This article challenges a recent proposal for the theoretical interpretation of L1 and L2 interaction that results from the abrupt change of language environment in internationally adopted children. According to this proposal (Pallier, Dehaene, Poline, LeBihan, Argenti, Depoux and Mehler, 2003; Ventureyra, Pallier and Yoo, 2004), such children experience a total loss of their L1, while, as adults, they exhibit a nativelike ultimate attainment of their L2. These authors suggest that what they see as a total loss of L1 allows a resetting of the neural network that normally subserves L1 retention and hence permits a complete acquisition of the L2. Data from two of our own research projects, one on L1 remnants in Korean adoptees in Sweden (see Park, forthcoming), and the other on age of acquisition and ultimate L2 attainment of Swedish (see Abrahamsson and Hyltenstam, in press), which included data from Latin American adoptees in Sweden among other participants, suggest (i) that L1 remnants are indeed maintained, (ii) that L2 attainment is not enhanced by severe L1 attrition, and (iii) that there is an age dimension to both the degree of L1 attrition and the level of L2 ultimate attainment in international adoptees. We therefore contend that a maturational interpretation of language replacement data is preferable.

Functional magnetic resonance imaging (fMRI) was used to compare brain activation from native Japanese (L1) readers reading hiragana (syllabic) and kanji (logographic) sentences, and English as a second language (L2). Kanji showed more activation than hiragana in right-hemisphere occipito-temporal lobe areas associated with visuospatial processing; hiragana, in turn, showed more activation than kanji in areas of the brain associated with phonological processing. L1 results underscore the difference in visuospatial and phonological processing demands between the systems. Reading in English as compared to either of the Japanese systems showed more activation in inferior frontal gyrus, medial frontal gyrus, and angular gyrus. The additional activation in English in these areas may have been associated with an increased cognitive demand for phonological processing and verbal working memory. More generally, L2 results suggest more effortful reading comprehension processes associated with phonological rehearsal. The study contributes to the understanding of differential brain responses to different writing systems and to reading comprehension in a second language.

We examined how the semantics of causal expressions in Russian and English might differ and how these differences might lead to changes in the way second language learners understand causal expressions in their first language. According to the dynamics model of causation (Wolff, 2007), expressions of causation based on CAUSE verbs (make, force) differ from expressions based on ENABLE verbs (let, help, allow) primarily in terms of the causee's inherent tendency toward an endstate, that is, the causee's physical or intentional inclination for a particular state of affairs. In Russian, the tendency appears to be based on internally derived forces, whereas in English, the tendency may be based on either internally or externally derived forces. In two experiments, English and Russian monolinguals and bilinguals described animations in which the causee's tendency was systematically varied. When the causee's tendency was ambiguous, English and Russian monolinguals’ descriptions differed, suggesting that the causal expressions differ in meaning across languages. Of primary interest, Russian–English and English–Russian bilinguals’ causal descriptions differed from those of monolingual speakers of their first language, and in the direction of the second language, even though they performed the task in the first language. This L2 → L1 transfer is explained in terms of the memory phenomenon of retrieval-induced reconsolidation.

In this paper, I investigate the status of the so-called “weaker” language, French, in French/German bilingual first language acquisition, using data from two children from the DuFDE-corpus (see Schlyter, 1990a), Christophe and François. Schlyter (1993, 1994) proposes that the “weaker” language in the unbalanced children she studied has the status equivalent to that of a second language (L2). I will verify this assumption on the basis of certain grammatical phenomena, such as the use of subject clitics, null subjects and negation, with respect to which L1 and L2 learners show different developmental patterns. The results indicate that the “weaker” language of the children analyzed in this study cannot be interpreted as an L2. Both children behave predominantly like monolingual and balanced bilingual L1 learners in both languages.

The work presented here aims to account for the structure of intrasentential code switching between English and Afrikaans within the framework of feature checking theory, a theory associated with minimalist syntax. Six constructions in which verb position differs between English and Afrikaans were analysed in terms of differences in the strength of particular features associated with functional categories, and the ability of verbs of either language to check these features. Predictions for the well-formedness of code-switched constructions were informed by data elicited from thirty fluently bilingual participants by means of relative judgements of visually-presented code-switched sentences and auditorily-presented code-switched utterances, and a sentence construction task. Findings indicated straightforward support for the predictions for adverb, focalisation, and topicalisation constructions, but less support for embedded that and wh clauses and yes-no questions. Alternative explanations for the latter results are proposed. The work suggests that the same mechanisms and devices proposed to account for monolingual data can also account for code-switching data.

This study explores the extent to which bilingual speakers in stable bilingual communities become fully bilingual in their two community languages. Growing evidence shows that in bilingual communities in which one language is very dominant, acquisition of the dominant language may be quite unproblematic across sub-groups, while acquisition of the minority language can be hampered under conditions of reduced input. In Wales, children are exposed to both English and Welsh from an early age, either in the home or at school, or both. The data reported here indicate that regardless of home language background, speakers develop equivalent, mature command of English, but that command of Welsh is directly correlated with the level of input in Welsh in the home and at school. Furthermore, maintenance of Welsh in adulthood may be contingent on continued exposure to the language. The data have implications for theories of bilingual acquisition in stable versus immigrant bilingual communities, for optimal conditions for bringing up bilingual children, and for theories of critical periods of acquisition.

A number of recent studies have argued that bilingual children's language comprehension and production may be affected by cross-linguistic influence. The overall aim of this study was to investigate whether the ability to judge the grammaticality of a construction in one language is affected by knowledge of the corresponding construction in the other language. We investigated how English–Italian and Spanish–Italian bilingual children and monolingual peers judged the grammaticality of plural NPs in specific and generic contexts in English and in Italian. We also explored whether language of the community, age, and the typological relatedness of the bilinguals’ two languages significantly affected their performance. While performance in English was overall poor, no significant differences existed between the English–Italian bilinguals and the monolinguals. In contrast, we found that knowledge of English affected the bilinguals’ ability to discriminate between grammatical and ungrammatical sentences in Italian. The English–Italian bilinguals were significantly less accurate than both the monolinguals and the Spanish–Italian bilinguals in a task where they simply had to rely on the local definite article cue to reject ungrammatical bare plurals in generic contexts. Language of the community and age also played a significant role in children's accuracy.

Spanish phonological development was examined in six sequential bilingual children at the point of contact with English and eight months later. We explored effects of the English vowel and consonant inventory on Spanish. Children showed a significant increase in consonant cluster accuracy and in vowel errors. These emerging sequential bilingual children showed effects of English on their first language, Spanish. Cross-linguistic transfer did not affect all properties of the phonology equally. Negative transfer may occur in specific areas where the second language is more complex, requiring reorganization of the existing system, as in the transition from the Spanish five-vowel to the English eleven-vowel system.