Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Cerebellar Contribution to Semantic Conflict Resolution
Poster Session A, Wednesday, September 30, 11:00 am - 1:00 pm, Wangari Maathai
Amanda Chirino Pérez1, Amanda Lebel1, Sigurd Alnes1, Fangfei Li2, Elicia Chau1, Richard Ivry1; 1UC Berkeley, 2The Hong Kong Polytechnic University
Introduction: During discourse, comprehension is facilitated by predictive mechanisms, allowing listeners to anticipate the speaker’s utterance and intent. Successful communication also requires flexibility on the part of the listener when their predictions are violated. When an expected word is replaced by an unexpected but plausible word, the system constructs an alternative meaning that competes with the predicted meaning. Resolution of the resulting conflict entails attentional processes that modulate the activation of the competing representations. We hypothesized that the cerebellum coordinates the dynamics of this attentional process, enabling the inhibition of competitors and the representational shift needed to resolve the semantic conflict. Methods: We tested patients with cerebellar degeneration (CD) and matched controls (n=60/group) in three experiments using independent cohorts. Experiment 1 utilized a semantic decision task where participants judged if a sentence was meaningful. We compared sentences ending in highly predictable words (Predictable), weakly constrained words (Low), unexpected but plausible words (Switching), and semantically anomalous words (Meaningless). Experiment 2 employed a lexical decision task evaluating word priming across different linguistic contexts. In Experiment 3, we recorded EEG during the semantic decision task, analyzing linguistic markers from event-related potentials (N400, post-N400 components) and time-frequency signals (theta power, 4–8 Hz). Results: Behaviorally, the CD group produced error rates comparable to controls. Reaction times were slower in all conditions. On the semantic decision task, the CD group showed a larger cost on Switching trials and a reduced cost on Meaningless trials. On the lexical decision task, the CD group showed faster RTs to verify the lexical status of the predicted word but, unlike controls, did not show negative priming to words related to the predictable word. These results suggest an over-reliance/activation of the predictable word and/or lack of semantic flexibility in the CD group. For example, the absence of negative priming indicates that the patients were unable to sharpen the attentional spotlight on the predictable word. The ERP analyses revealed a reduced N400 amplitude in the CD group to the final word on Switching trials, indicating facilitated lexical access to the unexpected word given the absence of inhibition from the predictable word. In Meaningless and Low conditions, N400 amplitude was preserved but their onset latency was delayed, suggesting late lexical access. Crucially, post-N400 positive ERPs indexing semantic integration and context reanalysis, were similar across conditions in the CD and Control groups, confirming preserved higher-level integrative processes. Time-frequency analyses showed pronounced midfrontal theta for all conditions relative to the Predictive condition, a marker of frontal engagement when the final word violated a semantic expectancy. Compared to the Controls, this response lasted much longer in the CD group on Switching and Low trials, indicating extended time required for conflict resolution. Conclusion: Core semantic processing and sentence-level integration remain intact in cerebellar degeneration, including the ability to use the current context to generate semantic predictions. However, the dynamics of semantic processing under conditions requiring cognitive flexibility is disrupted in this population, causing greater difficulty in making the representational shift required for comprehension in ambiguous contexts.
Topic Areas: Control, Selection, and Executive Processes, Disorders: Acquired