Questions in TED talks : a pragmatic approach to multimodal communication
Résumé
This paper examines questions and their gestural correlates in spoken monologic discourse, as observed in a corpus of 18 English TED talks (6 hours and 4 minutes, 1110 questions across 17 speakers). TED (Technology, Entertainment, and Design) talks are reported to be a hybrid and emerging genre ((Caliendo & Compagnone 2014); (Ludewig 2017)) that both relates to and differs from academic discourse and political speech. They are intended to be short, informational, and entertaining. TED Talks are web-mediated and widely shared to mass audiences. They are a digital medium (Belcher 2023) which belongs to the New Oratory as defined by Rossette-Crake (2019; 2022)). They are characterised by an asymmetrical relationship between a single speaker and a two-tier audience – both in person and remote. Although backchannel and gestural responses (such as hand-raising) may be invited, the common ground cannot be explicitly updated by an addressee. This contextual constraint has consequences on the grounding process (Clark & Brennan 1991) and on the gestures associated with questions (Shattuck-Hufnagel & Ren 2018). Questions are assumed to be produced by an ignorant speaker seeking information (Farkas, 2022) and are generally analysed as turn-taking triggers (Stivers et al., 2009). However, when questions cannot contribute to dialogue moves due to contextual constraints, their function is less clear. Interactive alignment is weaker than in conversation, coordination and synchronization with the audience being hindered by the non-interactional context (Pickering & Garrod 2004). In addition, questions are not homogeneous in terms of communicative function. The speaker may speak in their own name or reflect a generic vox populi (Celle 2009). The goal of this paper is to define the role of direct questions depending on the nature of the interlocutive relation and the gestures they are associated with. In order to carry out this study, all questions were segmented and annotated for syntactic, discursive and gestural parameters (hand gestures; head and eyebrow movement) using ELAN (Sloetjes & Wittenburg 2008). Questions appear to be mostly used for discourse-structuring and emphatic purposes, as strategies to attract the audience's attention and simulate interaction ((Cardo & Celle 2024; Cardo & Celle forthcoming)). They are an essential tool to steer the discourse to a speaker-controlled knowledge state. We hypothesize that the various functions of questions can be better captured by investigating their interlocutive properties, the nature of their answers (no answer, immediate or delayed answer by the speaker, answer by the audience or a specific addressee) and their gestural correlates. More specifically, we expect the emphatic and interactive functions of questions to be reflected in the gestures they co-occur with (Lelandais forthcoming). Our preliminary results indicate that verbless questions are associated with large hand gestures while declarative tag questions tend to be associated with two-handed gestures. As for wh-questions, they are significantly marked with hand beats. It remains to be determined if the presence of an answer affects these gestural correlates.