Lwenn Bussière-Caraes

We discuss the nature of retraction, what its felicity conditions and effects on a conversation are, how it compares to other speech acts, and how it relates to social power dynamics. On our account, retractions are proposals to update the context a particular way and, as such, can be rejected. Our formal account models the effect of a retraction on the context to restore the common ground to a prior state, computable from the conversational record. Our account also explains the status of speech acts that follow the target of retraction. Retractions are cognitively costly on our account, but, we argue, this is the right result.

In a certain picture of cooperative conversation, ‘silence gives assent’. However, in adversarial contexts, structured by power dynamics, silence may be a powerful expression of dissent. To reconcile these opposite interpretations, I propose an analysis of silence as the expression of a default attitude. Given pragmatic cues, participants infer the cooperativeness of conversational settings. Depending on cooperativeness, they assign a default attitude (of assent, of suspension of judgment, of dissent) to other participants, that they take intentional silence to express. This analysis takes the main effect of speech acts to be proposals to update the conversational Common Ground, and highlights the necessity of assent in conversational updates.

A speaker can express the same thought, true under the same conditions, while using different expressions and grammatical constructions. According to Frege, these are differences in colourings. Colourings may convey additional contents; in that, they resemble Gricean conventional implicatures. Sander (2019) argues that Gricean implicatures do not subsume the category of colourings, as some colourings do not communicate their content. I show that this argument relies on a notion of communication focused on the speaker's intentions. But a notion of communicative intentions where a speaker is responsible for the intentions her audience ascribes her possible. Under this notion, since so-called non-communicative colourings trigger specific inferences, a speaker who uses them communicates these inferences. Therefore, I vindicate the communicative role of colourings with content.