Task Schema Research Articles

Tracking Components of Bilingual Language Control in Speech Production: An fMRI Study Using Functional Localizers.

When bilingual speakers switch back to speaking in their native language (L1) after having used their second language (L2), they often experience difficulty in retrieving words in their L1. This phenomenon is referred to as the L2 after-effect. We used the L2 after-effect as a lens to explore the neural bases of bilingual language control mechanisms. Our goal was twofold: first, to explore whether bilingual language control draws on domain-general or language-specific mechanisms; second, to investigate the precise mechanism(s) that drive the L2 after-effect. We used a precision fMRI approach based on functional localizers to measure the extent to which the brain activity that reflects the L2 after-effect overlaps with the language network (Fedorenko et al., 2010) and the domain-general multiple demand network (Duncan, 2010), as well as three task-specific networks that tap into interference resolution, lexical retrieval, and articulation. Forty-two Polish-English bilinguals participated in the study. Our results show that the L2 after-effect reflects increased engagement of domain-general but not language-specific resources. Furthermore, contrary to previously proposed interpretations, we did not find evidence that the effect reflects increased difficulty related to lexical access, articulation, and the resolution of lexical interference. We propose that difficulty of speech production in the picture naming paradigm-manifested as the L2 after-effect-reflects interference at a nonlinguistic level of task schemas or a general increase of cognitive control engagement during speech production in L1 after L2.

Transfer of strategic task components across unique tasks that share some common structures.

Flexible, adaptive behaviour depends on the application of prior learning to novel contexts (transfer). Transfer can take many forms, but the focus of the present study was on "task schemas"-learning strategies that guide the earliest stages of engaging in a novel task. The central aim was to examine the architecture of task schemas and determine whether strategic task components can expedite learning novel tasks that share some structural components with the training tasks. Groups of participants across two experiments were exposed to different training regimes centred around multiple unique tasks that shared some/all/none of the structural task components (the kinds of stimuli, classifications, and/or responses) but none of the surface features (the specific stimuli, classifications, and/or responses) with the test task (a dot-pattern classification task). Initial test performance was improved (to a degree) in all groups relative to a control group whose training did not include any of the structural components relevant to the test task. The strongest evidence of transfer was found in the motoric, perceptual + categorization, and full schema training groups. This observation indicates that training with some (or all) strategic task components expedited learning of a novel task that shared those components. That is, task schemas were found to be componential and were able to expedite learning a novel task where similar (learning) strategies could be applied to specific elements of the test task.

Improving task generalization via unified schema prompt

Task generalization has been a long-standing challenge in Natural Language Processing (NLP). Recent research attempts to improve the task generalization ability of pre-trained language models by mapping NLP tasks into human-readable prompted forms. However, these approaches require laborious and inflexible manual collection of prompts, and different prompts on the same downstream task may receive unstable performance. We propose Unified Schema Prompt, a flexible and extensible prompting method, which automatically customizes the learnable prompts for each task according to the task input schema. It models the shared knowledge between tasks, while keeping the characteristics of different task schema, and thus enhances task generalization ability. The schema prompt takes the explicit data structure of each task to formulate prompts so that little human effort is involved. To test the task generalization ability of schema prompt at scale, we conduct schema prompt-based multitask pre-training on a wide variety of general NLP tasks. The framework achieves strong zero-shot and few-shot generalization performance on 16 unseen downstream tasks from 8 task types (e.g., QA, NLI, etc.). Furthermore, comprehensive analyses demonstrate the effectiveness of each component in the schema prompt, its flexibility in task compositionality, and its ability to improve performance under a full-data fine-tuning setting.

First language translation involvement in second language word processing.

Studies on bilingual word processing have demonstrated that the two languages in a mental lexicon can be parallelly activated. However, it is under discussion whether the activated, non-target language gets involved in the target language. The present study aimed to investigate the role of the first language (L1, the non-target one) translation in the second language (L2, the target one) word processing. The tasks of semantic relatedness judgment and lexical decision were both adopted, to explore the relation of the possible L1 involvement and the task demand. Besides, bilinguals with relatively higher and lower L2 proficiency were recruited, to clarify the potential influence of L2 proficiency. Results showed that the manipulation of L1 translation exerted an influence on bilinguals’ task performances, indicating that L1 translation was involved, but did not just serve as a by-product when bilinguals were processing L2 words. And about the influence of L2 proficiency, the higher proficiency bilinguals performed better than the lower proficiency ones when the L1 translations could be taken advantage of, indicating a better access to L1 translation in L2 word processing, as bilinguals’ L2 proficiency increased. As for the task demands, the L1 translation was partially involved in Experiment 1 while a full involvement was observed in Experiment 2, suggesting a differed depth of L1 translation involvement, if the task demands allowed. The present study supplemented the previous ones due to its participants (the intermediate bilinguals) and tasks (the tasks of semantic relatedness judgment and lexical decision); besides, it provided an interesting view into interpreting the “task schema” of the BIA+ model.

