Following the Conversation: Impacts of Set-Shifting and Topic-Shifting in Healthy Adults and Individuals with Traumatic Brain Injury

Abstract

Difficulty in conversational discourse abilities, marked by issues with processing topic structure, are a common characteristic of cognitive-communication disorders in individuals with traumatic brain injury (TBI). Despite progress in foundational word and sentence-level skills through speech therapy interventions, problems with conversational discourse tend to remain persistent. This persistence suggests a gap in understanding how non-linguistic cognitive processes influence conversation. In this set of experiments, I test whether cognitive mechanisms related to set-shifting contribute to processing topic shifts in conversation. The purpose of Experiment 1 was to determine if topic switches show characteristic behavioral signatures of set-shifting that emerge in non-linguistic tasks: longer response onset latencies and decreased information content efficiency. The results from the first experiment showed no differences between responses to new topics and the same topics for these measures; however, it was unclear whether these results would reflect responses to shifts in naturalistic conversation or if they simply were products of the experimental design. In Experiment 2, I examined the impacts of the topic switches in naturalistic conversation on language production in healthy adults and TBI patients. This includes measuring productivity, semantic complexity, semantic complexity, syntactic complexity, and fluency for responses to new topics and the same topics. I found that topic shifts elicited costs in terms of the number of words per utterance, verbs per utterance, revisions/rephrasing, and filled and unfilled pauses per syllable for both groups. These findings demonstrate that there are costs associated with switching topics that mirror non-linguistic shift costs and may suggest they arise from similar mechanisms.