Poster A43, Thursday, August 16, 10:15 am – 12:00 pm, Room 2000AB

Gray and white matter correlates of indirect speech act comprehension in behavioral variant frontotemporal degeneration

Meghan Healey1, Murray Grossman1;1University of Pennsylvania

Introduction: Language comprehension in a real-world context extends beyond decoding the phonetic, semantic, and syntactic components of speech. In addition, a listener must integrate these linguistic elements with non-linguistic, social elements in order to fully appreciate speaker meaning. A quintessential example of this is indirect speech acts, which are speech acts in which the intended speaker meaning is not directly encoded in the semantic content of the utterance itself. While little research has examined indirect speech acts, they are ubiquitous in daily life: consider, for example, the exchange: “Do you want some cake for dessert? / I’m on a very strict diet right now.” In contrast to this indirect reply, consider the direct reply “I do not want any cake for dessert.” Behavioral variant frontotemporal degeneration (bvFTD) is a young-onset neurodegenerative disease characterized by executive and social impairment due to progressive gray matter atrophy in frontal and anterior temporal regions. These patients do not have aphasia or a segmental language disorder. Here, we test the ability of these non-aphasic bvFTD patients with focal dementia to comprehend indirect speech acts such as the one above. Methods: Patients with bvFTD and a group of matched control subjects completed a novel indirect speech task. Participants read a short dialogue between two interlocutors (see example above) and performed a simple judgment task (i.e. “Does the reply mean yes or no?”). Dialogues consisted of a polar question and reply, and conditions varied according to inferential demand (direct, indirect). Patients also underwent high-resolution structural T1 imaging and diffusion tensor imaging. To explore the mechanism underlying patient impairment, a targeted neuropsychological battery was collected, probing language, executive, and social measures. Results: Results indicate that control subjects perform at ceiling in both direct and indirect conditions, with no significant difference observed between the two conditions (Z=-0.577, p=0.564). Patients are significantly impaired relative to controls in the indirect condition (U=24.50, p=0.047) and perform significantly worse in the indirect condition than their own performance in the direct condition (Z=2.810, p=0.005). Importantly, patients are not impaired relative to controls in the direct condition (U=34.00, p=0.210), suggesting that segmental language ability is not responsible for the decrement in indirect performance. Furthermore, correlation analyses indicated that patient performance in the indirect condition was significantly associated with both executive (e.g. Backward Digit Span, Trailmaking Test B), and social measures (e.g. Social Norms Questionnaire), but not language measures (e.g. Multi-Lingual Naming Test, Semantic Word-Picture Test). Imaging results showed impaired indirect speech comprehension in bvFTD is related to atrophy in a fronto-parietal network including orbitofrontal cortex, medial prefrontal cortex, insula, and precuneus (all p<0.005, uncorrected). DTI results related indirect speech performance to reduced fractional anisotropy specifically in the corpus callosum, uncinate fasciculus, superior longitudinal fasciculus, and inferior fronto-occipital fasciculus (all p<0.005, uncorrected). Conclusion: Patients with bvFTD have deficits in indirect speech act comprehension. Neuroimaging data support the hypothesis that indirect speech act comprehension encompasses extralinguistic regions beyond traditional peri-Sylvian language regions. Importantly, these extralinguistic regions related to indirect speech are traditionally associated with social cognition.

Topic Area: Meaning: Discourse and Pragmatics

Back