In their keynote contribution, Poplack, Zentz & Dion (henceforth PZD; Poplack, Zentz & Dion, 2011, this issue) propose an interesting “scientific test of convergence” (under section heading: “Introduction”) which contains criteria to check whether a particular feature in a given language in contact with another one is due to language contact or not. This is a valiant endeavor with a laudable goal. It is valiant because the answer to this question requires a complex investigation of the languages at issue. It is laudable since it is commonly believed that a given feature of a language in contact with another one is the result of convergence. This belief however is, in general, only a mere conjecture due to superficial similarities of the features at issue, for which no empirical evidence is provided. Yet, there is no doubt that PZD accomplish their endeavor in an outstanding manner. Based on a thorough study of substantial data from Canadian French and Canadian English, they demonstrate in a convincing way how it is possible to reveal whether a given feature is contact-induced or not.

The present study investigates voice onset times (VOTs) to determine if cognates enhance the cross-language phonetic influences in the speech production of a range of Spanish–English bilinguals: Spanish heritage speakers, English heritage speakers, advanced L2 Spanish learners, and advanced L2 English learners. To answer this question, lexical items with considerable phonological, semantic, and orthographic overlap (cognates) and lexical items with no phonological overlap with their English translation equivalents (non-cognates) were examined. The results indicate that there is a significant effect of cognate status in the Spanish production of VOT by Spanish–English bilinguals. These bilinguals produced /t/ with longer VOT values (more English-like) in the Spanish production of cognates compared to non-cognate words. It is proposed that the exemplar model of lexical representation (Bybee, 2001; Pierrehumbert, 2001) can be extended to include bilingual lexical connections by which cognates facilitate phonetic interference in the bilingual mental lexicon.

Various theories of developmental language impairments have sought to explain these impairments in modality-specific ways – for example, that the language deficits in SLI or Down syndrome arise from impairments in auditory processing. Studies of signers with language impairments, especially those who are bilingual in a spoken language as well as a sign language, provide a unique opportunity to contrast abilities across language in two modalities (cross-modal bilingualism). The aim of the article is to examine what developmental sign language impairments can tell us about the relationship between language impairments and modality. A series of individual and small group studies are presented here illustrating language impairments in sign language users and cross-modal bilinguals, comprising Landau-Kleffner syndrome, Williams syndrome, Down syndrome, Autism and SLI. We conclude by suggesting how studies of sign language impairments can assist researchers to explore how different language impairments originate from different parts of the cognitive, linguistic and perceptual systems.

The study investigated single-word spelling performance of 33 English- and 38 Greek-speaking monolingual children, and 46 English- and Greek-speaking bilingual children (age range from 6;7 to 10;1 years). The bilingual children were divided into two groups on the basis of their single-word reading and spelling performance in Greek. In line with predictions, we found that scores on an assessment of phonological awareness were a significant predictor of spelling in English for the bilingual children with stronger Greek literacy skill. Phonological awareness scores were also a strong predictor of spelling in Greek in the monolingual Greek-speaking children. For the bilingual children with weaker Greek literacy ability, spelling in English was predicted by performance in a test of visual memory. This was more in line with results for the monolingual English-speaking children, for whom spelling performance was predicted by visual memory and phonological awareness scores. Qualitative analysis of misspellings revealed that phonologically appropriate errors were significantly greater in the strong Greek literacy ability bilingual group than the weaker Greek literacy ability bilingual group. Stimulus analyses using regression techniques are also reported. The results are interpreted to suggest that in biliterates literacy processes are transferred from one language to the other (Mumtaz & Humphreys, 2002).

This article explores the widely documented difficulty with object clitics in the acquisition of French. The study investigates the effects of L1 transfer and processing limitations on the production and comprehension of object clitics in child L2 learners of French with different L1 backgrounds (Chinese, Spanish). The Spanish-speaking learners performed better than Chinese-speaking learners on clitic-related tasks, indicating a facilitative effect of transfer when the L1 also has object clitics. Yet no evidence was found for (negative) transfer of null objects from Chinese to French, as learners consistently rejected interpretations requiring referential null objects on a receptive task. The frequency of Chinese-speaking learners’ object omissions in production was negatively correlated with an independent measure of working memory (backward digit span), consistent with the hypothesis that object clitic omission is affected by processing limitations. These findings are discussed within a psycholinguistic model of syntactic encoding during language production.

Poplack, Zentz and Dion (PZD; Poplack, Zentz & Dion, 2011, this issue) examine the often unquestioned assumption that the existence of preposition stranding (PS) in Canadian French is linked to the presence of a contact situation with English in the North American context. Although this issue has been the topic of previous research from a syntactic perspective (Bouchard, 1982; Vinet, 1979, 1984), to my knowledge, it has never been explored using variationist sociolinguistic methods applied to a large corpus of spontaneous speech, with emphasis on code-switchers as potential agents of contact-induced change.

