Kitchen Russian: First-language object naming by Russian-English bilinguals

Abstract

Kitchen Russian: First-language Object Naming by Russian-English Bilinguals Barbara C. Malt (barbara.malt@lehigh.edu) Department of Psychology, Lehigh University, 17 Memorial Drive East Bethlehem, PA 18015 USA Aneta Pavlenko (apavlenk@temple.edu) Department of Curriculum, Instruction, and Technology in Education, Temple University, 451 Ritter Hall Philadelphia, PA 19122 USA Abstract A bilingual’s two lexicons are linked rather than isolated from each other. An implication of this linkage is that the contents of one might influence the contents of the other. We examined naming of common household objects by early, childhood, and late Russian-English bilinguals to assess a possible second language (L2) English influence on first language (L1) Russian naming patterns and typicality ratings. L2 influence was evident in the data even for late bilinguals. It was most pronounced with earliest arrival and entailed both narrowing and broadening of linguistic categories. Keywords: bilingualism; bilingual word use; semantics Introduction Bilinguals must be able to understand and communicate more or less the same thoughts using two different languages. At the lexical level, this feat is achieved via a cognitive architecture in which words of the two languages are connected indirectly through links to a common conceptual space and directly through links between word forms (e.g., Kroll & Stewart, 1994). These connections have a consequence for on-line comprehension and production: Activation of words in one language prompts activation of words in the other, requiring the bilingual to inhibit the words of the non-target language for each utterance (e.g., van Hell & Dijkstra, 2002). Another consequence of the interconnections, only recently beginning to be appreciated, is that the meanings associated with the word forms of one language may be influenced by the existence of the other lexicon. Many roughly corresponding words, even cognates, are not true translation equivalents across languages as used by monolinguals. For instance, the set of objects called botella by Argentinean Spanish monolinguals only partially overlaps with the set called bottle by American English monolinguals (Malt, Sloman, Gennari, Shi, & Wang, 1999). If there are interconnections between representations when two lexicons co-exist in one mind, the meanings encoded in one lexicon might influence those of the other. There are three potential directions of lexical influence between a bilingual’s two languages. Most intuitively plausible is an influence of the L1 (first-learned language) on an L2 (second-learned language), especially early in acquisition. Learners may import the meaning of an L1 word as their first attempt to connect meaning to an L2 word (e.g., Kroll & Stewart, 1994). More surprising is recent evidence for a mutual influence of two languages when they have been learned in parallel. Bilinguals who have been raised with two languages from birth differ from monolingual users of either language in their patterns of word use; they converge the patterns of their two languages toward each other (Ameel, Storms, Malt, & Sloman, 2005; see also Brown & Gullberg, 2008; Pavlenko & Jarvis, 2002). This effect occurs despite the fact that they have had the benefit of extended exposure to both languages from a young age and have achieved high levels of overall proficiency in both. These factors would seem to predict high mastery of the two lexical systems. Yet it might be that two lexicons being developed in parallel are made vulnerable by interleaved exposure to the words of the languages (possibly initially not even fully distinguished as belonging to different systems), and by the fact that neither lexicon is well-established when the other is being learned. What, then, should be expected of the third possible direction of influence, namely, an L2 influence on L1? Will speakers who begin life with one language but then acquire a second show any influence of the second-learned language on their use of the first? The considerations just mentioned suggest possibly not: Once one language is reasonably well- established, the content of its lexicon may be relatively invulnerable to any influence from a new one, even if there is competition between them in production. Empirically, however, there is growing evidence of L2 influence on well- established L1s in other linguistic domains from phonology to pragmatics (see, e.g., Cook, 2003; Pavlenko, 2000), indicating that earlier-learned representations can have some vulnerability to influence from later-learned ones. If an influence also exists for the lexical domain, it would suggest that the direct and indirect connections between representations of the two language systems leave them perpetually open to cross-language influence, possibly because of the fact that accessing lexical items of one language activates lexical items of the other. These activations may provide an opportunity for the stored memory traces of one language to be influenced by the other (Ameel, Malt, Storms, & von Assche, in press; Wolff & Ventura, in press).