The formation of French agent nouns (ANs) involves a large variety of morphological constructions, and particularly of suffixes. In this study, we focus on the semantic counterpart of agentive suffix diversity and investigate whether the morphological variety of ANs correlates with different agentive subtypes. We adopt a distributional semantics approach and combine manual, computational and statistical analyses applied to French ANs ending in -aire, -ant, -eur, -ien, -ier and -iste. Our methodology allows for a large-scale study of ANs and involves both top-down and bottom-up procedures. We first characterize agentive suffixes with respect to their morphosemantic and distributional properties, outlining their specificities and similarities. Then we automatically cluster ANs into distributionally relevant subsets and examine their properties. Based on quantitative analysis, our study provides a new perspective on agentive suffix rivalry in French that both confirms existing claims and sheds light on previously unseen phenomena.

French suffixations in -age, -ion and -ment are considered roughly equivalent, yet some differences have been pointed out regarding the semantics of the resulting nominalizations. In this study, we confirm the existence of a semantic distinction between them on the basis of a large scale distributional analysis. We show that the distinction is partially determined by the degree of technicality of the denoted action: -age nominals tend to be more technical than -ion ones. We examine this hypothesis through the statistical modeling of technicality. To this end, we propose a linguistic definition of technicality, which we implement using empirical, quantitative criteria estimated in corpora and lexical resources. We show to what extent the differences with respect to these criteria adequately approximate technicality. Our study indicates that this definition of technicality, while amendable, provides new perspectives for the characterization of action nouns.

Dans ce travail, nous examinons sur le plan distributionnel le sens de dérivés morphologiques, et plus précisément des noms d'agent déverbaux en -eur, -euse et -rice, et des noms d'action déverbaux en -age,ion et -ment. Nous utilisons une approche distributionnelle automatisée et un lexique dérivationnel. Nous proposons une représentation de l'information distributionnelle permettant d'examiner le sens prototypique des dérivés et l'instruction sémantique prototypique des suffixes. Nous montrons notamment que la différence entre les suffixes -eur, -euse et -rice ne relève pas seulement du genre et que les dérivés en -age, -ion et -ment présentent des profils spécifiques sur le plan distributionnel.Abstract. Contributions of distributional semantics for the semantic study of morphologically derived words. In this paper, we examine on a distributional level the meaning of morphologically derived words. We take a closer look at deverbal agent nouns formed with the French suffixes -eur, -euse and -rice, and nominalisations formed with the French suffixes -age, -ion and -ment. We combine a distributional approach and the use of a linguistic resource. We provide a representation of distributional information that allows us to examine the prototypical meaning of derivatives and the prototypical semantic instruction of suffixes. In particular we show that the distinction between the suffixes -eur, -euse and -rice is not limited to the gender. Moreover, we show that the suffixes -age, -ion and -ment show distributional specificities.

This article investigates the morphological diversity of agent nouns (ANs) in French. It addresses the questions of which nouns form a semantically coherent class of ANs, what their morphological properties are, and whether these properties correlate with agentive subtypes. To deal with these issues, a distributional semantics approach is adopted. The investigation is based on the distributional study of monosemous deverbal ANs ending in -eur, and on the examination of word similarities in the French Wikipedia corpus. It is shown that French ANs as a homogeneous distributional class display a large variety of morphological profiles. ANs can be affixed, converted, compound nouns, as well as opaque and morphologically simple nouns. Agentive affixes are diverse and correlated to the selection of bases from different lexical categories and semantic types. It appears that agent meaning in the nominal domain is not necessarily imported from the verbal domain, but can develop directly in the semantic structure of nouns. In addition, a distinction between functional, occasional and behavioral ANs, depending on whether they denote agents with an occupational status, agents in a particular event, or agents with a tendency to act in a certain way, is proven to be distributionally relevant. This distinction applies to all ANs, possibly in correlation with specificities as regards morphological type, base and affix selection. The study illustrates that with a careful processing of linguistic data, distributional semantics can help answering basic research questions and support fine-grained theoretical distinctions.

Affix rivalry is defined as the phenomenon of morphological competition where affixes and meaning are in a many-to-one relationship. Because of their poor semantic content, demonyms are perfect candidates for the investigation of selectional constraints in such a context. Indeed the morphological processes they originate from are characterized by their shared, straightfoward semantic relation, as they denote inhabitants linked to the toponym they derive from, which allows for the apparently simplified scrutinization of non-semantic properties. Investigations suggest a more nuanced and complex reality. The present study provides a quantitative and statistical investigation of the rivalry between French - ois, - ais, - ien and - éen suffixes. It notably relies on phonological and morphological features. Its contribution pertains to the use of statistical modeling to provide a quantitative description and to the integration of extralinguistic features on the nature of geographical proximity in a quantitative approach. The study shows that while the model cannot accurately predict the suffix of a given demonym based on these features, it still draws on the main tendencies underlying affix rivalry in the formation of French demonyms.