An instance-based model account of the benefits of varied practice in visuomotor skill

Exposing learners to variability during training has been demonstrated to improve performance in subsequent transfer testing. Such variability benefits are often accounted for by assuming that learners are developing some general task schema or structure. However much of this research has neglected to account for differences in similarity between varied and constant training conditions. In a between-groups manipulation, we trained participants on a simple projectile launching task, with either varied or constant conditions. We replicate previous findings showing a transfer advantage of varied over constant training. Furthermore, we show that a standard similarity model is insufficient to account for the benefits of variation, but, if the model is adjusted to assume that varied learners are tuned towards a broader generalization gradient, then a similarity-based model is sufficient to explain the observed benefits of variation. Our results therefore suggest that some variability benefits can be accommodated within instance-based models without positing the learning of some schemata or structure.

The effects of an augmented-reality ubiquitous writing application: a comparative pilot project for enhancing EFL writing instruction

The use of technology such as online software has been examined in English as a foreign language (EFL) writing contexts. However, few studies have incorporated context-aware ubiquitous technology into EFL writing instruction. To develop multimodal and digital literacy of target EFL undergraduates in this pilot project, the researchers designed and implemented an augmented-reality context-aware ubiquitous writing (ARCAUW) application, which aimed to increase long-term memory, motivation, and self-regulated cognition in participants’ writing development. The pilot project compared the writing outcomes and learner perceptions of the proposed ARCAUW writing mode against the baseline mobile-assisted, classroom-based writing mode. The two writing modes differed in the way that metacognitive scaffolding tasks were carried out. Pre- and post-test results showed that (a) although both modes led to significant improvement in writing the process analysis essay, ARCAUW was conducive to the development of task schema in long-term memory, motivation, and self-regulation in writing, and (b) additional cognitive processing during AR-based learning led to mixed results in writing performance. It was concluded that ARCAUW should be incorporated into mobile-assisted writing courses to reach optimal instructional outcomes. Finally, a five-step procedure was offered to help EFL practitioners design and implement ubiquitous writing in their own teaching contexts.

Attentional control in interpreting: A model of language control and processing control

AbstractInterpreting is a complex bilingual task, placing high demands on both language control (i.e., source language not interfering in target language production) and processing control (i.e., multi-tasking carried out in concert under time pressure). On the basis of empirical evidence in the literature, we propose an attentional control model to account for both language control and processing control. Specifically, language control in interpreting is achieved by a structural framework of language-modality connections (established in interpreting training and stored as task schema), and by focused attention that helps build, strengthen and adapt the framework through monitoring, target enhancement, task disengagement, shifting, and working memory. In contrast, processing control in interpreting is achieved by divided attention via coordination and working memory, and by language processing efficiency that includes mastery of both languages and the appropriate use of interpreting strategies. Implications of this model for general bilingual language control are discussed.

Transcranial direct current stimulation influences bilingual language control mechanism: evidence from cross-frequency coupling.

How to better suppress the interference from the non-target language when switching from one language to the other in bilingual production? The current study applied transcranial direct current stimulation over the right dorsolateral prefrontal cortex to modulate language control measured by cross-frequency coupling. We found that switching to L2 was more modulated by F4-F3 alpha-beta phase-amplitude compared to switching to L1 after receiving the anodal stimulation at the language task schema phase. These findings suggest that anodal stimulation affects the selection of the target language task schema by enhancing the activation of frontal areas and facilitating the coordination between the left and the right frontal hemispheres.

Designing a Revision Tool to Distinguish Surface-level and Text-based Writing Feedback

This paper presents a design case that describes the design, development, and user experience testing of a Google Docs revision add-on. The add-on is an instructional, peer review tool intended to help students distinguish surface-level feedback from text-based feedback in order to develop their revision task schema. Eleven secondary teachers completed a survey about using the add-on for instructional purposes, and 56 secondary students completed a survey after using the tool to provide feedback to a peer’s writing and make changes to their writing based on feedback provided to them through the tool. Thematic analyses revealed recommendations for modifications and additions to the tool. Next steps include researching the effects of the updated add-on on secondary students’ revision task schema development.

Effects of Naming Language and Switch Predictability on Switch Costs in Bilingual Language Production.

