Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

N400 Splitting Reveals Unexpectedly Early Effects of Semantics on ERPs

Poster Session C - Sunday, March 8, 2026, 5:00 – 7:00 pm PDT, Fairview/Kitsilano Ballrooms

Pinning He^1,2 (pinning.he@mail.mcgill.ca), J. Bruno Debruille^1,2; ¹McGill University, ²Douglas Mental Health University Institute

For many researchers, the N400 ERP appears too late to index initial semantic processing. Nevertheless, no purely semantic effects have been consistently reported on ERPs prior to the N400. However, this absence may result from averaging across participants who have effects in opposite directions on earlier ERPs. The present study aimed to separate participants based on amplitude differences between their N400 to ambiguous words and their N400 to unambiguous words. This was done first in the data of an experiment run in Paris (N = 94). There, participants read short stories that ended either with semantically univocal and coherent endings, equivocal, or incoherent words. These endings were identical across the three conditions, while the short story context they ended was changed across participants. Participants were median-split into two groups based on the results of the subtraction of the ERPs to coherent from the ERPs to equivocals in the 300-500 ms time-window. Very early effects of condition were observed. Additionally, a significant interaction of group x condition was found in a very early time window ( i.e., 40-120 ms). To verify whether this interaction was replicable, we analyzed the data of the same experiment run in Montreal on different participants (N = 91). A group x condition interaction was found again. We are thus reporting these results to open a discussion as to the nature of the semantic processes that could be at stake at such an early latency.

Topic Area: LANGUAGE: Semantic