The study investigated phonological processing in bilingual reading for meaning. English–French and French–English bilinguals performed a category verification task in either their first or second language. Interlingual homophones (words that share phonology across languages but not orthography or meaning) and single language control words served as critical stimuli. The interlingual homophones and their control words were not members of the categories, but their interlingual homophone mates were category members (e.g., A vegetable: shoe, where chou in French means “cabbage”). The bilinguals made more errors and had longer decision latencies on homophones than on their control words, providing evidence for cross-language phonological activation of meaning. Results are discussed with respect to the Bilingual Interactive Activation Model (BIA+).

This study investigated the development of second language online auditory processing with ab initio German learners of Dutch. We assessed the influence of different levels of background noise and different levels of semantic and syntactic target word predictability on word-monitoring latencies. There was evidence of syntactic, but not lexical-semantic, transfer from the L1 to the L2 from the onset of L2 learning. An initial stronger adverse effect of noise on syntactic compared to phonological processing disappeared after two weeks of learning Dutch suggesting a change towards more robust syntactic processing. At the same time the L2 learners started to exploit semantic constraints predicting upcoming target words. The use of semantic predictability remained less efficient compared to native speakers until the end of the observation period. The improvement and the persistent problems in semantic processing we found were independent of noise and rather seem to reflect the need for more context information to build up online semantic representations in L2 listening.

In recent years, multiple studies have shown that the languages of a bilingual interact during processing. We investigated sign activation as deaf children read words. In a word–picture verification task, we manipulated the underlying sign equivalents. We presented children with word–picture pairs for which the sign translation equivalents varied with respect to sign phonology overlap (i.e., handshape, movement, hand-palm orientation, and location) and sign iconicity (i.e., transparent depiction of meaning or not). For the deaf children, non-matching word–picture pairs with sign translation equivalents that had highly similar elements (i.e., strong sign phonological relations) showed relatively longer response latencies and more errors than non-matching word–picture pairs without sign phonological relations (inhibitory effects). In contrast, matching word–picture pairs with strongly iconic sign translation equivalents showed relatively shorter response latencies and fewer errors than pairs with weakly iconic translation equivalents (facilitatory effects). No such activation effects were found in the word–picture verification task for the hearing children. The results provide evidence for interactive cross-language processing in deaf children.

This study investigated correspondence between different measures of bilingual language proficiency contrasting self-report, proficiency interview, and picture naming skills. Fifty-two young (Experiment 1) and 20 aging (Experiment 2) Spanish–English bilinguals provided self-ratings of proficiency level, were interviewed for spoken proficiency, and named pictures in a Multilingual Naming Test (MINT); in Experiment 1, the Boston Naming Test (BNT) was also used. Self-ratings, proficiency interview, and the MINT did not differ significantly in classifying bilinguals into language-dominance groups, but naming tests (especially the BNT) classified bilinguals as more English-dominant than other measures. Strong correlations were observed between measures of proficiency in each language and language-dominance, but not degree of balanced bilingualism (index scores). Depending on the measure, up to 60% of bilinguals scored best in their self-reported non-dominant language. The BNT distorted bilingual assessment by underestimating ability in Spanish. These results illustrate what self-ratings can and cannot provide, illustrate the pitfalls of testing bilinguals with measures designed for monolinguals, and invite a multi-measure goal-driven approach to classifying bilinguals into dominance groups.

Researchers on bilingual processing can benefit from computational tools developed in artificial intelligence. We show that a normalized Levenshtein distance function can efficiently and reliably simulate bilingual orthographic similarity ratings. Orthographic similarity distributions of cognates and non-cognates were identified across pairs of six European languages: English, German, French, Spanish, Italian, and Dutch. Semantic equivalence was determined using the conceptual structure of a translation database. By using a similarity threshold, large numbers of cognates could be selected that nearly completely included the stimulus materials of experimental studies. The identified numbers of form-similar and identical cognates correlated highly with branch lengths of phylogenetic language family trees, supporting the usefulness of the new measure for cross-language comparison. The normalized Levenshtein distance function can be considered as a new formal model of cross-language orthographic similarity.

The present study examined the language development of children adopted from China to examine possible early age effects with respect to their use of complement clitics, lexical diversity and verb morphology. We focused on these aspects of French because they distinguish second language learners of French and native French-speaking children with language impairment from children learning French as a native language and, in the case of object clitics and certain verb tenses, are relatively late to emerge in native speakers. Thus, it might be expected that they would be susceptible to the delayed onset of acquisition of French experienced by internationally adopted children. Language samples of twelve adopted children from 3;6 to 4;8 living in French-speaking families were analyzed and compared to those of non-adopted monolingual French-speaking children of the same age, sex and socio-economic status. The adopted and control children had similar levels of socio-emotional adjustment and non-verbal intellectual abilities. The adopted children exhibited accelerated language development in general, and there were no significant differences between the internationally adopted and control children with respect to lexical diversity and verb tense. However, the adopted children made significantly more errors using complement clitics, and in particular object clitics, compared to the non-adopted children. The results are discussed in terms of possible effects related to delayed age of acquisition of French.

