Anouch Bourmayan, Pascal Ludwig, and Morgan Moyer are organizing a “Valence in Language” workshop at Sorbonne Université on June 13, 2025. Invited speakers are Diana Mazzarella (Université de Neuchâtel), Joshua Knobe (Yale University), and Nicole Gotzner (University of Osnabrück).
Abstracts for talks can be submitted before February 28, 2025. The call reads:
It is widely agreed that Frege’s On Sense and Reference set the foundations for contemporary philosophy of language, as well as formal semantics. It should not come as a surprise, then, that affective meaning, which tracks speakers’ subjective feelings and attitudes, has been almost completely dismissed in both disciplines as an unsuitable object of study. Indeed, Frege’s misgivings about the relevance of psychological aspects is one of the hallmarks of his approach to logic and formal language. A way of rephrasing Frege’s worries would be to say that the affective information associated with a word is necessarily subjective, and, as such, irrelevant to the study of meaning that aims at objective and hence shareable aspects of meaning.
This view has remained largely unchallenged, and the dismissal of the relevance of affective information sank even deeper down as this referentialist semantics approach to natural language reified in the mid 20th century with the melding of technical and philosophical advances from Tarski, Davidson, Montague, and Lewis, and then standardized with the formalism in Heim and Kratzer (1998).
However, the last 20 years have seen a flourishing of interest in such phenomena, including recent proposals by, e.g. McCready (2020) on expressives, Cepollaro (2020) or Hess (2021) on slurs, and Jeshion (2021) for a taxonomy of pejorative meaning. Nonetheless, the mainstream still views these phenomena as generally irrelevant to the study of meaning proper, in part since they are thought to manifest in a minimal or exceptional part of the lexicon.
In parallel, the field of cognitive psychology has extensively explored the significance of valence in language. Following an early idea from Wundt (1907), Zajonc (1980, 2000) has defended the general hypothesis that affective responses may precede conceptual recognition, that is, may be evoked with minimal stimulus input and virtually no cognitive processing. Regarding language more specifically, as early as 1957 Osgood introduced the semantic differential technique which allowed him to define the affective connotation of words – not only specific classes of words but “plain vanilla” words – along three underlying dimensions, the first of which was valence. Other models of semantic differentials were subsequently developed, including those by Mehrabian and Russell (1974), Bradley and Lang (1999) and Warriner et al. (2013). Overall, all the studies confirmed that valence is the most significant dimension of the three parameters, being the most stable and the most informative one. Further, with advances in psycho- and neurolinguistic methodologies in the last decades, the Affective Primacy hypothesis found support at the level of linguistic content, comparing affective to descriptive dimensions of meaning (see, among others, Bargh et al. 1989, Kousta et al. 2009, Gaillard et al. 2006 or Ponz et al. 2014).
In this workshop, we would like to examine the idea that valence has a greater role in language than has been generally acknowledged. Indeed, a word’s valence might be an important aspect of the meaning of many more words than those that are recognized as “expressives”. That is, expressivity could be a broad and ubiquitous phenomenon rather than a feature specific to only certain terms.
