worked out by Anderson, and "the Morphology by itself" hypothesis formulated by Aronoff, looking for the consequences on the definition of morphology itself.

Our analysis of french nouns with collective reference which are constructed on the basis of a noun with the suffixes -ade, -aie, -aille and -uге such as colonnade, chênaie, chevelure and grenaille or prêtraille has a double end: (i) to account of the exact part of the suffixation in the fact that a noun may have a collective reference; (ii) to progress in the analysis of collective throughout the part-whole relations, and, in a very specific manner, throughout two of these relations, the portion- mass relation (which connects grain to grenaille, for instance), and a relation which has never been studied for itself but has only been evocated in a constrastive way connected with other meronymic relations, the member-collection relation (which connects colonne with colonnade and cheveu with chevelure for instance). It appears that one can predict the type of whole named by a derived noun according to the suffix which has been used to construct it. So it will be proved that the lexical units do have a lexical specific meaning which is independent of the context; so it will be shown how to express plurality in the inflexional and derivational fields.

Suffixed nouns with -ET have up to fithteen different meanings. These meanings are distributed between a Referent pole and a Speaker pole. They are structured as a radial category, the central meaning of which is "smallness", on the model of an Idealized Cognitive Model. This model correlates all the morphological processes yielding the derived -ET nouns. Only the suffixed nouns grouped together within the Referent pole do involve the idea of diminishing and are formed by classical lexeme formation rules. For the others, the link between the base and the derived noun on the one hand, and the semantic import tied to the suffix on the other are accounted for by two different mechanisms: the suffixation says that the derived noun's referent denotes something familiar to the speaker. As for the link between the base and the derived noun, it is established by very general semantic mechanisms which also operate in nominal compounds.

Spanish hypocoristics are known for the consonantal strenghtenings they display (cf. Francisco → Pancho, Sergio → Checo, etc.). These strengthenings have been analyzed (Piñeros 2000) as a manifestation of the emergence of the unmarked triggered by the different propensities of consonants to function as onsets. We will analyze in the very same terms two classes of hypocoristics which have until now resisted analysis: those formed via consonantal reduplication (e.g. Francisco → Quico) and those formed via a process of syncope (e.g. Francisco → Paco). In both cases, the speaker is able to extract from the string of segments a consonant most apt to play the role of onset than that which would have been obtained via apocope or apheresis.

This article presents a detailed analysis of suppletion phenomena in French conjugation. The analysis rests on the postulation of a network of dependency relations between the various stems of a verb lexeme. We then discuss the relevance on inflectional classes to the description of French conjugation; we conclude that there is no compelling case in favor of inflectional classes of French verbs.