We begin with the hypothesis that all people are “bilingual” because every language contains ingredients from several grammars, just as English exhibits both an Anglo-Saxon and a Latinate vocabulary system. We argue that the dominant grammar is defined by productivity and recursion in particular. Although current evidence is sparse, in principle, for a child who shows Specific Language Impairment (SLI) in a bilingual environment, richer modules in one grammar may help trigger more obscure modules in another language. Thus, if one language has a rich case system, it may help a child see an impoverished case system in another grammar. Examples from prepositional systems, wh-movement, recursive possessives and others are discussed. In general, a second language can be beneficial to the SLI child in the acquisition of both languages. Minimalism offers a level of abstraction where these cross-language connections can most naturally be stated.

Though many previous studies have reported enhanced cognitive control in bilinguals, few have investigated if such control is modulated by language proficiency. Here, we examined the inhibitory control of high and low proficient Hindi–English bilinguals on an oculomotor Stroop task. Subjects were asked to make a saccade as fast as possible towards the appropriate colour patch among competitors and distractors suppressing an eye movement evoked by the meaning of the word. High proficient bilinguals quickly oriented their attention towards the correct colour patch while effectively controlling the Stroop interference compared with low proficient subjects, on both colour and direction words. High proficient bilinguals also had fewer saccadic errors and demonstrated overall faster saccadic latency on all trial types. The results provide strong evidence for enhanced oculomotor control in proficient bilinguals compared with the less proficient ones.

This study investigates the differential effects of language attrition in two diverse linguistic domains: verb placement and object expression. Linguistic phenomena at the syntax – discourse interface, such as object expression, have been shown to be more vulnerable to attrition than narrow syntax properties, such as verb placement. This study aims to test this hypothesis by analysing spoken data from Portuguese–German bilinguals who have moved away from the dominant German environment. The results show that the speakers who have lost continued German input after the age of eleven years exhibit difficulties regarding object expression in German but do not reveal any relevant syntactic deficits in the domain of verb placement.

In this study we examined how the introduction of a parallel finger-tapping task influences second language (L2) speech encoding mechanisms and monitoring processes, and how the level of proficiency impacts the efficiency and accuracy of L2 performance under single and dual task conditions. The results indicate that imposing dual task demands had a negative effect on the accuracy of lexical selection and the efficiency of error-correction processes. We argue that this can be explained with reference to attentional bottleneck effects on lexical selection processes and on monitoring. The findings also reveal that the level of L2 competence influenced both the speed and the accuracy of speech encoding processes and the efficiency of monitoring.

Adults and children learning a second language show difficulty accessing expressive vocabulary that appears accessible receptively in their first language (L1). We call this discrepancy the receptive–expressive gap. Kindergarten Spanish (L1) – English (L2) sequential bilinguals were given standardized tests of receptive and expressive vocabulary in both Spanish and English. We found a small receptive–expressive gap in English but a large receptive–expressive gap in Spanish. We categorized children as having had high or low levels of English exposure based on demographic variables and found that the receptive–expressive gap persisted across both levels of English exposure. Regression analyses revealed that variables predicting both receptive and expressive vocabulary scores failed to predict the receptive–expressive gap. The results suggest that the onset of the receptive–expressive gap in L1 may have been abrupt. We discuss possible mechanisms underlying the phenomenon.

We are most grateful to our commentators for their careful reading of our Keynote Article (henceforth KA) and their incisive observations on contact-induced change, and for the many challenging and thought-provoking issues they raise. We welcome the opportunity to respond to (some of) them, especially since, perhaps not surprisingly, these are symptomatic of the very issues in the field of contact linguistics that prompted us to write this KA in the first place.

This study asked whether bilinguals construct a language-independent level of information structure for the sentences that they produce. It reports an experiment in which a Polish–English bilingual and a confederate of the experimenter took turns to describe pictures to each other and to find those pictures in an array. The confederate produced a Polish active, passive, or conjoined noun phrase, or an active sentence with object–verb–subject order (OVS sentence). The participant responded in English, and tended to produce a passive sentence more often after a passive or an OVS sentence than after a conjoined noun phrase or active sentence. Passives and OVS sentences are syntactically unrelated but share information structure, in that both assign emphasis to the patient. We therefore argued that bilinguals construct a language-independent level of information structure during speech.

The possible advantage of bilingual children over monolinguals in analyzing word meanings from verbal context was examined. The subjects were 40 third-grade children (20 bilingual and 20 monolingual) recruited from independent schools in the USA. The two groups of participants were compared on their performance on a standardized test of receptive vocabulary and an experimental measure of word meanings, the Word–Context Test. Results revealed that on average, the bilingual children had smaller vocabularies in English. The bilinguals deduced the meaning from context of more words than the monolingual children, although there were no differences between groups on the rate of reaching the target meanings for words on which they were successful, and on the quality of their definitions. Moreover, bilingual children approached the task differently and they showed greater flexibility when analyzing word meanings from verbal context, thus indicating that bilinguals may be more efficient vocabulary learners than monolinguals.

