Domain General vs. Domain Specific Language

Domain general vs. domain specific language and cognitive processing in bilingual aphasia INTRODUCTION Bilinguals routinely manage (i.e., control) two languages, but how this skill effects cognitive control is under debate: association between language control (LC) and cognitive control (CC) vs. a dissociation between these control mechanisms (e.g., Calabria et al., 2011; Prior & • an Gollan, 2011). Current study: • Bilingual adults with aphasia (BAA) offer a unique opportunity to explore the relationship between LC and CC because they are often reported to present with LC deficits. Few studies have investigated LC and CC mechanisms in this population (Dash & Kar, 2014; Gray & Kiran, in press; Green et al., 2010; Verreyt et al., 2013). RESULTS Analyses: 1. Congruency effect: indicates that control mechanisms are functioning in that domain. Formula: accuracy: (congruent – incongruent); RT: (incongruent – congruent). • Non-linguistic tasks: two-way ANOVAs were performed to evaluate the effect of condition (congruent and incongruent) x target (color and shape or red and black) for percent accuracy and zRT for NHBA and BAA. • Linguistic tasks: two-way ANCOVAs were performed to evaluate the effect of condition (congruent and incongruent) x target (English and Spanish) for percent accuracy and RT for NHBA and BAA and a measure of language experience was the covariate. OBJECTIVES 1. Determine whether deficits in language inhibition are specific to the language domain or are indicative of a more general cognitive deficit. Similarity between language control and cognitive control: Resistance to Distractor Interference Bilingual Language Processing (type of cognitive control) Both languages are active. To access the target Inhibition of non-target stimuli that is simultaneously language, the non-target language must be inhibited. presented with the target stimuli. NHBA: F(1, 38) = 1.47, p = .23; BAA: F(1, 24) = 4.29, p < .05 NHBA F(1, 37) = 1.72, p = .20; BAA F(1, 23) = .85, p = .37) NHBA: F(1, 38) = 33.15, p < .001; BAA: F(1, 24) = 22.15, p < .001 NHBA F(1, 37) = 33.62, p < .001; BAA F(1, 23) = 2.05, p = .17 NHBA: F(1, 38) = 34.82, p < .001; BAA: (F(1, 24) = 8.12, p < .01 NHBA F(1, 37) = .05, p = .82; BAA F(1, 23) = .07, p = .79) NHBA: F(1, 38) = 119.54, p < .001; BAA F(1, 24) = 15.98, p < .001) NHBA F(1, 37) = 19.68, p < .001; BAA F(1, 23) = 3.54, p = .07) 2. Determine if there is an effect of task complexity. PARTICIPANTS • 20 Spanish-English neurologically healthy bilingual adults (NHBA), age: M = 51; SD= 13. • 13 Spanish-English bilingual adults with aphasia (BAA), age: M = 48; SD= 12. purple = Spanish dominant *NHBA and BAA were matched on age and education and all filled out the Language Use Questionnaire (Kiran et al., 2010). green = English dominant 2.Conflict Ratio: indicates the magnitude of conflict. Formula: accuracy: (congruent– incongruent)/incongruent RT : (incongruent – congruent)/congruent (Green et al. (2010). BAAHigh (higher dx scores): BAA3, BAA7, BAA8, BAA12, BAA13 BAALow (lower dx scores): BAA1, BAA2, BAA4, BAA5, BAA6, BAA9, BAA10, BAA11 Diagnostic testing in English and Spanish for all BAA. BAA1 Standardized Tests BAT Receptive Language Expressive Language BAT Part C Translation Translation Boston Naming Test Expressive Language PALPA (25) /EPLA (26) BAA2 BAA3 BAA4 BAA5 BAA6 BAA7 BAA8 BAA9 BAA10 BAA11 BAA12 BAA13 E S E S E S E S E S E S E S E S E S E S E S E S E S 36% 80% into S 20% 14% 0% into E 30% 67% 80% into S 75% 88% 100% into E 85% 75% 100% into S 45% 74% 35% into E 65% 75% DNT into S 50% 68% DNT into E 85% 18% DNT into S 35% 63% 80% into E 30% 44% DNT into S 30% 54% DNT into E 20% 89% 100% into S 60% 67% 65% into E 80% 98% 100% into S 55% 85% 75% into E 80% 47% 65% into S 55% 70% 70% into E 70% 88% 100% into S 100% 75% 95% into E 95% 25% 0% into S 40% 17% 0% into E 0% 95% 100% into S 100% 98% 100% into E 100% 91% 100% into S 55% 46% 10% into E 75% 0% 0% 53% 57% 57% 10% 38% 13% 0% 23% 0% 0% 63% 8% 82% 25% 15% 18% 65% 50% 0% 0% 80% 70% 100% 8% 93% refused 72% 78% 69% 58% 65% 70% DNT 74% 