Cognitive Control, Semantic Processing, and Semantic Ambiguity Resolution in Individuals with Aphasia

Abstract

Event Abstract Back to Event Cognitive Control, Semantic Processing, and Semantic Ambiguity Resolution in Individuals with Aphasia Wiltrud Fassbinder1*, Rebecca Hunting Pompon2, Mohammed Aldhoayan3, Hyun Seung Kim1, Hyunsoo Yoo4, Johannes Hüsing5, Kevin Dalziel6, Jeremy Mancini7 and Malcolm R. McNeil1 1 University of Pittsburgh, Communication Sciences and Disorders, United States 2 University of Delaware, Communication Science and Disorders, United States 3 University of Pittsburgh, Health Information Management, United States 4 The University of Texas at El Paso, Rehabilitation Sciences, United States 5 Universitätsklinikum Heidelberg, Koordinierungszentrum für Klinische Studien (KKS), Germany 6 VA Northern California Health Care System, Speech Language Pathology, United States 7 Jefferson University Hospitals, Thomas Jefferson University, United States INTRODUCTION Several imaging and lesion-based studies suggest that cognitive control plays an important role in semantic ambiguities resolution[1]. However, two recent behavioral studies indicate that such interference resolution may be more related to verbal semantic abilities than to cognitive control[2,3], with better semantic abilities resulting in more ambiguity interference[2]. While deficits in cognitive control are well documented in individuals with aphasia (IWA)[4,5,6], the effects of such deficits for ambiguity resolution have not been investigated in conjunction with semantic abilities. This study investigates how in IWA, cognitive control and lexical-semantic abilities contribute to comprehension efficiency and success in ambiguity resolution. It was predicted that for IWA, deficits in cognitive control result in increased ambiguity interference and decreased accuracy, whereas deficits in lexical-semantic processing result in decreased ambiguity interference, on a sentence comprehension task (SCT) requiring lexical ambiguity resolution. PROCEDURES Ninety-five IWA completed all study tasks. Cognitive control was indexed as the RT ratio of incongruent/neutral items in a picture-word interference task (PWI ratio). Lexical semantic processing was measured as accuracy in a category judgment task (CJ accuracy). On the SCT, participants read sentences with a disambiguating verb and an ambiguous object, followed by a probe word related to alternative meaning of the ambiguous object (He drank the port today – docks)[7]. In the control condition, ambiguous objects were replaced with unambiguous words. All timing measures were modeled in linear mixed effect models, and accuracy was fitted using logistic mixed regression models[8]. Subjects and items were entered as random effects (including intercepts and slopes[9]), and ambiguity (ambiguous/control), PWI ratio, and CJ accuracy as fixed effects. RESULTS RT IWA with higher PWI ratios responded faster in the SCT than those with lower ratios (95% CI [-14.4 – -2.8]). Main effects for ambiguity (95% CI [-.8 – -.04]) and lexical-semantic processing (95% CI [-1.4 – .2]) were modulated by an interaction (95% CI [.2 – 1]): Ambiguity interference was only evident in IWA with higher CJ accuracy (see Figure 1). Accuracy Ambiguous items were less accurate than unambiguous items (95% CI [.2– 1]). DISCUSSION Cognitive Control: The lack of relationship between PWI interference and SCT ambiguity for RT and accuracy suggests that different aspects of cognitive control are required in the two tasks6: In the PWI task, inhibition is task-driven and mainly proactive, whereas in the SCT inhibition is mainly reactive. However, higher PWI interference is associated with faster judgments, suggesting that it reflects an underlying cognitive factor relevant to processing efficiency, possibly processing speed. Further analyses of available data will examine this possibility. Lexical-semantic processing: Faster lexical semantic processing was associated with faster RTs and more ambiguity interference, consistent with the hypothesis that better semantic processing results in faster activation of alternative meanings, causing more interference[4]. However, differences in CJ accuracy did not affect SCT accuracy, suggesting that this interference was successfully resolved. For individuals with poorer lexical-semantic processing, RTs were slower, and there was less/no ambiguity interference, suggesting that competing word meanings were not sufficiently re-activated to interfere with the decision time or comprehension success. Figure 1