|
|
ENQUIRE PROJECT DETAILS BY GENERAL PUBLIC |
| Project Details |
| Funding Scheme : | General Research Fund | |||||||||||||||||||||||||||
| Project Number : | 15606119 | |||||||||||||||||||||||||||
| Project Title(English) : | Neurocognitive bases of immediate learning success and lexical consolidation of second language words in older adults: a multi-modal (MRI/EEG) study on medial temporal lobe functions | |||||||||||||||||||||||||||
| Project Title(Chinese) : | 第二語言詞彙即時學習及強化的神經認知基礎： 內側顳葉功能的 MRI/EEG 研究 | |||||||||||||||||||||||||||
| Principal Investigator(English) : | Prof Wang, William Shi-yuan | |||||||||||||||||||||||||||
| Principal Investigator(Chinese) : | ||||||||||||||||||||||||||||
| Department : | Dept of Electronic Engineering | |||||||||||||||||||||||||||
| Institution : | The Hong Kong Polytechnic University | |||||||||||||||||||||||||||
| E-mail Address : | wsyw@cuhk.edu.hk | |||||||||||||||||||||||||||
| Tel : | ||||||||||||||||||||||||||||
| Co - Investigator(s) : |
|
|||||||||||||||||||||||||||
| Panel : | Humanities, Social Sciences | |||||||||||||||||||||||||||
| Subject Area : | Psychology and Linguistics | |||||||||||||||||||||||||||
| Exercise Year : | 2019 / 20 | |||||||||||||||||||||||||||
| Fund Approved : | 779,060 | |||||||||||||||||||||||||||
| Project Status : | Completed | |||||||||||||||||||||||||||
| Completion Date : | 30-6-2023 | |||||||||||||||||||||||||||
| Project Objectives : |
|
|||||||||||||||||||||||||||
| Abstract as per original application (English/Chinese): |
Language is made possible in the brain by exaptation, in which existing neurobiological systems supporting specific cognitive functions are recruited to subserve language functions. Because the medial temporal lobe (MTL) memory system, especially the hippocampus, is important for initial storage of recently encountered events (episodic encoding) and their subsequent transformation into long-term episodic/declarative memory (memory consolidation), the MTL likely plays a central role in learning words for both first language (L1) and second language (L2). However, there has been surprisingly few empirical works on its role in L2 learning. Instead, many studies reported that L2 learning success is determined by regions that support working memory, e.g., inferior parietal lobe (IPL). It is unclear whether the lack of control in phonological awareness towards L2 in previous works has contributed to the emphasis of IPL over MTL. In addition, older adults are under-represented in L2 research. However, they are generally known to utilize different brain networks than younger adults, casting further doubts regarding the exact brain network that subserves L2 learning across the population.
This project investigates how ageing affects the functional connections of the MTL during L2 (Italian) learning. Older and younger native Cantonese adults who have little exposure to Italian and pass a test of phonological awareness in both languages will be recruited. They will learn hundreds of L2 short phrases during a four-week L2 vocabulary learning programme. In addition, subjects will attend a pre-learning MRI session such that how the brain measures at the earliest phase of learning determine learning success can be investigated. The lexical consolidation process of newly learnt words will be tracked by event-related potentials (ERPs) at regular intervals, and by a post-learning MRI session, in which subjects will revise recently learnt L2 words or learn new L2 words. There are three primary hypotheses: (1) the individual variations in L2 learning success are strongly associated with the task-related functional connectivity between the hippocampus and other regions (e.g., IPL and ATL), (2) the relative roles of working memory and semantic memory in learning words are different across age; and (3) the lexical consolidation process is faster and more complete for younger than older adults. These hypotheses will be tested by network-based dynamic causal modelling and standard voxel-wise statistical parametric mapping for fMRI data, and by conventional ERP analysis. Collectively, our findings will clarify the functional connections of MTL structures that strongly contribute to L2 learning success during adulthood. 