Poster D45, Friday, August 17, 4:45 – 6:30 pm, Room 2000AB

The neural basis of shared discourse: fMRI evidence on the relation between speakers’ and listeners’ brain activity when processing language in different states of ambiguity

Karin Heidlmayr1,2, Kirsten Weber1,2, Atsuko Takashima1,2, Peter Hagoort1,2;1Max Planck Institute for Psycholinguistics, Wundtlaan 1, 6525XD Nijmegen, The Netherlands, 2Donders Institute for Brain, Cognition and Behaviour, Radboud University, 6500 HB Nijmegen, The Netherlands

In natural language use, information is exchanged between speakers and listeners seldom in the form of isolated words or sentences, but usually in the frame of larger discourse contexts, such as narratives or expository texts. To create and understand meaning in a discourse, links between successive utterances within the text, but also with world-knowledge beyond the discourse have to be made. The goal of the present study was to identify the neural bases of meaning construction between speakers and listeners when processing naturalistic texts. The successful exchange of conceptual information between speakers and listeners requires a balance between new information (uncertainty, entropy) and previously known information (redundancy). However, in suboptimal conditions, uncertainty can be exceedingly high, e.g. if within-text information is ambiguous and background information insufficient, leading to the listener’s lack of understanding of the utterance. Thus, in the present fMRI pseudo-hyperscanning study, we aimed at investigating how the amount of given contextual information influences the understanding of ambiguous discourse. Critically, to assess the neural activity related to the sharing of meaning between interlocutors, i.e. the conceptual pact, the relationship of the neural activation associated with meaning creation between the speakers and listeners was assessed. An fMRI experiment using a variant of the ambiguous text paradigm (Dooling & Lachman, 1971; Bransford & Johnson, 1972) was designed, i.e. conceptually ambiguous expository texts were presented with preceding contextual information that in some cases did and in others did not facilitate the extraction of a coherent meaning. Ten speakers produced ambiguous texts preceded by a highly informative title in the scanner and 18 listeners subsequently listened to these texts, preceded by either a highly informative, intermediately informative (highly/intermediately informative title conditions) or no title at all (non-informative condition). Shared conceptual processing was expected to be reflected by neural coupling that varied with comprehension-relevant contextual information, while linguistic and sensory information was equal across conditions. BOLD activation analyses as well as inter-subject correlation (ISC) analyses (Hasson et al., 2004) between the speakers’ and the listeners’ BOLD time courses were carried out. Besides a strong involvement of the left-lateralized language network (Hagoort, 2017) in all three conditions, preliminary BOLD data revealed activation of the left dorso-medial prefrontal cortex in the highly informative title condition, indicating its role in establishing text coherence. Moreover, ISC analyses showed an alignment between speakers’ and listeners’ BOLD time courses in medial frontal and parietal regions when ambiguous texts were preceded by a highly or intermediately informative title. In the absence of a title, no such alignment was found. These regions overlap with parts of the default-mode network (Fox & Raichle, 2007) and have previously been suggested to play a role in integrating information over larger time scales when processing discourse (Ferstl et al., 2008; Simony et al., 2016). The present data indicate that the construction of coherent meaning and the representation of situation models in discourse rely on the involvement of medial parietal and frontal regions, which show alignment in the BOLD activation across interlocutors who share discourse information.

Topic Area: Meaning: Discourse and Pragmatics