Bilingual Task Research Articles

Oscillatory features of German–English sentence processing: Evidence from an EEG study

Aims and objectives: In this study, we investigated the neural oscillatory patterns occurring during sentence-level overt bilingual processing tasks on 46 German–English bilinguals with the intention of finding out if there were task-specific oscillations for this language pair. Methodology: Four tasks, including Language 2 (L2) listening, Language 1 (L1) speaking, L2 shadowing (i.e., repeating a segment of an utterance), and backward interpreting (i.e., orally translating L2 sentences to L1), were implemented with the technique of electroencephalography (EEG). Data and analysis: Data included demographic indices, neural oscillation metrics, and oral production scores from participants. Static spectral analysis, time–frequency analysis, principal components analysis, cross-frequency coupling, and inferential statistics were applied to the data. Findings: The sensory task and articulatory tasks differed greatly in the delta–theta band and the beta–gamma band in terms of time–frequency dynamics. Analyses also revealed variations in a frontal/prefrontal theta component and a distinct gamma component across the four tasks. Moreover, we found a right fronto-temporal theta–gamma coupling unique to backward interpreting. Originality: The task-wise comparison experiment paradigm has rarely been implemented in neuroimaging research on language, especially so in EEG research on bilingualism. Significance: The findings provide novel insights for the understanding of language conversion mechanism and may benefit the training of professional interpreters.

Bilingual phrase induction with local hard negative sampling

AbstractBilingual lexicon induction focuses on learning word translation pairs, also known as bitexts, from monolingual corpora by establishing a mapping between the source and target embedding spaces. Despite recent advancements, bilingual lexicon induction is limited to inducing bitexts consisting of individual words, lacking the ability to handle semantics‐rich phrases. To bridge this gap and support downstream cross‐lingual tasks, it is practical to develop a method for bilingual phrase induction that extracts bilingual phrase pairs from monolingual corpora without relying on cross‐lingual knowledge. In this paper, the authors propose a novel phrase embedding training method based on the skip‐gram structure. Specifically, a local hard negative sampling strategy that utilises negative samples of central tokens in sliding windows to enhance phrase embedding learning is introduced. The proposed method achieves competitive or superior performance compared to baseline approaches, with exceptional results recorded for distant languages. Additionally, we develop a phrase representation learning method that leverages multilingual pre‐trained language models. These mPLMs‐based representations can be combined with the above‐mentioned static phrase embeddings to further improve the accuracy of the bilingual phrase induction task. We manually construct a dataset of bilingual phrase pairs and integrate it with MUSE to facilitate the bilingual phrase induction task.

Development of language competence of future defectologists in the context of bilingualism

Relevance. The development of special education in Kazakhstan is associated with legislation in the education sector, the country's policy in relation to children with special needs, the level of development of special (correctional) science as an integrated (psychological, pedagogical, medical) field of knowledge. The system of higher education in Kazakhstan is forced to revise the goals and strategies for the development of professional education of future special education teachers. One of the main tasks of which is the transition from the qualification training of a specialist to competence based, by changing the training strategy, where the student learns to achieve the set goal, work for the intended result, i.e. the development of professional competencies is expected. Purpose. The purpose of the study is the scientific and theoretical substantiation of the need to develop the language competence of future defectologists with the consideration of bilingualism and educational and methodological support of the process. Methodology. The fundamental method of the research was pedagogical design, during which a pedagogical model of the language competence development in future special education teachers was designed. Within the framework of a comprehensive program for the development of the language competence in special education teachers, the following methods have been developed: an elective course �Fundamentals of the speech culture of a defectologist�; project complexes, bilingual creative tasks, exercises; methodological measures. Results. During the approbation of the developed model, it was revealed that the introduction of the proposed active methods, means and forms of teaching into the language teaching process can be an effective way to solve the issue under consideration. Conclusions. Theoretical substantiation and the results of experimental work in the future can be used in the accumulation of content and the development of educational recommendations, teaching aids and methods of teaching languages. Keywords: special education; language peculiarity; communication; specialist

Audio-visual Stroop matching task with first- and second-language color words and color associates

AbstractIn the audio-visual Stroop matching task, participants compare one Stroop stimulus dimension (e.g., the color of a written word) to a second stimulus (e.g., a spoken word) and indicate whether these two stimuli match or mismatch. Slower responses on certain trials can be due to conflict which occurs between color representations (semantic conflict) or due to conflict between responses evoked by task comparisons (response conflict). The contribution of these conflicts has been investigated with color word distracters. This is the first study which explores how two types of first- and second-language words affect audio-visual matching. Native French speakers performed a bilingual Stroop matching task with intermixed French (L1) and English (L2) color words (Experiment 1) and color associates (Experiment 2) presented in congruent and incongruent colors simultaneously with spoken French color words. Participants were instructed to indicate whether the spoken word “matches” or “mismatches” the font color, while ignoring written word meaning. Interestingly, the results were similar for the critical “mismatch” trials for both French and English words. The responses were the fastest on trials in which task comparisons activate fewer response alternatives, supporting the assumption of the response conflict account.