語言是一個非常複雜的行為，需要各個神經生理系統支援不同的語言功能。由於內側顳葉（MTL）的記憶系統，尤其是海馬體，掌管情節編碼（即儲存剛遇到的事件）以及記憶鞏固（將其轉化為長期情節／陳述性記憶），所以 MTL 很可能在第一語言（L1）及第二語言（L2）詞語學習上尤其重要。然而，有關 MTL 在L2學習中的作用的實證研究卻很少。相反，許多研究指出L2學習的成功率取決於支援工作記憶的區域，例如下頂葉（IPL）。 目前我們尚未清楚對下頂葉的這種重視是否因為之前的研究並沒有控制參加者的L2語音意識（phonological awareness）。此外，L2的文獻缺乏以老年人為對象的研究。然而，我們已知道老年人會利用與年輕人不同的大腦網絡。這進一步令我們對 L2 學習的確切大腦網絡產生疑問。 本研究旨在了解老化在學習 L2（意大利語）的過程中如何影響 MTL 的功能連結。我們會招募年老和年輕的廣東話母語者。他們需通過這兩種語言的語音意識測試，但不能有太多接觸意大利語的經驗。參加者將在為期四週的L2詞彙學習計劃中學習數百個L2短語。此外，參加者將會在計劃前參加一節磁力共振影像掃描（MRI），以便了解在最早期學習階段中的哪些大腦指標與學習的成功率特別有關。至於新學習單詞的詞彙鞏固過程，則會以事件相關電位（ERPs）定期追踪。研究共有三個主要假設：（一）L2學習成功率的個體差異與海馬體和其他區域（如IPL和ATL）之間的任務相關功能連結有密切的關係;（二）工作記憶及語義記憶於學習單詞的作用在不同年齡段是不同的;（三）年輕人的詞彙鞏固過程比老年人更快更完整。這些假設會透過動態因果建模（DCM）和標準體素統計參數圖（SPM），以及傳統的事件相關電位分析，來多方面進行驗證。總括而言，我們的研究結果將闡明內側顳葉結構中有哪些功能連結對成年期間學習 L2 特別重要。 |
|||||||||||||||||||||||||||
| Realisation of objectives: | The medial temporal lobe, especially the hippocampus, is critical for memory formation and consolidation, as emphasized in the complementary learning systems theory (McClelland et al., 1995). In the language domain, its role in foreign language learning (FLL) has also been highlighted, e.g., in Declarative–Procedural model (Ullman, 2001, 2004), but the evidence has remained scarce. In particular, it is plausible that long-term declarative memory would be recruited to a larger extent during FLL in older than younger learners by virtue of the lifelong growth in crystallized knowledge, but this hypothesis had yet to be tested. The present project aimed to fill these gaps and examined the role of medial temporal lobe and declarative memory in FLL. Italian was chosen as the target language, mainly due to its high grapheme-phoneme correspondence and its use of the Latin alphabet. Two phases of investigation (PHASE I & II) were undertaken. In PHASE I (Months 1-18), we focused on how the surface-based morphological features of the brain such as cortical thickness are associated with FLL performance. A total of 25 older and 8 younger learners were recruited in the first phase, with the completion rate being 84% and 100%, respectively. Each learner attended a 1.5-hour introductory Italian lesson, followed by an MRI session. Each learner was then enrolled to a five-week intensive Italian programme that comprised two learning components: (1) ten intensive computerized vocabulary learning lessons, and (2) four revision sessions, for tracking the process of lexical consolidation. Although these sessions focused on vocabulary building, basic grammatical components and rules (e.g., article, gender, present tense, adjectives) were also taught. A 1.5-hour Italian final test was administered post-learning. Phonological tests on Cantonese, English, and Italian, as well as a neuropsychological test battery comprising six standard cognitive tests were administered pre- and post-learning. In PHASE II (Months 19-42), the study protocols were fine-tuned with two notable changes: (1) fMRI data were collected both pre- and post- learning; (2) a phonology lesson on Cantonese and English was given prior to all Italian learning components. Taking advantage of the affordable rate of the then new MRI facility (managed by University Research Facility of Behavioral and Systems Neuroscience, UBSN, HKPolyU), an independent cohort of 33 younger learners (YL), 34 older learners (OL), and 23 older controls (OC) was recruited. The completion rates were 69.7%, 70.6%, and 78.3%, respectively. Results achieved: Regarding Objective 1 (To identify the functional connections of the hippocampus that subserve the immediate learning success and lexical consolidation of L2 words in adulthood), we found that the old words, which had already been acquired, elicited stronger activation in the left precuneus (X=-12, Y=-66, Z=30) and left calcarine fissure (X=-12, Y = -70, Z=20), while new words, which were being learnt for the first time inside the scanner, differentially activated the left parahippocampal gyrus (X=-36, Y=-26, Z=-20) and