In two experiments, we tested the role of lexical repetition, cognates, and second language (L2) proficiency in the priming of code-switches, using the structural priming technique. Dutch–English bilinguals repeated a code-switched prime sentence (starting in Dutch and ending in English) and then described a target picture by means of a code-switched sentence (also from Dutch into English). Low- and high-proficient speakers of L2 English were tested in Experiments 1 and 2, respectively. We found that the participants’ tendency to switch at the same position as in the prime sentence was influenced by lexical repetition between prime sentence and target picture and by the presence of a cognate in prime and target. A combined analysis showed that these lexical effects were stronger in the high-proficient than in the low-proficient L2 speakers. These results provide new insights into how language-related and speaker-related variables influence code-switching in sentences, and extend cognitive models of lexical and combinatorial processes in bilingual sentence production.

Many attempts have been made to teach native Japanese listeners to perceptually differentiate English /r–l/ (e.g. rock–lock). Though improvement is evident, in no case is final performance native English-like. We focused our training on the third formant onset frequency, shown to be the most reliable indicator of /r–l/ category membership. We first presented listeners with instances of synthetic /r–l/ stimuli varying only in F3 onset frequency, in a forced-choice identification training task with feedback. Evidence of learning was limited. The second experiment utilized an adaptive paradigm beginning with non-speech stimuli consisting only of /r/ and /l/ F3 frequency trajectories progressing to synthetic speech instances of /ra–la/; half of the trainees received feedback. Improvement was shown by some listeners, suggesting some enhancement of /r–l/ identification is possible following training with only F3 onset frequency. However, only a subset of these listeners showed signs of generalization of the training effect beyond the trained synthetic context.

This study examined temporal reference marking in texts written by Dutch deaf children and adults who differed in sign language proficiency. The temporal reference marking systems in Dutch and Sign Language of the Netherlands (SLN) differ substantially, with Dutch having a wide range of lexical and morphological markers of temporal reference, and SLN relying on lexical marking of temporal reference. The results showed that the youngest proficient signers had difficulties with tense morphology: they avoided the marked past tense form in narratives and omitted verbs, but showed no problems with lexical marking of temporal reference. In the older proficient signing writers, verb morphology emerged, and in proficient signing adults temporal reference marking resembled that of the hearing adults. This study shows that in order to gain more insight into deaf people's writing, it is important to adopt a bilingual perspective and take variations in sign language proficiency into account.

Although all bilinguals encounter cross-language interference (CLI), some bilinguals are more susceptible to interference than others. Here, we report on language performance of late bilinguals (Russian/German) on two bilingual tasks (interview, verbal fluency), their language use and switching habits. The only between-group difference was CLI: one group consistently produced significantly more errors of CLI on both tasks than the other (thereby replicating our findings from a bilingual picture naming task). This striking group difference in language control ability can only be explained by differences in cognitive control, not in language proficiency or language mode.

The purpose of this study was to conduct an acoustic examination of the obstruent stops produced by Korean–English bilingual children in connection with the question of whether bilinguals establish distinct categories of speech sounds across languages. Stop productions were obtained from ninety children in two age ranges, five and ten years: thirty Korean–English bilinguals, thirty monolingual Koreans and thirty monolingual English speakers. Voice-Onset-Time (VOT) lag at word-initial stop and fundamental frequency (f0) in the following vowel (hereafter vowel-onset f0) were measured. The bilingual children showed different patterns of VOT in comparison to both English and Korean monolinguals, with longer VOT in their production of Korean stop consonants and shorter VOT for English. Moreover, the ten-year-old bilinguals distinguished all stop categories using both VOT and vowel-onset f0, whereas the five-year-olds tended to make stop distinctions based on VOT but not vowel-onset f0. The results of this study suggest that bilingual children at around five years of age do not yet have fully separate stop systems, and that the systems continue to evolve during the developmental period.

It remains unclear whether adult-learned second language (L2) depends on similar or different neurocognitive mechanisms as those involved in first language (L1). We examined whether English past tense forms are computed similarly or differently by L1 and L2 English speakers, and what factors might affect this: regularity (regular vs. irregular verbs), length of L2 exposure (length of residence), age of L2 acquisition (age of arrival), L2 learners’ native language (Chinese vs. Spanish), and sex (male vs. female). Past tense frequency effects were used to examine the type of computation (composition vs. storage/retrieval). The results suggest that irregular past tenses are always stored. Regular past tenses, however, are either composed or stored, as a function of various factors: both sexes store regulars in L2, but only females in L1; greater lengths of residence lead to less dependence on storage, but only in females; higher adult ages of arrival lead to more reliance on storage. The findings suggest that inflected forms can rely on either the same or different mechanisms in L2 as they do in L1, and that this varies as a function of multiple interacting factors.