In this experiment we set out to investigate how predictive processing may modulate alignment between interlocutors in a dyadic setting (Pickering & Garrod, 2013). The experiment presents a novel interactional task where participants are involved in a partially controlled association game where speaker A names a picture and speaker B responds with a semantically related word. Importantly, the predictability for the upcoming object is manipulated. Data has been collected from 20 dyads, and the results show a prediction effect with a mean difference of 400 ms between predictable and non-predictable conditions. Crucially, this prediction effect was not only present in speaker A who had to name the predictable or unpredictable object, but also for the interlocutor. To our knowledge, this is the first study to demonstrate speaker-listener prediction effects in a dyadic interaction. This will be further tested in a dual-EEG setting to explore this question at the neural level.

A behavioural and two EEG hyper-scanning experiments are presented which investigate how predictive processing modulates the way interlocutors align behaviourally and at the level of the brain (Hasson, 2012; Pickering & Garrod, 2007). In the experiments interlocutors engaged in dyadic interactions performing a semi-controlled semantic association game and where the p predictability of the semantic associations was manipulated. The behavioural results showed that both interlocutors were around 400 ms faster in the predictable versus non-predictable conditions The results of the two EEG studies aim at demonstrating (1) whether we observe brain-to-brain synchronisation between the interlocutors at the level of word semantics, and (2) whether prediction enhances this synchronisation. To our knowledge, this is the first study to directly demonstrate prediction effects in an interaction.

The semantic representation of abstract words is a topic of discussion within the embodied cognition framework. Existing theories propose the involvement of emotional, linguistic, and social experiences in abstract word representation. Focusing on ‘socialness' — a variable with limited empirical evidence — our study explores whether abstract words are associated with richer social experiences. Two semantic categorization tasks (explicit and implicit) with socialness priming and a lexical decision task with socialness priming were conducted to examine the effect of socialness on abstract words recognition. Additionally, we run similar experiments with affective priming to examine the effect of valence on abstract words recognition. Our results indicate that only valence facilitates the recognition of abstract words and only in the explicit task. Conclusively, we find no evidence supporting a non-strategic effect of socialness and valence on abstract words recognition, thus challenging existing theories on the grounding of abstract words in social information and emotion.