Poster A42, Thursday, August 16, 10:15 am – 12:00 pm, Room 2000AB

The exceptional role of the first person: Evidence from natural story processing

Matthias Schlesewsky1, Ingmar Brilmayer2, Alexandra Werner3, Beatrice Primus2, Ina Bornkessel-Schlesewsky1;1Centre for Cognitive and Systems Neuroscience, School of Psychology, Social Work and Social Policy, University of South Australia, Adelaide, Australia, 2Department of German Language and Literature I, University of Cologne, Cologne, Germany, 3Department of English and Linguistics, Johannes Gutenberg-University, Mainz, Germany

The use of natural stories (e.g. narratives) provides us with a new way of studying linguistic information processing in the human brain. It allows us to replicate results from previous, controlled studies, but also opens up possibilities for investigating dependencies that span larger time units. This is, for example, relevant for research about the maximal size of temporal receptive windows. It can also increase our understanding about the role of predictions in language processing. In narratives, there are “global” predictions in the sense that protagonists will recur several times throughout the story, while temporal (“local”) predictions about this recurrence are relatively imprecise (i.e. it is difficult to predict precisely when a protagonist will be rementioned). Here, we present initial EEG observations on the processing of narrative dependencies. Participants listened to a German audio book version of The Little Prince by Antoine de Saint-Exupery (recording by Will Quadflieg, chapters 1–17, excluding chapters 5,6 and 14). The narrative includes passages written from the perspective of the Little Prince, but also dialogical passages from the perspective of the Little Prince’s interlocutors, as well as a third person narrator. Twenty-five, monolingual native speakers of German (14 female; mean age 24.4, range 20–29) with normal hearing participated in this experiment. For the current analysis we focused on two main aspects: a. Given the exceptional role of the first person (“I”) in comparison to second and third person, we investigated differences in the processing of first, second and third person pronouns. The exceptionality of the first person is motivated from a typological as well as processing perspective and may reflect a more general distinction between self vs other; b. We calculated the so called referential distance (RD) for every pronoun under examination. RD is calculated by counting the number of sentences that lie between the current mention of a discourse referent and its last mention in prior discourse. RD ranges from 0 to 20, while 20 is also assigned to referents without prior mention. Only referentially unambiguous pronouns encoded as grammatical subject were evaluated. In total, the recording contained 79 first person, 35 second person and 95 third person singular pronouns. Results indeed revealed a difference between pronouns: between 150 and 250 milliseconds, first person pronouns showed a strong positivity as opposed to third person pronouns (P300), which in turn elicited more positive-going ERPs than second person pronouns (1 > 3 > 2). In addition, ERPs following first-person pronouns were nearly unaffected by referential distance. For second-and third-person pronouns, by contrast, P3 amplitude decreased with increasing referential distance. Given the previously demonstrated sensitivity of the P300 for self-relevant behaviour, our results provide the first evidence that, even in narratives, the first person serve as an attentional cue for self relevance. In addition, the independence from RD for the first, but not the second and third person, could indicate that this type of attentional cue is prediction independent and constitutes a default in information processing in general.

Topic Area: Meaning: Discourse and Pragmatics