The Impact of Interpreting Training Experience on the Attentional Networks and Their Dynamics.

Interpreting, a complicated and demanding bilingual task, depends heavily on attentional control. However, few studies have focused on the interpreters' advantages in attention, and the findings so far have been inconsistent. Meanwhile, the connection between attentional networks and other cognitive abilities, such as working memory (WM), has rarely been explored in interpreters. The present study investigated whether interpreting experience (IE) contributed to the attentional networks of bilinguals and explored the link between interpreters' attention and WM. Three groups of Chinese-English bilinguals, differing only in their duration of interpreting training (the More-IE group, the Less-IE group, and the No-IE group), completed the Attention Network Test (ANT). Results showed that only the alerting network was more efficient in the More-IE group than in the Less-IE and No-IE groups; moreover, the dynamics between the alerting and executive networks were significant only in the More-IE group. Furthermore, we found a negative correlation between the executive effect and the working memory capacity (WMC) in the More-IE group. Our study validated and provided empirical support for the Attentional Control Model, stimulating further research into neurocognitive mechanisms of advanced second language learning.

Phonetic Dissimilarity and L2 Category Formation in L2 Accommodation.

Many studies of speech accommodation focus on native speakers with different dialects, whereas only a limited number of studies work on L2 speakers' accommodation and discuss theories for second language (L2) accommodation. This paper aimed to fill the theoretical gap by integrating the revised speech learning model (SLM) with the exemplar-based models for L2 speech accommodation. A total of 19 Cantonese-English bilingual speakers completed map tasks with English speakers of Received Pronunciation and General American English in two separate experiments. Their pronunciations of THOUGHT and PATH vowels, and fricatives [z] and [θ] were examined before, during, and after the map tasks. The role of phonetic dissimilarity in L2 accommodation and L2 category formation in the revised SLM (SLM-r) were tested. First, the results suggested that global phonetic dissimilarity cannot predict Hong Kong English (HKE) speakers' accommodation patterns. Instead, the segment-specific phonetic dissimilarity between participants and interlocutors was found to be positively correlated with the participants' degree of accommodation. In addition, HKE speakers who did not form a new L2 category of [z] were found to significantly accommodate toward their interlocutor, suggesting that L2 accommodation might not be constrained by phonological category. An integrated exemplar model for L2 accommodation is proposed to explain these findings.

Cross-language interference in bilinguals with neurodegenerative disorders

AbstractThe aim of the study is to explore the differential effects of neurodegenerative disorders on cross-language interference and facilitation, by comparing the performance in a bilingual Stroop task of Catalan–Spanish bilinguals with different neurodegenerative disorders [Mild Cognitive Impairment (MCI), n = 16; Alzheimer's disease (AD), n = 16; and Parkinson's disease (PD), n = 16)] and healthy controls (n = 14). Interference is the difference in naming latencies between the incongruent (colour words not matching with their ink colour) and control (non-colour words) conditions. Facilitation is the difference between the control and congruent conditions (colour words matching with their ink colour).The results showed that AD and MCI patients, but not PD, showed a larger interference effect than healthy controls. Nevertheless, the patient groups showed the same facilitation effect as healthy controls.The contribution of cortical and subcortical brain areas to cross-language interference suppression is discussed in relation to the type of brain degeneration.

Learning bilingual word embedding for automatic text summarization in low resource language

Studies in low-resource languages have become more challenging with the increasing volume of texts in today's digital era. Also, the lack of labeled data and text processing libraries in a language further widens the research gap between high and low-resource languages, such as English and Indonesian. This has led to the use of a transfer learning approach, which applies pre-trained models to solve similar problems, even in different languages by using bilingual or cross-lingual word embedding. Therefore, this study aims to investigate two bilingual word embedding methods, namely VecMap and BiVec, for Indonesian – English language and evaluates them for bilingual lexicon induction and text summarization tasks. The generated bilingual embedding was compared with MUSE (Multilingual Unsupervised and Supervised Embeddings) as the existing multilingual word created with the generative adversarial network method. Furthermore, the VecMap was improved by creating shared vocabulary spaces and mapping the unshared ones between languages. The result showed the embedding produced by the joint methods of BiVec, performed better in intrinsic evaluation, especially with CSLS (Cross-Domain Similarity Local Scaling) retrieval. Meanwhile, the improved VecMap outperformed the regular type by 16.6% without surpassing the BiVec evaluation score. These methods enabled model transfer between languages when applied to cross-lingual-based text summarization. Moreover, the ROUGE score outperformed classical text summarization by adding only 10% of the training dataset of the target language.

