This study investigates evidential meanings, referring to the information sources available to the speaker (i.e., direct witnessing, report, or inference), expressed through syntactic and lexical strategies in French. Across three experimental studies, we examined French subordinate structures, including complement, relative, pseudo-relative, and infinitive ECM clauses, that encompass perception or reporting verbs. We used witness rating, information source identification, and discourse completion tasks administered to a total of 221 French speakers. The results show that (i) infinitive ECM clauses are unambiguously associated with direct witnessing; (ii) pseudo-relative and relative clauses with voir are, though less strongly than infinitive, associated with direct witnessing; (iii) although still being sometimes used in reference to direct witnessing, complement clauses with voir are primarily suitable in inferential contexts based on resultant states; and finally (iv) complement clauses of dire are associated with reported sources. Our results show that indirect evidentiality is marked by complement clauses only, whereas direct evidentiality distributes over infinitive ECM, pseudo-relative and relative clauses with no significant difference between the two types of relatives. We conclude that indirect evidentiality in French is syntactically associated with one type of subordinate structure, whereas direct evidentiality is not.

In this article, we examine nominal modal auxiliary phrases (NMAPs) in French. More specifically, we study NMAPs where the nominal base consists of the nouns obligation (‘obligation’), nécessité (‘necessity’) and probabilité (‘probability’). At first glance, these appear to correspond to the deontic interpretation of “obligation”, the alethic interpretation of “necessity”, and the epistemic interpretation of “high probability” associated with the modal verb devoir. We demonstrate that, from a syntactic perspective, these NMAPs can be introduced by the verbs avoir (‘to have’), être (‘to be’), se trouver (‘to find oneself’), se voir (‘to see oneself/be/’), and /se/ mettre (‘to put/set /oneself/’), with the exception of probability constructions with the noun probabilité, which can only be introduced by the verb avoir. From a semantic perspective, we show that the noun in “deontic” NMAPs is not exclusively obligation, but often nécessité as well. Similarly, while the noun in “alethic” NMAPs is most frequently nécessité, the noun obligation is not entirely excluded. Regarding NMAPs involving probabilité, the modal interpretation is almost always “alethic possibility” and not the “epistemic” interpretation of “high probability.”

Some lexicographers, language observers, grammarians, and linguists have sporadically noted the rise in usage of the sequence “dans le cadre de” (which can be translated as “in the context of/as part of”) in contemporary French writing. However, no in-depth study has been devoted to this complex preposition, even though it presents several remarkable features. The aim of this study is to provide insights into how this expression, which derives from the prepositional phrase “dans le cadre de NP”, which refers to a space enclosed by a physical boundary, has become so spectacularly successful in written usage, especially from the 1950s and particularly in the press. In order to do this, we first check its status as a complex preposition by applying syntactic, semantic, and statistical criteria. We then explore its network of meanings, starting with those of the noun “cadre” (Engl. ‘frame’), which forms its lexical core. Exploring textual databases enables us to study, synchronically and diachronically, how these meanings are distributed in texts, thereby highlighting the relevance of the criteria of genre, type, and topicality. Finally, we show that incorporating external historical considerations into the analysis provides a better understanding of the remarkable rise in the use of this complex preposition in the contemporary press.

Dans cet article, nous étudions l’alternance entre deux formes concurrentes au sein du paradigme du verbe « aller », où les formes ALLER coexistent avec les formes ÊTRE (issues du paradigme du verbe « être »). Grâce à un travail approfondi et minutieux sur des données de corpus oraux, associé à des analyses inférentielles robustes combinant des forêts aléatoires et la modélisation bayésienne, nous montrons que cette concurrence est influencée par plusieurs variables linguistiques telles que le type de sujet, la personne, la négation, le type de complément, mais aussi – et surtout – par des variables concernant le locuteur et l’interaction, telles que l’âge, le genre, la région d’origine ou le nombre de participants présents. Allant plus loin que les travaux précédents sur cette alternance, nos analyses montrent des interactions significatives entre ces deux ensembles de facteurs. Nous défendons la nécessité de prendre en compte les règles prescriptives lors de l’étude de la concurrence morphologique, mais aussi leur intégration différenciée par les locuteurs en fonction de leur profil social.

This study investigates the polysemy of the French verb sentir, traditionally studied among perception verbs, from a usage-based perspective through a behavioural profile analysis conducted on a heterogeneous written and spoken corpus. The aim is to assess whether the annotated senses of sentir form a structured distributional structure rather than a simple inventory of discrete lexical meanings. The analysis relies on hierarchical clustering in order to reconstruct distributional proximities among the verb’s different uses. The results show that the senses of sentir are not randomly distributed, but instead display a coherent distributional structure. The most robust finding concerns an opposition between uses involving property or emanation and uses centered on the experiencer’s perception. The analysis also identifies a relatively stable network core in which the sense “to experience an internal bodily state” emerges as the best candidate for a distributional reference point. The study further proposes a topological interpretation distinguishing an internal experiential zone, a modality-specific sensory zone, and a referential periphery, interpreted through several semantic organizational mechanisms, including metaphor, metonymy, specialization, subjectification, and schematization.

Restrictive relatives in French are often framed as problematic because vernacular usage departs from the symmetrical form–function mapping upheld by prescriptive grammars. Where the standard system privileges a diversified inventory of syntactically specified relativizers, spontaneous speech recurrently converges on the generalized subordinator que, often interpreted as a simplification or error. Drawing on a corpus of sociolinguistic interviews with speakers of Acadian French in Nova Scotia, this study examines the conditioning of relativization in a spoken minority variety using a variationist framework incorporating mixed-effects regression models. The analysis reveals a structured probabilistic system dominated by three high-frequency markers (qui, que/qu’, dont), whose distribution reflects both internal and external dimensions of relativization. These dimensions are examined through structural, processing-related, stylistic, and social factors, including syntactic function, dependency length, antecedent properties, matrix construction type, and sociolinguistic characteristics of the speakers. In non-subject relatives, the que ∼ dont alternation is structured primarily by oblique function and further conditioned by dependency length and style. Subject relatives remain near-categorical in qui, though rare non-canonical que/qu’ forms are consistently licensed by resumptives and weakly specified antecedents.