In everyday life, we are often involved in social situations that lead us to share ideas or to get involved in cooperative tasks. In such situations, we use language to build upon previously shared information and develop newly shared knowledge with our addressee. For example, after establishing with someone that an abstract shape “looks like a cat”, the same abstract shape can later be referred to as “the cat”. Here, mutual knowledge of “the cat” has been explicitly established. Though similar examples happen in real life conversations, it is also very common that speakers refer to things that they already knew before the interaction. In such cases, choices of referential expressions can also be adapted based on inferences of the addressee’s prior knowledge of the referent.
Inferring another person’s mental states, including their knowledge, requires a cognitive ability called theory of mind. In a previous study, we showed that speakers indeed adapt how they refer to different movie characters they know depending on how likely it is that their addressee also knows them. Importantly, participants with better theory of mind abilities showed a greater adjustment to the addresses’ inferred knowledge, establishing a significant link between theory of mind and choices of referential expressions during verbal interactions.
The addressee in our last study was always a confederate who answered as if she had prototypical knowledge of the movie characters, ‘knowing’ all the well-known characters and none of the little-known characters. Information about people’s prototypical knowledge of the characters and information provided through feedback during the task thus always converged.
This project aims to improve on our innovative task to further examine the relative importance of these two different sources of information on theory of mind inferences about an addressee’s knowledge. If the adjustment observed for the presentation of well-known and little-known movie characters in our previous study was tailored to the needs of the specific person with whom our participants interacted, then this adjustment should be reduced when facing an atypical addressee who shows an equal knowledge of all characters. These adjustments could take the form of making sure to give enough information for proper identification, of refraining from saying unnecessary information, or both. In addition we aim to corroborate and further explore the relationship between theory of mind abilities and adjustment of references for the addressee. Moreover, given that speakers often provide over-informative references, we will examine how such referential choices affect naïve listeners’ appraisal of the speaker’s knowledge and their confidence in these judgments.
Contributions and advantages: Our previous studies support the feasibility and importance of the proposed project. This project will for instance have important implications for the current theoretical debates about the collaborative or egocentric nature of language production. Our novel approach will also allow studying the impact of theory of mind inferences during participative interactions and will provide important new knowledge on the relationship between theory of mind inferences and different verbal production. Our innovative methods should also foster new research in that field. Our team is involved in translating new knowledge and methods for clinical research and the present project could thus have critical clinical implications through the development of new evaluation or treatment approaches for disorders linked with theory of mind or language production deficits.