Bilingual lips: Motor activation for second language word processing

Aims and objectives: Current research suggests that motor articulatory representations are relevant for optimal speech processing under difficult circumstances. The challenging situation of a second language (L2) listener mimics this sub-optimal condition; therefore, activation of motor brain regions to support L2 processing is expected. This study aimed to investigate if the motor system associated with speech articulators exerts an influence on processing bilingual acoustic inputs. Methodology: Thirty proficient bilinguals performed twice on a word-to-picture matching task: (1) while producing constant contraction of the orbicularis oris (OO) muscle interfering with the articulator’s movement and (2) while producing constant contraction of the first dorsal interosseus (index finger) muscle. Data and analysis: Accuracy and reaction times (RTs) were collected from the bilingual task during the experimental and control conditions. Two-way repeated measures analyses of variance were conducted to assess if constant contraction of articulators yielded a significant effect on word recognition in L1 and L2. Findings: Word recognition was significantly slower in L2 compared with the native language (L1) during articulator contraction. Task accuracy in L1 was marginally better during the finger muscle contraction compared with OO muscle contraction, and no difference in accuracy was found in L2. An interesting distinct speed-accuracy trade-off strategy for L1 and L2 was observed. The findings support the hypothesis of a motor system facilitatory effect on processing acoustic inputs in bilinguals. Originality: The role of motor components in language processing has been studied in challenging linguistic environments, but little has been done to identify its role in L2 processing. This study used an innovative behavioral strategy to interfere with articulatory muscles during word recognition. Significance: Our results provide evidence of motor system processing support to L2 word recognition. In addition, a bilingual tendency to sacrifice speed for accuracy in L2 compared with L1 is suggested.

Language switching and domain-general control in interpreters

The purpose of this paper is to examine whether and how training and professional experience in interpreting affect task switching in this bilingual population. In the first experiment, we compared a group of interpreting students to a group of translation students using the bilingual categorization task to assess their domain-specific language switching before and after training. In the second experiment, we added a group of professional interpreters to the participants in experiment 1 to test prepotent response inhibition using the Simon task (domain-general). First, the results showed training-related improvement in the bilingual categorization task in both student groups, indicating a similar effect for translation and interpreting training. Second, both student groups showed better performance on the Simon task compared to professional interpreters, but only on response times and not on accuracy. The correlation analyses of the two tasks in student groups only showed significant correlations between the global RTs and supported the hypothesis that proactive language control may depend more on inhibition than on the switching-specific factor. Considering language background, the lower onset age of L2 acquisition (AOA2) in the interpreting students (compared to the translation students) was significantly correlated with the congruency effect in the Simon task, indicating an impact of language background on domain-general control. Results were discussed in light of the different engaging elements, including task specificity, training length, research method, and participants' linguistic profile.

The relationship between bilingual language control and language dominance: An empirical study of visual language perception

Bilingual visual perception is an under-researched area in psycholinguistics and has yielded contradictory results regarding language control. Two theories were developed to account for visual language perception in bilinguals – the Inhibitory Control model (Green, 1986) and the Bilingual Interactive Activation model (Grainger and Dijkstra, 1992). Even though these two accounts have opposite predictions for asymmetrical language control (different for the bilingual’s L1 and L2), most research up to date found evidence only for symmetrical control (Macizo et al., 2012; Orfanidou and Summer, 2005). This study aims to investigate the influence of language dominance on the visual language perception of bilinguals and providing evidence for the accountability of the models mentioned above. Thirty-one Hungarian-English bilinguals were recruited for this study. Participants’ language dominance was operationalized by a complex score using the Bilingual Language Profile questionnaire (Birdsong et al., 2012). The bilingual lexical decision task was used to investigate the differences in the cognitive processing of the two languages. Target stimuli were carefully matched for their visual and linguistic features to eliminate the potential confounding influences on their perception during task performance. For unbalanced, L1 dominant bilinguals, asymmetrical, dominance-related switching cost was observed, indicating the relevance of the Inhibitory Control model. Faster L2 processing correlated with a richer history of L2-associated experiences and more balanced bilingualism. However, no correlations were found with the frequency of language use, language attitudes, and only weak correlations were observed with language proficiency. The current research proposes a methodological framework for measuring the influence of linguistic background on language switching cost that could ensure comparability between further studies.