62% 74% 93% 66% 100% 98% 71% 89% 95% 85% 63% 57% 94% 94% 95% 63% PPT-picture version CLQT Cognitive Domains attention memory executive functions language visuospatial skills composite severity clock drawing CLQT Subtests Symbol Cancelation Symbol Trails Design Generation 83% 90% 92% 94% 50% 90% 88% 98% 63% 85% 52% 96% 96% DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT DNT Mild Moderate WNL Mild WNL Mild Mild Severe Severe Severe Severe Severe Severe Severe Mild Severe Severe Severe Mild Moderate Mild Moderate WNL WNL WNL Mild Mild WNL WNL WNL WNL WNL WNL WNL WNL WNL Moderate Mild WNL WNL Mild WNL Mild Severe Severe Mild Mild Moderate WNL Severe Severe Severe Severe Severe Severe Severe WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL WNL not WNL not WNL WNL not WNL not WNL WNL WNL not WNL not WNL not WNL WNL not WNL not WNL not WNL WNL not WNL not WNL not WNL WNL not WNL WNL WNL WNL WNL WNL not WNL WNL not WNL not WNL not WNL not WNL not WNL WNL WNL WNL WNL WNL WNL Note: BAT = Bilingual Aphasia Test, Receptive Language = average of Pointing, Semi-complex commands, Complex commands, Verbal auditory discrimination, semantic categories, synonyms, antonyms I & II, Semantic acceptability subtests; Expressive Langauge = object naming, Expressive Language = BNT in English and Spanish; PALPA 25 = Imageability and Frequency Visual Lexical Decision 25 on the Psycholinguistic Assessments of Language Processing in Aphasia, and the Spanish translation, EPLA 26 = Decisión lexica visual: imaginabilidad y frecuencia 26 on la Evaluación del Procesamiento Lingüístico en la Afasia, PPT = Pyramids and Palm Trees Test, CLQT: Cognitive LInguistic Quick Test, into S = translating from English into Spanish, into E = translating into English from Spanish, DNT = did not test, WNL = within normal limits. METHODS 1. Non-linguistic Flanker: decide which way the target arrow is pointing. 2. Linguistic Flanker: decide if the target word is English or Spanish. NHBA: F(1, 38) = 1.47, p = .23; BAAHigh: F(1, 8) = .80, p = .40; BAALow: F(1, 14) = 4.04, p = .06 3. Non-linguistic Triad: select the response that matches the target by shape or color. 4. Linguistic Triad: select the word that is semantically related to the target. NHBA: F(1, 38) = 34.82, p < .001; BAAHigh: F(1, 8) = 20.88, p < .01; BAALow: F(1, 14) = 1.60, p = .103 NHBA F(1, 37) = 29.85, p < .001 BAAHigh F(1, 7) = 7.9, p < .05 BAALow F(1, 13) = .54, p = .48 NHBA: F(1, 38) = 33.15, p < .001; BAAHigh: F(1, 8) = 5.55, p < .05; BAALow: F(1, 14) = 41.74, p < .001 NHBA F(1, 37) = 1.72, p = .20; BAAHigh F(1, 7) = 1.37, p = .28; BAALow F(1, 13) = 1.55, p = .23 NHBA F(1, 37) = .05, p = .82; BAAHigh F(1, 7) = 3.05, p = .12; BAALow F(1, 13) = 1.78, p = .21 NHBA F(1, 37) = 19.68, p < .001; BAAHigh F(1, 7) = .51, p = .505; BAALow F(1, 13) = .485, p = .50 NHBA: F(1, 38) = 119.54, p < .001; BAAHigh: F(1, 8) = 31.77, p < .001; BAALow: F(1, 14) = 7.57, p < .05 Congruency checklist for all tasks for NHBA, BAAHigh and BAALow. NHBA (n = 20) Task BAA (n = 13) BAAHigh (n = 5) BAALow (n = 8) acc RT acc RT acc RT acc RT Non-linguistic Flanker Linguistic Flanker LExperience.cov X    X   X X X X X X X X X Non-linguistic Triad Linguistic Triad LExperience.cov           X    X X Note: The congruency effect is notated by a ; no congruency effect is notated by an X, NHBA = neurologically healthy adults, BAAHigh = bilingual adults with high diagnostic scores, BAALow = bilingual adults with aphasia iwth low diagnostic scores, LExperience.cov = language experience cov = covariate. CONCLUSIONS HYPOTHESES Potential Outcomes Domain General Domain Specific Effects of Task Cognitive Cognitive Task Complexity Control Control    Non-linguistic Flanker X X X    Non-linguistic Triad X X X    Linguistic Flanker X X X    Linguistic Triad X X X Note: the congruency effect is notated by a ; no congruency effect is notated by an X. These outcomes should be identified in accuracy and/or latency for NHBA and BAA. 1. NHBA (and BAAHigh)  Domain General Cognitive Control 2. BAALow  Domain Specific Cognitive Control 3. BAAHigh vs BAALow: separable patient groups