نامگذاری تصاویر اسامی هم ریشه و غیرهم ریشه در زبان پریشی دوزبانه

Previous research has found differences in the speed and accuracy of responses involving concrete cognate nouns and non-cognate nouns in a range of written and “on-line” tasks using neurologically unimpaired, bilingual adults. The present study investigated whether cognateness affects verbal confrontation naming performance in balanced French/English bilinguals (N = 15 aphasic and 15 non-aphasic subjects). Subjects met selection criteria for equal proficiency, regular use, and early acquisition of both languages. Results of a picture naming test show that cognate pictures were more often correctly named in both languages than were non-cognates. Some error types and self-correction behaviors also varied with cognate status. There were similarities between the results of this study and those of previous studies of monolingual naming. Some error types and self-correction strategies appear to be unique to bilingual speakers. Theoretical questions and treatment applications arising from these findings are outlined.

As the demographics of many areas of the world change, speech-language pathologists are seeing increasing numbers of bilingual aphasic adults. Bilingualism is no longer an occasional feature in neurogenic language disorders, but is “a phenomenon every clinic must be prepared to cope with” (Paradis, 1995, p. 219). Bilingual aphasia is, therefore, a topic of increasing clinical interest. Much of the pioneering psycholinguistic research on bilingualism asked, “Do bilinguals have one lexicon or two?” Although it was a useful starting point for a new field of study, this question is now seen as too broad to be meaningful Altenberg 1989, Diller 1974, Durgunoglu and Roediger 1987 and Snodgrass 1993. Just as the question “Is aphasia therapy efficacious?” has been replaced by “What type of therapy is efficacious for whom?” (to paraphrase Wertz, 1987), the question of “one lexicon or two” has been replaced by “What kinds of lexical stores or processes exist for what kinds of bilinguals for what types of words and for which language tasks?” This study asks whether cognate status (cognates are word pairs with similar form and the same meaning in two languages) influences naming accuracy and error types on a confrontation naming task. How words are represented or retrieved in the bilingual lexical system(s) is not addressed for two main reasons. First, the inferences drawn from lexical processing studies are controversial. Differences across a bilingual’s two languages or differences between bilingual and unilingual performance are often interpreted as revealing the organization of the bilingual lexicon. This practice is called into question by studies by Kolers and Roediger (1984) and by Durgunoglu and Roediger (1987) showing that the same subjects produce results consistent with both a combined store (items from both languages in a single lexicon) and a dual store (separate, language-specific lexicons) depending upon the experimental task. Diller (1974) and Altenberg (1989) have both raised theoretical objections to the single/dual store debate. It seems prudent, therefore, to follow Snodgrass’s (1993) advice: What can we conclude from this review about the organization of the bilingual lexicon? The major conclusion I reached is that investigators were concluding too much about the process underlying each of the tasks on the basis of too few data. I would urge researchers of the bilingual lexicon to spend more time collecting data so that we can be sure that a particular pattern really exists before making sweeping theoretical statements about the meaning of this pattern. (p.110) Second, while we await the resolution of the broad questions of how bilingual language is organized, bilingual aphasic patients require assessment and treatment. Knowing the influence of cognate status on confrontation naming in bilingual adults with aphasia can be of immediate practical benefit. If cognates are easier to produce than non-cognates, treatment might focus more on the latter, for instance. Although there is little clinically oriented literature on how to assess and treat bilingual aphasia (Roberts, 1998), the studies of lexical performance in non-aphasic adults have begun to identify the relevant variables. Among these are the level of bilingual proficiency of subjects and the type of stimuli used.

The findings will be presented in answer to the three research questions. Influence of Cognate Status on Naming Accuracy The cognate pictures were more often correctly named in both languages and more often correctly named in English than the non-cognate pictures. Table 3 displays the relevant means and standard deviations. For each word, the maximum score was the number of subjects in each group (15). A one-way ANOVA with alpha set at .01 was carried out to test for the effect of cognateness on the naming accuracy of the 50 cognate and 50 non-cognate words. There was a significant effect for cognateness in the aphasic group but not the normal group for “Both correct” (F(1,98) = 8.20, p < .005) and for “English correct” (F(1,98) = 9.64, p < .005). Table 3. Mean Number of Subjects Accurately Naming in French and English*+legendlegend Normal group Aphasic group Variable Cognate Non-cognate Cognate Non-cognate Both languages correct Mean 12.72 11.52 8.32 6.08 (SD) (2.89) (3.80) (3.90) (3.92) Range 5–15 1–15 2–14 0–14 French correct Mean 13.04 13.00 10.36 9.84 (SD) (2.67) (2.32) (3.99) (3.86) Range 6–15 7–15 2–15 2–15 English correct Mean 13.48 12.36 9.64 7.40+ (SD) (2.29) (3.52) (3.46) (3.75) Range 6–15 7–15 3–14 0–14 Neither correct* Mean 1.20 1.16 3.32 3.84 (SD) (2.10) (2.01) (3.51) (3.90) Range 0–8 0–7 0–12 0–13 French only correct Mean .32 1.48 2.04 3.76 (SD) (.62) (1.81) (1.53) (2.11) Range 0–2 0–7 0–7 1–9 English only correct Mean .76 .84 1.32 1.32 (SD) (1.04) (0.94) (1.06) (1.20) Range 0–3 0–3 0–4 0–4 * Not tested statistically. + Significant difference between cognates and non-cognates (p ⩽ .005) for aphasic group only legend Significant difference between cognates and non-cognates for both groups (p ⩽ .005). legend “French correct” and “English correct” = mean number of subjects correctly naming each word in French or English, independent of whether it was correctly named in the other language. Table options Responses that were correct in only one language showed a cognate effect, but only in French. A picture was correctly named only in French significantly more often if it was a non-cognate than a cognate (Normals F(1,98) = 18.38, p < .005 and Aphasics F(1,98) = 21.98, p < .005).