Switch costs are defined as the phenomenon that bilinguals have worse performance in switch trials relative to non-switch trials. Bilinguals’ naming language and switch predictability have been found to influence the magnitude of switch costs. However, how these two factors modulate switch costs in different phases (i.e., lemma activation and language selection) during language production remains unclear. Most previous studies using the language switching paradigm did not dissociate lemma activation from language selection, because the language cue was either presented simultaneously with or prior to a stimulus. Therefore, here we modified the language switching paradigm by presenting a digit stimulus prior to a visual cue. This allowed us to dissociate lemma activation from language selection, and thus we were able to investigate the mechanisms underlying the effects of naming language and switch predictability on switch costs during the two different phases in language production. Unbalanced Indonesian-Chinese bilinguals were required to name digits in either their L1 (Indonesian) or L2 (Chinese), and their reaction times and electrophysiological responses were recorded. The behavioral results showed the effects of switch predictability on switch costs, with responses in switch trials being slower than those in non-switch trials in the low switch predictability condition, while there was no significant difference in response times between switch trials and non-switch trials in the high switch predictability condition. The event-related potential results showed that neither naming language nor switch predictability affected switch costs during the lemma activation phase, but both did so during the language selection phase, particularly at the language task schema competition stage. The results imply that naming language and switch predictability affect switch costs mainly during the language task schema competition stage.

The plasticity of lexical selection mechanism in word production: ERP evidence from short-term language switching training in unbalanced Chinese–English bilinguals

The present study examined the plasticity of the lexical selection mechanism in bilingual word production by training a group of unbalanced Chinese–English bilinguals with a language switching task. The experimental group received an 8-day language switching training, while the control group received no training. Before and after training, the behavioral and event-related potential (ERP) data of both groups in a cued picture naming task were collected. ERP results revealed a training effect such that after training, the N2 peak latency in cue-locked ERPs was shortened only in the experimental group. These results suggest that short-term language switching experience could improve the efficiency to establish the target language task schema, and that the language control mechanism of word production in unbalanced bilinguals could be modulated by language switching experience.

An fMRI study of executive control during translation and oral reading in Cantonese-English bilingual speakers

An fMRI study of executive control during translation and oral reading in Cantonese-English bilingual speakers

Switch Costs Occur at Lemma Stage When Bilinguals Name Digits: Evidence from Language-Switching and Event-Related Potentials.

Switch costs are generally found in language switching tasks. However, the locus where switch costs occur during bilingual language production remains unclear. Several studies that used a cued language-switching paradigm have attempted to investigate this question in bilingual language production, but researchers have not reached a consensus. Moreover, we are interested in where switch costs occur when language selection occurs after lemma activation. Previous studies have not investigated this question because most previous studies presented language cues before or along with the stimuli. Therefore, we used a modified cued language-switching paradigm with a combined event-related potentials (ERPs) technique to explore the locus of switch costs during bilingual language production. The cue and stimulus were separated and presented in two different presentation sequences in which Indonesian–Chinese bilingual speakers were instructed to name digits in their L1 or L2 according to the color of the cue. The ERPs related to the cue and stimulus for two presentation sequences were measured. In the stimulus-cue sequence, the analysis that was time-locked to cues revealed a reversed switch cost as early as 220 ms after the cue onset; furthermore, a switch cost was shown in L1 with a late stage post-cue onset. The results suggested that when language selection occurred after lemma activation, the switch costs mainly occurred at the lemma selection stage. In the cue-stimulus sequence, the analysis that was time-locked to cues did not reveal significant main effects of switching, whereas the analysis that was time-locked to digits yielded a switch cost, again indicating that switch costs mainly occurred at the lemma selection stage rather than at the language task schema competition stage. Overall, our results indicated that when bilinguals spoke digits aloud in the language switching task, switch costs mainly occurred at the lemma selection stage.

Electrophysiological evidence for domain-general inhibitory control during bilingual language switching.

This paper presents an experiment that explored the role of domain–general inhibitory control on language switching. Reaction times (RTs) and event–related brain potentials (ERPs) were recorded when low–proficient bilinguals with high and low inhibitory control (IC) switched between overt picture naming in both their L1 and L2. Results showed that the language switch costs of bilinguals with high–IC were symmetrical, while that of bilinguals with low–IC were not. The N2 component failed to show a significant interaction between group, language and task, indicating that inhibition may not comes into play during the language task schema competition phase. The late positive component (LPC), however, showed larger amplitudes for L2 repeat and switch trials than for L1 trials in the high–IC group, indicating that inhibition may play a key role during the lexical response selection phase. These findings suggest that domain–general inhibitory control plays an important role in modulating language switch costs and its influence can be specified in lexical selection phase.

A control process model of code-switching

Code-switching (CS) is central to many bilingual communities and, though linguistic and sociolinguistic research has characterised different types of code-switches (alternations, insertions, dense CS), the cognitive control processes (CPs) that mediate them are not well understood. A key issue is how during CS speakers produce the right words in the right order. In speech, serial order emerges from a speech plan in which items are represented in parallel. We propose that entry into the mechanism for speech planning (a competitive queuing mechanism) is governed by CPs best suited to the particular types of code-switches. Language task schemas external to the language network govern access. In CS, they are coordinated cooperatively and operate in a coupled or in an open control mode. The former permits alternations and insertions whereas the latter is required for dense CS. We explore predictions of this CP model and its implications for CS research.

