Читать книгу The Concise Encyclopedia of Applied Linguistics - Carol A. Chapelle - Страница 228

Languages differ from one another in the way their vocabularies carve up conceptual space and the physical world. For instance, both Russian and English have separate words for glasses and cups (stakany and chashki in Russian), but the exact reference of these words differs between these languages: Paper cups are called stakanchiki (small glasses) in Russian (Pavlenko, 2005). Variation in the expression of concepts across languages exists for many semantic domains, perhaps all, and it occurs across both distantly related and closely related languages (Majid, Jordan, & Dunn, 2015). One of the best‐known examples of differential word‐to‐concept mapping across languages concerns the semantic domain of color concepts. Languages vary widely in the number of color words they possess to describe the color spectrum and, of course, the number of color words used in a specific language has consequences for the exact reference of each of these words: The smaller the number of color words, the larger the range of hues referred to by each of them.

The consequence of this cross‐language variability in color terminology for color categorization and representation in bilinguals has been examined since around 1960. One study concerned a detailed investigation of color naming in Navaho–English bilinguals and Navaho and English monolinguals (Ervin, 1961). Ervin first performed a detailed contrastive analysis of the color systems of Navaho and English. This analysis revealed, for instance, that litso, the closest Navaho translation of yellow, is the favored response of monolingual Navaho speakers to hues across a much larger part of the color spectrum than the range of hues exciting yellow in monolingual speakers of English. Assuming an influence from the colors' names in the nontarget language, Ervin expected the response probabilities in the target language to differ between the bilinguals and monolinguals. For instance, when presenting a yellowish color patch and inviting a color response in Navaho, the bilinguals were expected to produce fewer litso responses than the monolingual Navaho controls. These and other predictions from the contrastive analysis were borne out by the data.

Ervin explained these results in terms of coactivation in bilingual memory of the representation of the presented color's name in the nonresponse language. For example, a Navaho–English bilingual may say tatLqid (‘green’) to a color patch that most Navaho monolinguals would call litso (‘yellow’) because the common English name for the depicted color (green) is coactivated with litso. Coactivated green then activates tatLqid via a connection between these two words in memory and tatLqid emerges as the response. An alternative account of such semantic accents is in terms of differences in the conceptual representations of color words between bilinguals and monolinguals. This is how Caskey‐Sirmons and Hickerson (1977) explained the results of a similar study wherein English L2 speakers with various Asian languages as L1 were tested. Specifically, these authors assumed that, in the course of acquiring L2 English, broader conceptual representations of color words had been formed by merging the concepts associated with L1 color words and those associated with these words' closest L2 translations. But also here an interpretation of the results in terms of co‐activation of lexical representations in the nontarget language cannot be ruled out (De Groot, 2014), and a similar indeterminacy may apply to the different behavior of monolinguals and bilinguals in tasks that probe bilingual conceptual representation in other semantic domains (e.g., Ameel, Malt, Storms, & Van Assche, 2009, a study that examines the representation of common household objects).