QuBisM - Questions, Bias, Multimodality: Negotiating Meaning in Interaction

The most straightforward reason for asking a question is to gather information, as in Do you have a bike? Yet questions can also request confirmation of a belief, as in Don't you have a bike? The latter question is a so-called biased question because the speaker expresses a bias toward, i.e., is leaning toward, the possibility that the interlocutor owns a bike whilst asking whether this is true. The meaning layers expressed by different biased questions have been the issue of intense research in the past two decades. Yet, their exact conditions of use in various communicative situations, the interplay of diverse linguistic cues signalling different types of biases, and the issue of how we acquire as children the intricate form-meaning relations of such multi-layered functionality are poorly understood. In the proposed project, we seek to fill this gap in order to arrive at a more nuanced account of biased questions. Biased questions are an ideal way to study how people negotiate complex layers of meaning through word order, intonation and gestures, which will further our understanding of communication in general. While previous research has focused on the word order and meaning of 'classic' biased questions, which are essentially negative yes-no questions (e.g., Don't you have a bike? Do you not have a bike?), the goal of our project is to unravel the nuanced meanings such questions can express in comparison to constructions which also express a bias but have not been treated under this term, like tag questions (You have a bike, don't you?) and formulaic suggestions (Why don't you get a bike?). We aim to tease apart the roles of sentence structure, meaning, the immediate context, sentence melody and gesture (head, hands, body movements and facial gesture) in all these biased questions (broadly defined) in conversation. With new tools and frameworks for multimodal research emerging, we have now the means to study these linguistic cues and their interaction in detail. Indeed, there is initial evidence that speakers use different gestural means in different question types in English, and crosslinguistic research shows that gesture is used to mark many different meaning aspects in conversation, highlighting the need for gesture research when studying subtle meaning differences. We will conduct our research on English and German. These two languages use similar morphosyntactic means to mark different types of bias, but there is evidence that the interplay of prosody and morphosyntax differs between them. We will also investigate how biased questions can be learned in the early years (children aged 4 to 6) in our cross-linguistic comparison. Complementing our empirical and analytical contributions on the expression and learnability of biased questions, we will contribute a theoretical account of the division of labour across linguistic modules in identifying different types of bias. This will feed our understanding of how different kinds of linguistic means contribute to interpretation in human language more generally. If these theories are to reflect cognitive reality, the assumed structures must be learnable and their development traceable in the course of first language acquisition. These theoretical considerations make biased questions an ideal springboard for interdisciplinary engagement with researchers, for instance in psychology for investigating the development of Theory of Mind, language pathology for exploiting developmental disorders manifesting in social and linguistic behaviour, or computer scientists who seek to improve machine-to-human-interaction.