This study compared the comprehension of syntactically simple with more complex sentences in Italian–English adult bilinguals and monolingual controls in the presence or absence of sentence-level interference. The task was to identify the agent of the sentence and we primarily examined the accuracy of response. The target sentence was signalled by the gender of the speaker, either a male or a female, and this varied over trials, where the target was spoken in a male voice the distractor was spoken in a female voice and vice versa. In contrast to other work showing a bilingual disadvantage in sentence comprehension under conditions of noise, we show that in this task, where voice permits selection of the target, adult bilingual speakers are in fact better able than their monolingual Italian peers to resist sentence-level interference when comprehension demands are high. Within bilingual speakers we also found that degree of proficiency in English correlated with the ability to resist interference for complex sentences both when the target and distractor were in Italian and when the target was in English and the distractor in Italian.

The purpose of this study was to determine if different language measures resulted in the same classifications of language dominance and proficiency for a group of bilingual pre-kindergarteners and kindergarteners. Data were analyzed for 1029 Spanish–English bilingual pre-kindergarteners who spanned the full range of bilingual language proficiency. Parent questionnaires were used to quantify age of first exposure and current language use. Scores from a short test of semantic and morphosyntactic development in Spanish and English were used to quantify children's performance. Some children who were in the functionally monolingual range based on interview data demonstrated minimal knowledge of their other languages when tested. Current use accounted for more of the variance in language dominance than did age of first exposure. Results indicate that at different levels of language exposure children differed in their performance on semantic and morphosyntax tasks. These patterns suggest that it may be difficult to compare the results of studies that employ different measures of language dominance and proficiency. Current use is likely to be a useful metric of bilingual development that can be used to build a comprehensive picture of child bilingualism.

A corpus analysis of phonological word-forms shows that English words have few phonological neighbors that are Spanish words. Concomitantly, Spanish words have few phonological neighbors that are English words. These observations appear to undermine certain accounts of bilingual language processing, and have significant implications for the processing and representation of word-forms in bilinguals.

The purpose of this article was to test the predictions of a speech production model of cross-linguistic influence in French–English bilingual children. A speech production model predicts bidirectional influence (i.e., bilinguals’ greater use of periphrastic constructions like the hat of the dog relative to monolinguals in English and reversed possessive constructions like *chien chapeau to refer to a dog's hat). In contrast, other explanations predict unidirectional influence from French to English. Possessive constructions were elicited from preschool French–English bilingual children as well as preschool French and English monolinguals within the same age range. The results showed bidirectional influence, consistent with a speech production model.

According to the nonce borrowing hypothesis (NBH), “[n]once borrowings pattern exactly like their native counterparts in the (unmixed) recipient language” (Poplack & Meechan, 1998a, p. 137). Nonce borrowings (Sankoff, Poplack & Vanniarajan, 1990, p. 74) are “lone other-language items” which differ from established borrowings in terms of frequency of use and recognition. Lone other-language items are singly occurring words from the “donor” language which are preceded and followed by words or phrases from the “recipient” language. Whether such other-language words belong only to the donor language (and are classed as codeswitches) or to both the donor and the recipient language (and are classed as borrowings) is both a theoretical and a practical issue. Poplack & Meechan (1998a) suggest that this question can be settled by measuring the linguistic integration of donor-language words, so that infrequent donor-language words which behave like their recipient-language counterparts are categorised as (nonce) borrowings. This suggests that frequency of use need play no role in the extent to which other-language items are linguistically integrated into the recipient language. We challenge this hypothesis with an analysis of soft mutation on English-origin verbs in Welsh, which shows that integration is related to frequency.

The study examined individual differences in code-switching to determine the relationship between code-switching frequency and performance in verbal and non-verbal task switching. Seventy-eight Cantonese–English bilinguals completed a semi-structured conversation to quantify natural code-switching, a verbal fluency task requiring language switching, and two non-verbal switching tasks. Participants who engaged in more conversational code-switching showed smaller costs in verbal task switching than those who switched languages less frequently. Participants performed similarly to bilinguals in previous studies on non-verbal switching tasks, but in this case performance was not linked to the degree of conversational code switching. The difference in the influence of code-switching for verbal and non-verbal executive control tasks indicates a dissociation between domains for the mechanism of task switching.

