Lexical attrition in younger and older bilingual adults

Abstract

Healthy monolingual older adults experience changes in their lexical abilities. Bilingual individuals immersed in an environment in which their second language is dominant experience lexical changes, or attrition, in their first language. Changes in lexical skills in the first language of older individuals who are bilinguals, therefore, can be attributed to the typical processes accompanying older age, the typical processes accompanying first‐language attrition in bilingual contexts, or both. The challenge, then, in understanding how lexical skills change in bilingual older individuals, lies in dissociating these processes. This paper addresses the difficulty of teasing apart the effects of ageing and attrition in older bilinguals and proposes some solutions. It presents preliminary results from a study of lexical processing in bilingual younger and older individuals. Processing differences were found for the older bilingual participants in their first language (L1), but not in their second language (L2). It is concluded that the lexical behaviour found for older bilinguals in this study can be attributed to L1 attrition and not to processes of ageing. These findings are discussed in the context of previous reports concerning changes in lexical skills associated with typical ageing and those associated with bilingual L1 attrition.

• Ageing
• bilingual
• lexical attrition
• compound words
• lexical decision

Notes

1. Whereas most participants were right‐handed individuals, ten individuals who reported being left‐handed were included. Their distribution across the three participant groups was comparable (about 10% of each group).

2. For the English lists, compound and single words were selected that were judged to be familiar to native speakers of Hebrew who are L2 users of English. Word length (in letters) and written word frequency (from CELEX) of the primes and targets in the various conditions were not equivalent across all the lists. Hebrew word frequency was not available for many of the Hebrew items. The authors plan to obtain frequency ratings from bilingual participants and revisit the word frequency distributions across the two LDTs.

3. One of the research questions of the larger study, not discussed in details in the present paper, was to compare and contrast the pattern of constituent priming in Hebrew and in English, in an attempt to dissociate the roles of the first constituent, the head of the compound, and the compound modifier. For that reason, it was important to obtain the data from the Hebrew only experiment first.

4. The two LDTs reported here were part of a larger session that included four LDTs as well as a naming test for the older participants and a translation‐judgement task for the NY‐younger group. These results are not reported in the current paper.

5. Only by‐subject analyses are reported here.

6. Segalowitz (1991) concluded that whereas recognition speed for L1 decreased with increased L2 proficiency, there was little evidence for differential automatic L1 processing of more‐ or less‐proficient bilinguals.

7. The authors were surprised by the large variance found for the NY‐older group in their RTs in L1, as compared to the younger group. Of course, there is always a correlation between response latency and variance size as well as between the group size and the variance size. Moreover, large variance has been reported previously for older participants on many tasks (e.g. Christensen, Mackinnon, Jorm, Henderson, Scott, & Korten, 1994), but the variance in the older group's L1 RTs appears particularly large as compared to their variance in their L2 RTs.