Poster Slam Session C
Friday, August 17, 10:15 – 10:30 am, Room 2000C, Chair: Mairéad MacSweeney

Syntactic and thematic mechanisms of subject-verb integration in aphasia and typical sentence comprehension

Jennifer Mack1, Cynthia K. Thompson2;1UMass-Amherst, 2Northwestern University

Introduction. In both neurotypical adults and adults with agrammatic aphasia, subject-verb integration is relatively easy in unergative sentences, in which the subject receives the thematic role of Agent (i.e., a volitional actor) and originates in a pre-verbal syntactic position. Greater neurocognitive resources are required for unaccusative sentences, in which the subject is a Theme (i.e., an undergoer) and originates in a post-verbal position [1]. For example, in neurotypical listeners, sentence subjects are reactivated more quickly in unergative than unaccusative sentences [2]. Two visual-world eye-tracking experiments tested competing accounts of this finding. Experiment 1 compared unergative to unaccusative structures (e.g., The child … jumped/collapsed …, where “…” indicates additional linguistic material), whereas Experiment 2 compared processing of “mixed” structures (i.e., with Theme arguments but pre-verbal traces) to unaccusative sentences (e.g., The table … creaked/collapsed …). The syntactic hypothesis [2,3] predicts earlier reactivation for sentences with pre-verbal vs. post-verbal traces in both experiments. The thematic hypothesis proposes that listeners make online thematic predictions (i.e., animate subject – Agent; [4]), and thus predicts earlier reactivation when thematic predictions are confirmed (unergative vs. unaccusative sentences in Experiment 1) but no difference when thematic structure is matched (mixed vs. unaccusative sentences in Experiment 2). For listeners with aphasia, the syntactic hypothesis [3,5] predicts greater processing delays for sentences with post-verbal vs. pre-verbal traces, whereas the thematic hypothesis [4,6] predicts no differences across conditions, due to impaired thematic prediction processes. Methods. The participants were neurotypical young (n=20) and older adults (n=15) and adults with stroke-induced agrammatic aphasia (n=8). Participants listened to sentences, viewed visual arrays containing an image of the sentence subject and three distractors, and answered occasional comprehension questions. Mixed-effects logistic regression models quantified the time course of reactivation of the sentence subject. Results. In the first 500 ms after verb onset, neurotypical adults showed faster reactivation in unergatives vs. unaccusatives in Experiment 1, but no differences across conditions in Experiment 2. In the same time window, listeners with aphasia showed no effects of condition in either experiment. Conclusion. The results supported the thematic hypothesis, suggesting that thematic predictions guide subject-verb integration and impaired predictive processes may contribute to sentence comprehension deficits in aphasia. The results also suggest that left inferior frontal regions, often damaged in agrammatism, support linguistic prediction (cf. [6,7]). References. [1] Thompson, C. K., & Meltzer-Asscher, A. (2014). In A. Bachrach, I. Roy, & L. Stockall (Eds.), Structuring the argument: Multidisciplinary research on verb argument structure (pp. 141-168). Amsterdam: John Benjamins. [2] Koring, L., Mak, P., & Reuland, E. (2012). Cognition, 123(3), 361-379. [3] Burkhardt, P., Piñango, M. M., & Wong, K. (2003). Brain and Language, 86(1), 9-22. [4] Meyer, A. M., Mack, J. E., & Thompson, C. K. (2012). Journal of Neurolinguistics, 25(1), 31-43. [5] Love, T., Swinney, D., Walenski, M., & Zurif, E. (2008). Brain and Language, 107(3), 203-219. [6] Mack, J. E., Ji, W., & Thompson, C. K. (2013). Journal of Neurolinguistics, 26(6), 619-636. [7] Nozari, N., Mirman, D., & Thompson-Schill, S. L. (2016). Brain and Language, 157-158, 1-13.

Topic Area: Language Disorders

Poster C40

Back to Poster Slam Sessions