The relation between perception and production in social meaning is often taken to be transparent, with social meaning associations learned from observations of language use. However, recent work has suggested that this relation is often more complex than previously thought. Here, we present new data comparing the social meaning of realized variable liaison in spoken French, couched within the framework of the pragmatic sociology of critique. We recall data from a recent matched guise experiment showing that listeners associate the realization of liaison with meanings like “professionalism”, specifically in social situations where efficacy and expertise are at issue. Basing ourselves on this finding, we use a production task, presenting these same social situations to amateur and professional actors. We find that our participants do not exploit the social meaning potential of variable liaison, producing liaison at lexically-determined rates on a par with previous corpus studies. We discuss this discrepancy between perception and production, which suggests that the link between the two is dependent on the linguistic variable under investigation.

This article focuses on the problem of the categorization of the adverb sans doute as an epistemico-modal marker or an evidential marker. Starting from a critical appraisal of an analysis proposed by Bourmayan et Ashino, in which sans doute is considered fundamentally epistemico-modal, three hypotheses are presented, which depart from theirs. First, sans doute has as its main lexical element an inferential evidential value: it signals that the content qualified by the adverb is the result of a defeasible type of inference from clues/premises. Second, due to the defeasibility of the inference, the statement containing the adverb takes on, invariably, a pragmatic value of non-certainty, not lexically coded, however. Sans doute is thus not an epistemico-modal marker by itself in contemporary French. Third, the adverb combines these values with an epistemic value of what we call “posture of certainty”. Different from epistemic modality, which refers to an evaluation in terms of certainty, and/or to the epistemic cognitive state linked to that, posture of certainty refers to what can be called the tone or epistemic behavior of confidence the speaker shows. Sans doute therefore indicates without contradiction, a defeasible and, thus, uncertain hypothesis, presented with a tone of full confidence.

Ces dernières décennies, certains travaux ont cherché à montrer que le plus-que-parfait tend à développer des emplois de passé perfectif dépourvus de toute notion d’antériorité. Le but du présent article est de reconsidérer ces analyses, en accordant davantage d’importance à l’anaphoricité de ce temps verbal. Après avoir rappelé les principales propriétés du plus-que-parfait, l’auteur note que les exemples discutés dans ces travaux concernent essentiellement deux contextes d’emploi. Plusieurs de ces exemples sont ici repris et réanalysés. L’auteur en conclut que ces travaux sous-estiment le rôle de l’implicite dans le fonctionnement anaphorique du plus-que-parfait, en exigeant inutilement la présence d’une expression temporelle susceptible d’interpréter l’anaphorique (antécédent). Il en résulte que l’hypothèse parfois avancée d’une aoristisation de ce temps verbal, outre son coût théorique, s’avère inappropriée.

In this article I begin by presenting some examples from Early-Modern French of how, using a large multi-genre database (Frantext), it is possible to track the spread of change and how it embeds across different genres (the genres in Frantext ranging from correspondence and first-person travel narratives to essays and poetry). These case studies allow us to explore a number of questions, including which changes seem to diffuse “from above” and which rather come “from below”; and which textual sources best reflect “authentic” usage or indeed are closest to reflecting spoken usage. They also raise questions as to whether it is possible to create a continuum of genres with those that are more “progressive” or are early adopters of change at one end, and those which are more resistant to change at the other. I then discuss some of the challenges surrounding this work, including the theorization (or lack of theorization) of text types and genres in various corpora and the privileging of other factors when they are elaborated. Consequently, there is currently a lack of comparability between different corpora in the way they categorize and calibrate different text types and genres.