Electroencephalographic oscillations of alpha and beta rhythms during phrase-guessing procedure.

Phrases-guessing is one of the essential reasoning abilities in problem solving for human beings. However, it is still an open question about why individuals perform differently during the same reasoning task. In this study, we utilized a bilingual phrase-guessing task to explore the neural activities under the individually different performances with electroencephalography. Participants who had no knowledge of Greek were required to guess the meaning of a Greek phrase (long or short in length) by making an either-or selection as to which translation-equivalent Chinese word corresponds to Greek word. Names of color were used as experimental stimuli for which two Chinese words denoted the same color with one as a conventional color name and the other as a novel color name. The experiment yielded length of phrases (long vs. short) and novelty of phrases (novel vs. conventional) as variables. The behavioral results revealed significant length-by-novelty interaction on the number of selections. However, neither main effects nor interactive effects were found on response time. Further, the amplitude spectrums of high alpha rhythm, low alpha rhythm, and low beta rhythm during the task were positively associated with the participants' number of selections for a long Greek phrase with a novel and complex Chinese phrase (LNc) and a short Greek phrase with a conventional Chinese phrase (SCo), while negatively correlated with the response time of selections for LNc and SCo. Our findings suggested that the consistency between participants' behavior and electrophysiological oscillations (alpha and beta bands) could be employed as biomarkers for decoding the phrase-guessing procedure.

Interlinguistic conflict: Word-word Stroop with first and second language colour words.

The congruency (or Stroop) effect is a standard observation of slower and less accurate colour identification to incongruent trials (e.g. "red" in green) relative to congruent trials (e.g. "red" in red). This effect has been observed in a word-word variant of the task, when both the distracter (e.g. "red") and target (e.g. "green") are colour words. The Stroop task has also been used to study the congruency effect between two languages in bilinguals. The typical finding is that the congruency effect for L1 words is larger than that for L2 words. For the first time, the present report aims to extend this finding to a word-word variant of the bilingual Stroop task. In two experiments, French monolinguals performed a bilingual word-word Stroop task in which target word language, language match, and congruency between the distracter and target were manipulated. The critical manipulation across two experiments concerned the target language. In Experiment 1, target language was manipulated between groups, with either French (L1) or English (L2) target colour words. In Experiment 2, target words from both languages were intermixed. In both experiments, the congruency effect was larger when the distracter and target were from the same language (language match) than when they were from different languages (language mismatch). Our findings suggested that this congruency effect mostly depends on the language match between the distracter and target, rather than on a target language. It also did not seem to matter whether the language-mismatching distracter was or was not a potential response alternative. Semantic activation of languages in bilinguals and its implications on target identification are discussed.

Tracking lexical access and code switching in multilingual participants with different degrees of simultaneous interpretation expertise.

With the worldwide increase in people speaking more than one language, a better understanding of the behavioural and neural mechanisms governing lexical selection, lexical access in multiple languages and code switching has attracted widespread interest from several disciplines. Previous studies documented higher costs when processing a non‐native (L2) than a native (L1) language or when switching from L2 to L1. However, studies on auditory language reception are still scarce and did not take into account the degree of switching experience. Accordingly, in the present study, we combined behavioural and electrophysiological measurements to assess lexical access in L1 and L2 as well as code switching in professional simultaneous interpreters, trainee interpreters, foreign language teachers and Anglistics students, while the participants performed a bilingual auditory lexical decision task. The purpose of this study was to expand the knowledge on code switching in auditory language processing and examine whether the degree of simultaneous interpretation experience might reduce switching costs. As a main result, we revealed that L2 compared to L1 trials, as well as switch compared to non‐switch trials, generally resulted in lower accuracies, longer reaction times and increased N400 amplitudes in all groups of participants. Otherwise, we did not reveal any influence of switching direction and interpretation expertise on N400 parameters. Taken together, these results suggest that a late age of L2 acquisition leads to switching costs, irrespective of proficiency level. Furthermore, we provided first evidence that simultaneous interpretation training does not diminish switching costs, at least when focusing on lexical access.

