Cognitive Control in Bilingual Aphasia

Cognitive control in bilingual aphasia INTRODUCTION • Recent research has identified the complex interplay between language and cognitive control in normal bilingual individuals (Costa, 2008; Luk, 2010; Paap, 2013). • Evidence shows that bilingual individuals can effectively inhibit non-linguistic and linguistic information. • In bilingual aphasia we see language control impairments (e.g. pathological codeswitching, asymmetrical translation ability); however, it is not clear whether bilingual patients have a domain specific or domain general control impairment. • No study yet has systematically examined cognitive control in bilingual aphasia to determine whether deficits in language inhibition are specific to the language domain or are indicative of a more general cognitive deficit. Domain Specific vs. Domain General Cognitive Control in Bilingual Aphasia PARTICIPANTS RESULTS Specific Aim 1: Flanker Task Controls and patients exhibit the congruency effect: longer RT on incongruent condition compared to the congruent condition. 30 neurologically healthy Spanish-English bilingual adults (NHBA) (21 women, M=48; SD=14). 9 Spanish-English bilingual adults with aphasia (BAA) (4 women, M=54, SD=14). Patient P1 MPO 124 Age 63.4 P2 96 63 P3 26 43 P4 197 38.9 P5 32 67.3 P6 30 76.4 P7 16 35 P8 31 55.6 P9 12 48.4 Response Time • BAA completed the Pyramids and Palm Trees Test (PPT), Bilingual Aphasia Test (BAT) in Spanish and English, The Boston Naming Test (BNT). PPT Patients Mean (SD) BNTEnglish BNTSpanish BATBATBATComprehension Comprehension Semantic English Spanish English 85 (14.0) 21 (28.1) 21 (22.1) 58 (28.1) BATSemantic Spanish BAT-Word Recognition: English into Spanish BAT-Word Recognition: Spanish into English BATTranslation: English into Spanish BATTranslation: Spanish into English 68 (22.7) 45 (21.5) 53 (18.8) 69 (23.3) 71 (26.8) 11 (22.1) 19 (21.5) • NHBA and BAA completed the Language Use Questionnaire (LUQ): language ability rating (LAR), lifetime exposure, current exposure, hours report, education history and confidence, family proficiency. Established language dominance for each participant from LUQ Language Ability Rating (LAR) self-report: LAR English *minus* LAR Spanish. METHODS Flanker Task Specific Aim 2: Linguistic Task Accuracy • NHBA are more accurate on Tr, Un, and UnTr conditions, demonstrating semantic interference. • The BAA model is not significant although trends show higher accuracy for semantically related conditions. Response Times OBJECTIVES Determine whether deficits in language control are specific to the language domain or are indicative of a more general cognitive deficit. Linguistic Task Instructions: Are the two words are related or unrelated? Hypotheses 1. Neurologically healthy bilingual adults will complete both tasks successfully (intact cognitive control). 2. Bilingual adults with aphasia will perform more poorly on the linguistic task compared to the non-linguistic task (evidence for domain specific cognitive control) OR bilingual adults with aphasia will perform poorly on both tasks (evidence for domain general cognitive control). 3. Language proficiency will effect response time outcomes on the linguistic task. RESEARCH POSTER PRESENTATION DESIGN © 2012 www.PosterPresentations.com NHBA and BAA exhibit the congruency effect, thus demonstrating non-linguistic cognitive control. Linguistic Task NHBA: Accuracy & Response Time NHBA are more accurate on Tr, Un, and UnTr conditions indicating that NHBA exhibit semantic interference on accuracy. However, NHBA are fastest on congruent conditions (Tr, S, STr) indicating an RT processing benefit from semantic information. English targets: F(4, 44) = .661, p > .05 Spanish targets: F(4, 44) = .899 p > .05 Specific Aim 3: Effect of Language Proficiency Accuracy & Response Time BAA do not show an interfering or facilitory effect from processing semantic information, suggesting inefficient linguistic processing. Although individual RT analysis does reveal that the majority of BAA do show faster RTs for semantically related conditions. Effect of Language Proficiency • NHBA: RT is effected by language dominance on congruent and incongruent non-translation conditions. A similar trend is observed on the STr condition. These results indicate that NHBA exhibit more efficient processing when going into their dominant language. • BAA: Language dominance influences RT on the semantic nontranslation condition only. BAA are efficient with processing semantically related words in their dominant language only. IMPLICATIONS Results suggest that in BAA linguistic control mechanisms are dissociable from non-linguistic control. Domain Specific Deficits in BAA NHBA: RT is related to language dominance on non-translation (congruent and incongruent) conditions. RTs on direct Translation are faster than on Unrelated Translation. The Semantic Translation trend is that RTs are shorter when translating into the dominant language. BAA : Language dominance effects RT on the Semantic non-translation condition. Specific Aim 2: Examine the nature of language control in a linguistic task that requires active inhibition of the non-target language. • Do neurologically healthy bilingual adults and bilingual adults with aphasia will exhibit similar patterns on a semantic interference task that requires language control? Specific Aim 3: Examine the effect of language proficiency on language processing. • How does language proficiency for Spanish-English neurologically healthy bilingual adults and Spanish-English bilingual adults with aphasia effect speed of processing on translation and nontranslation conditions that vary by semantic relationship? English targets: F(4, 149) = 25.352, p < .001 Spanish targets: F(4, 149) = 19.436, p < .001 Flanker Task BAA: Instructions: Is the red arrow pointing left or right? Specific Aim 1: Examine the nature of cognitive control requiring inhibition in a non-linguistic task. • Do Spanish-English neurologically healthy bilingual adults and Spanish-English bilingual adults with aphasia exhibit similar patterns on the Flanker task (Erickson and Erickson, 1974)? CONCLUSIONS Future Directions Test additional non-linguistic tasks that require more demanding cognitive control compared to the Flanker Task. Sample stimuli: • Stimuli consist of word pairs with various relationships (e.g. direct translation, semantic, unrelated) and language directionality (e.g. English-English, Spanish-Spanish, EnglishSpanish and Spanish to English). • All stimuli were controlled for frequency and cognates. Test two additional linguistic tasks that require varying degrees of lexical access, i.e. word identification and word generation. Identifying different aspects of lexical access and deficits in our patients will allow us to tease apart levels of semantic processing.