Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Theory of mind predicts conversational success in early right hemisphere stroke recovery
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Tatiana Schnur1, Andrea Suazo1, Margaret Blake2; 1McGovern Medical School, University of Texas Health Houston, 2University of Houston
Successful conversation requires speakers to track what they know, what their partner knows, and how these differing perspectives shape dialogue. While central to theories of pragmatic language, direct evidence linking theory of mind (ToM) to conversational success after right hemisphere damage (RHD) remains limited. RHD provides a powerful test case because many patients experience consequential communication impairments extending beyond basic language abilities, creating an opportunity to identify socio-cognitive mechanisms that support interactive dialogue. We tested whether ToM explains individual differences in conversational success in 33 adults with acute/subacute RHD (median 5 days post-stroke). Sixteen age- and education-matched controls provided a normative benchmark for task performance. We measured conversational success using a collaborative picture-difference task where participants and coordinators identified discrepancies between similar scenes, requiring participants to coordinate privileged visual knowledge with a partner’s partially overlapping knowledge. A nonverbal false belief task minimized language demands while assessing two ToM components: spontaneous inference of another person’s perspective under lower self-perspective conflict (low conflict) and explicit belief reasoning while managing conflict between self- and other-perspective (high self/other conflict). Baseline cognitive tasks assessed visual inattention, inhibitory control, and working memory. Using hierarchical linear mixed-effects models within the RHD group, we tested whether ToM predicted conversational success beyond demographic, stroke-timing, baseline cognitive, and conversation-partner effects. Adding ToM significantly improved model fit over the base model (Δχ²(2) = 8.6, p = .014; ΔR² = .05). Critically, explicitly inferring others’ perspectives under high self/other conflict predicted conversation success (β = 0.51, p = .02), demonstrating that participants who were better able to reason about a partner’s knowledge while managing interference from their own privileged perspective identified more picture differences during conversation. No other predictors were significant (p’s > .10). The lower-conflict ToM condition showed a weaker association (β = 0.37, p = .097), suggesting that the main difficulty in this collaborative task was not merely representing another person’s belief, but comparing that belief with one’s own knowledge to guide communication. The full model explained substantial variance in conversational success (fixed effects: 36%; fixed plus random effects: 50%). These findings provide the first empirical evidence in unilateral RHD that conversational success depends on managing conflict between self- and other-perspectives, rather than just inferring interlocutor knowledge. This supports models where cooperative dialogue requires speakers to maintain distinct representations of self and other, compute the mismatch between them, and use that mismatch to decide what information needs to be communicated. By testing in the acute/sub-acute phase, we better link disrupted ToM- mechanisms to functional communication before longer-term reorganization or compensatory strategies dominate performance. Clinically, these results advocate for socio-cognitive screening and ToM-targeted interventions for post-stroke communication impairments.
Topic Areas: Disorders: Acquired, Meaning: Discourse and Pragmatics