An ERP study of inhibition of non-target languages in trilingual word production

The present study examined the locus where inhibition of non-target languages is exerted during trilingual word production by analyzing the cue-locked and stimulus-locked ERPs respectively in the n-2 language repetition paradigm. During the experiment, Uighur–Chinese–English trilinguals overtly named Arabic digits in one of their three languages according to a visually presented cue while their behavioral and electrophysiological responses were recorded. The behavioral data revealed insignificant n-2 repetition costs. Cue-locked ERPs revealed also only tiny or marginally significant n-2 repetition effects over some midline electrodes. The stimulus-locked ERP data showed a more negative ERP component elicited by the n-2 repetition trials than the n-2 non-repetition trials around 250ms after stimulus onset, but no significant difference in this ERP effect between different languages was found. The results indicate that inhibition of non-target languages occurs at the lemma selection phase rather than the language task schemas phase during trilingual language production.

Language conflict in translation: An ERP study of translation production

Although most bilinguals can translate with relative ease, the underlying neuro-cognitive processes are poorly understood. Using event-related brain potentials (ERPs) we investigated the temporal course of word translation. Participants translated words from and to their first (L1, Dutch) and second (L2, English) language while ERPs were recorded. Interlingual homographs (IHs) were included to introduce language conflict. IHs share orthographic form but have different meanings in L1 and L2 (e.g., “room” in Dutch refers to cream). Results showed that the brain distinguished between translation directions as early as 200 ms after word presentation: the P2 amplitudes were more positive in the L1→L2 translation direction. The N400 was also modulated by translation direction, with more negative amplitudes in the L2→L1 translation direction. Furthermore, the IHs were translated more slowly, induced more errors, and elicited more negative N400 amplitudes than control words. In a naming experiment, participants read aloud the same words in L1 or L2 while ERPs were recorded. Results showed no effect of either IHs or language, suggesting that task schemas may be crucially related to language control in translation. Furthermore, translation appears to involve conceptual processing in both translation directions, and the task goal appears to influence how words are processed.

Inhibition of non-target languages in multilingual word production: Evidence from Uighur–Chinese–English trilinguals

The present study examined the hypothesis whether non-target languages are inhibited during multilingual language production by examining the n-2 language repetition cost. In two experiments, Uighur–Chinese–English trilinguals named Arabic digits in one of their three languages according to a visually presented cue. Significant n-2 repetition costs were obtained in both experiments, which indicate that inhibition exists during multilingual word production. In addition, in Experiment 1, it was also found that the n-2 repetition cost was reduced when cues were highly compatible with the task, which means non-target languages are less inhibited. In Experiment 2, the n-2 repetition cost was increased at a shorter CSI. Taken together, these results indicate that inhibition of non-target languages occurs during multilingual language production, and that efficiency of establishing the target language task schema has an effect on the inhibitory control process.

Multiple task demands in action disorganization syndrome

We present novel evidence for specific, disruptive effects of multi-step task demands on the production of everyday life tasks in two patients with Action Disorganization Syndrome (ADS). Experiment 1 demonstrated that the patients were impaired at carrying out everyday life tasks but improved when they instructed the examiner to perform the tasks. Experiments 2 and 3 showed that the patients improved in their own performance when the demands were reduced – by eliminating the need to use their own task schema (Experiment 2) and by reducing the need for error monitoring (Experiment 3). Experiment 4 reduced sequential effects in task performance further by having patients perform the same individual actions as before but now out of the task context. The data indicate that ADS patients are vulnerable to multiple demands made during the performance of everyday life tasks and show abnormal, disruptive effects of three specific factors: (i) using a self-driven schema, (ii) error monitoring and (iii) the activation of prior actions from the same task. The implications for understanding ADS are discussed.

Strategies to Translate Information Technology (IT) Terms

The rendezvous amid translation theories and models is the quest for an approach to translation units, which is especially significant in translating scientific terms outside as well as inside a context. The research seeks to explore two principles of translating terms, ―recombination of semantic components‖ and ―functional equivalence‖, which helps look through terms outside the context and determine if the translation units in translating the IT terms are morphemes, words, or phrases. Based on these term translation principles and Vinay & Darbelnet’s (1977) translation procedures, six procedures of translating terms–direct borrowing, ―calque‖ or loan translation, literal translation, transposition, modulation, and adaptation–are developed. Besides, Coordinate Translation Model with three axes: X axis–translation unit axis, Y axis–concept equivalent axis, and Z axis–language task schema axis is proposed to locate appropriate translation units for translating IT terms inside the context.