The aim of the present study was to investigate the attentional mechanisms of multilingual children with differential degrees of language competence. For this purpose, 118 children (61 female/57 male; mean age 10.9 years (SD = 0.29); early acquisition multilinguals) from the Ladin valleys in South Tyrol, Italy, performed the Attentional Network Test (ANT). Our results proved that proficiency levels in early multilingual children may play a crucial role in the development and enhancement of the alerting component of the attentional system. Interestingly enough, we were able to deduce that linguistic competence rather than competence in other skill domains may have a decisive role in the alerting component. We suggest that the peculiarity of highly competent multilinguals relies on their ability to better detect, and consequently react faster to, the target stimulus than their less competent multilingual peers.

Lexical access in bilinguals is known to be largely non-selective. However, most studies in this area have involved bilinguals whose two languages share the same script. This study aimed to examine bilingual lexical access among bilinguals whose two languages have distinct scripts. Korean–English bilinguals were tested in a phoneme monitoring task in their first or second language. The results showed a simultaneous activation of the non-target language in a monolingual task, suggesting non-selective lexical access even among bilinguals whose two languages do not share the same script. Language dominance did not affect the pattern of results.

This study aimed to investigate the capacity of coordinating comprehension and production processes and the role of phonological working memory in simultaneous interpreting. To this end we evaluated the Articulatory Suppression (AS) effect in three groups of participants, monolingual controls, students of interpreting and professional interpreters. Three variables were examined, the material to be studied (words, pseudo-words), the complexity of the articulations (simple, complex) and the articulatory rate (participants produced their speech at their own rate). Monolingual controls showed AS effect in all study conditions; students of interpreting showed AS effect in complex study conditions and professional interpreters showed AS effect only when they studied pseudo-words and produced complex articulations. These results suggest that coordinating comprehension and production processes in interpreters is mediated by the retrieval of lexical–semantic information and the distribution of the speech.

In two experiments we investigated the relationship between the working memory skills of sign language interpreters and the quality of their interpretations. In Experiment 1, we found that scores on 3-back tasks with signs and words were not related to the quality of interpreted narratives. In Experiment 2, we found that memory span scores for words and signs under oral articulatory suppression were related to the quality of interpreted narratives. We argue that the insensitivity to articulatory suppression in memory span tasks reflects the interpreters' ability to bind information from multiple sources in episodic memory. This enhanced ability leads to less reliance on the retention of information from the source language in memory during interpreting, and will positively affect the quality of interpretations (Padilla, Bajo & Macizo, 2005). Finally, in contrast to previous studies on the memory spans for signs and words (Hall & Bavelier, 2010), we found that the memory spans scores for spoken words and signs were equally large. We argue that the use of a large set of phonologically complex stimuli in the present study may have stimulated participants to use a speech-based code to store and retain the signs in short-term memory.

In this rejoinder to Shana Poplack's response to Stammers & Deuchar (this issue), we argue that our reformulation of the nonce borrowing hypothesis (NBH) to include specific reference to frequency was needed in order to make the hypothesis more precise and testable. Furthermore, in order to test the assumption that codeswitching (CS) and borrowing (B) are two distinct categories, it was necessary to suspend this assumption in our study. This led us to find support for a possible CS/B distinction, but not for the categorical integration of all borrowings regardless of frequency. In discussing our methods, we maintain that soft mutation is an appropriate measure of morphosyntactic integration in Welsh, and is no more purely phonetic than any other morphosyntactically triggered process.

Stimuli used in cued language switching studies typically consist of digits or pictures. However, the comparability between both stimulus types remains unclear. In the present study, we directly compared digit and picture naming in a German–English language switching experiment. Because digits represent a semantic group and contain many cognates, the experiment consisted of four conditions with different stimulus sets in each condition: digits, standard language switching pictures, pictures depicting cognates, and semantically-related pictures. Digit naming caused smaller switch costs than picture naming. The data suggest that this difference can be attributed to phonology. Both methodological and theoretical implications are discussed.

We report the results of an affective priming study conducted with proficient sequential German and French bilinguals to assess automatic affective word processing in L1 and L2. Additionally, a semantic priming task was conducted in both languages. Whereas semantic priming effects occurred in L1 and L2, and significant affective priming effects were found in L1, affective priming effects in L2 were only found for participants with high levels of language immersion and frequency of L2 use. These results suggest that for sequential bilinguals the intensity of L2 use largely determines whether emotional words in L2 automatically activate their affective connotations.