Slide Slam B14
P600 and task effects
Veena D. Dwivedi1, Pratik Nath1, Brent Dryczewycz1, Aaron Ayub1; 1Brock University
We conducted two experiments to investigate the influence of task demands on sentence processing using event-related potential (ERP) methods. Stimuli from Osterhout and Holcomb (1992) were used to examine effects on the P600 component, a marker of syntactic anomaly. Participants read sentences (critical words are underlined) such as (i) The broker persuaded _*to_ conceal the transaction _was_ sent to jail vs. (ii) The broker planned _ to_ conceal the transaction *_was_ sent to jail. In Experiment 1, 25 participants read these sentences with yes/no comprehension questions (e.g., Did the broker plan something?), in contrast to Osterhout & Holcomb’s use of grammaticality acceptability tasks. We expected to replicate findings from the original work, where P600 effects were found at ‘to’ in (i) vs. (ii) and at ‘was’ in (ii) vs. (i). While P600 effects at ‘to’ did not replicate, the P600 effect downstream at ‘was’ did. In Experiment 2, 21 participants read these sentences, where no questions were asked at critical trials. Reduced P600 effects were expected due to lack of attention at critical trials. We did not find any evidence of P600 effects at ‘to’ or ‘was’ at all. Instead, a nearly significant long-lasting positivity effect was found at the final word ‘jail’ in (ii) vs. (i) suggesting that participants wait to integrate for meaning. Both experiments did find a significant long-lasting negativity at 'conceal', suggesting that participants were sensitive to deletion associated with the reduced relative clause. Results are discussed in terms of noisy channel hypotheses of sentence processing.