From unified phrase representation to bilingual phrase alignment in an unsupervised manner

AbstractSignificant advances have been achieved in bilingual word-level alignment, yet the challenge remains for phrase-level alignment. Moreover, the need for parallel data is a critical drawback for the alignment task. This work proposes a system that alleviates these two problems: a unified phrase representation model using cross-lingual word embeddings as input and an unsupervised training algorithm inspired by recent works on neural machine translation. The system consists of a sequence-to-sequence architecture where a short sequence encoder constructs cross-lingual representations of phrases of any length, then an LSTM network decodes them w.r.t their contexts. After training with comparable corpora and existing key phrase extraction, our encoder provides cross-lingual phrase representations that can be compared without further transformation. Experiments on five data sets show that our method obtains state-of-the-art results on the bilingual phrase alignment task and improves the results of different length phrase alignment by a mean of 8.8 points in MAP.

Bilingual Stroop Effect in High and Low Proficient Nepali-English Bilinguals

This research article presents experimental psycholinguistic study on high and low proficient Nepali – English bilingual adults. Language comprehension was measured using lexical decision tasks and language production was measured using verbal fluency task. Executive control was measured using bilingual Stroop task in mouse tracking paradigm. The results show high proficient bilinguals show the bilingual advantage in language comprehension, language production and executive control tasks.

Language lexicons for Hindi-English multilingual text processing

<span lang="EN-US">Language identification (LI) in textual documents is the process of automatically detecting the language contained in a document based on its content. The present language identification techniques presume that a document contains text in one of the fixed set of languages. However, this presumption is incorrect when dealing with multilingual document which includes content in more than one possible language. Due to the unavailability of standard corpora for Hindi-English mixed lingual language processing tasks, we propose the language lexicons, a novel kind of lexical database that augments several bilingual language processing tasks. These lexicons are built by learning classifiers over English and transliterated Hindi vocabulary. The designed lexicons possess condensed quantitative characteristics which reflect their linguistic strength in respect of Hindi and English language. On evaluating the lexicons, it is observed that words of the same language tend to cluster together and are separable over language classes. On comparing the classifier performance with existing works, the proposed lexicon models exhibit the better performance.</span>

Cross-Lingual Embeddings Using a Temporally Aligned Comparable Corpus: A Case Study For Manipuri–English

Cross-lingual embeddings facilitate cross-language learning, bridging the gap between rich-resource and low-resource languages. This study provides and assesses unsupervised cross-lingual embeddings generation methods for the low-resource Manipuri–English language pair. Manipuri is a resource-poor language spoken in India’s northeastern regions. The embeddings are evaluated on the language pair bilingual dictionary induction task. Furthermore, we propose a method to improve the cross-lingual embeddings by exploiting a temporally aligned comparable corpus. Lack of supervision has always been an issue for learning models, especially in low-resource settings. The proposed method takes advantage of the temporal alignments and provides the much-needed supervision to improve the alignment between Manipuri and English language pair. We observe that the proposed model consistently outperforms all the corresponding baselines from various experimental results.

Language switch costs in a lexical decision task: Symmetry and cognitive correlates

Language switch costs have been explored less in receptive tasks than in productive tasks, and previous studies have produced mixed findings with regard to switch cost symmetry and the relationship of switch costs to executive function. To address these unresolved gaps, one hundred Chinese–English bilingual adults completed a bilingual lexical decision task and three tasks measuring executive function, and we used mixed effects models and correlational analyses to answer the two research questions. The results showed asymmetry with larger costs into the second language, but this was qualified by interactions with response sequence effects. No evidence was found for a relationship between switch costs and inhibition or shifting. Together, rather than supporting a model involving top-down control mechanisms as has been suggested to account for switch cost patterns in productive tasks, these findings support a bottom-up, activation-based model of bilingual word recognition and receptive language switching.

Establishing Diagnostic Skills in Novice Bilingual Clinicians: A Scaffolded Approach

This study sought to scaffold administration performance of a standardized bilingual screener to sufficient levels of accuracy for data collection using principles of Cognitive Load Theory by managing task complexity when training pre-service clinicians. Before training administration skills, two students were given copies of the manual for the Bilingual English Spanish Oral Screener (BESOS) and asked to administer the protocol independently. During the intervention phase, students were scaffolded through administration tasks of increasing complexity and given explicit instruction, which included tailored goals, modeling and feedback. Performance for four skills was assessed using a fidelity rubric and analyzed using visual analysis. Performance varied per skill but overall scores were higher during the intervention phases than during the baseline phase for both students. In addition, accuracy of performance maintained across client participants showing patterns of generalization. Although the data are limited, scaffolding training skills for pre-service clinicians appears supportive in training administration skills for bilingual tasks. The level of support may vary per skill and per language. Future research may seek to investigate other clinical skills and tasks.