About time we make sense: Distinct neural processes engaged during temporal sequencing and coherence building in discourse
Fairbrother, Wonja M.
- In the real world, causes always come before effects, while in communication, events can be described in either this canonical temporal order using causal connectors such as "and so", or in non-canonical order using connectors like "because". Using event-related potentials (ERPs), we determined whether the canonical sequencing of events influences the establishment of causal coherence, or vice ver... read moresa, during online discourse processing. Two-clause cause/effect sentences were created, in which we fully crossed temporal sequencing of events with discourse coherence, yielding four experimental conditions (example set: Fred was hungry [and so/*because] he ate...; Fred ate [*and so/because] he was hungry..."). 32 participants read these sentences, presented word-by-word (450ms, ISI: 100ms), and made acceptability judgments at the end of each sentence. Participants' working memory spans were assessed with an Automated Reading Span task. At anterior electrode sites, ERPs to critical words ("ate/hungry") in clauses appearing in non-canonical sequence evoked a larger sustained negativity starting between 400-500ms than in clauses appearing in canonical temporal sequence. At centro-parietal sites, ERPs to critical words in incoherent clauses evoked a larger negativity between 300-500ms (an N400 effect) and a larger positivity between 700-900ms (a P600 effect) than in coherent clauses. A main effect of working memory span between 200-300ms after the critical word and sentence-final word did not interact with either coherence or canonicity. Together, these results suggest that during discourse comprehension, establishing the temporal sequencing of events and establishing their causal coherence are driven by distinct neural mechanisms and are not influenced by individual differences in working memory span.read less