right middle cingulate gyrus (X=4, Y=30, Z=36). This main effect of OLDNEW interacted with RUN (run 1/2) at a more medial portion of the parahippocampal gyrus (X=-16, Y=-22, Z=-18), due to the growing difference between the old and new words in the second run. Our results underlined the lexical consolidation function of the hippocampus, showing that new words are actively maintained in the hippocampus during vocabulary learning. They provided new empirical support for the CLS theory and the DP model, both of which predicted a prominent role of the medial temporal lobe in FLL. Regarding Objective 2 (To examine the age-related differences in the relative role of working memory and semantic memory in studying L2 words), in Phase I of the project, we found that in older learners, the immediate learning and post-learning dictation score were best predicted by the delayed recall. In contrast, while immediate learning scores were similarly best predicted by delayed recall in younger controls, the post-learning dictation score was most strongly associated with English phonological awareness. These findings are consistent with our hypothesis that episodic memory could play a more crucial role than phonological processes in FLL in older adult learners. In Phase II of the project, the language learners were given a 1.5-hr phonology lesson prior to all Italian learning components, such that the participants would have better phonological awareness towards both their native language (L1) and English (L2). Because of this pedagogical improvement across the two phases, we found that phonological awareness became a much more prominent predictor of the vocabulary learning performance in older adults in Phase II. Nonetheless, the roles of episodic and semantic memory towards vocabulary learning had remained highly significant. For Objective 3 (To investigate the L2 lexical consolidation process, especially with respect to the age-related differences in the initial consolidation rate and the extent of long-term memory consolidation), it was originally hypothesized that younger and older learners would show distinct trajectories of learning, with the former group showing larger improvement over time, in terms of learning new words and retaining old words. However, there were no age-related difference in the rates of long-term consolidation of old words and immediate learning success of new words. In fact, older learners were as good or even better than the younger learners in most accuracy measures of Italian learning performance. Objective 4 (To explicate the intrinsic neural factors (grey matter volume, cortical thickness and functional connectivity) that underlie the brain’s readiness to learn and retain L2 words in both younger and older adults) was investigated by analyzing how the neural factors predict translation performance during in-class quizzes and final test. Using anatomical measurements and a cognitive test battery, we examined the hypothesis that older adults may engage semantic and episodic memory functions than other cognitive functions traditionally considered the most important for FLL (e.g., working memory capacity and phonological awareness). At the cortical level, in older learners, the immediate learning success of vocabulary was most strongly associated with the morphological features of the left pars orbitalis (χ2=79.54, df=20, p<.001), while the long-term memory retention was most strongly associated with the right entorhinal cortex (χ2=48.06, df=16, p=.001), which is known to be important for memory formation. Convergent evidence was also found based on the pattern of cognitive associations. At the subcortical level, the sub-cortical volumes of both the left and right thalami were a significant predictor of the L2 vocabulary maintenance in older adults (left: χ2=14.09, df=4, p=.036; right: χ2=13.54, df=4, p=.036; FDR-corrected). Our results are consistent with a prominent role of declarative memory functions in vocabulary learning in older learners. The analysis was reported in an article in Frontiers in Human Neuroscience (Fong et al., 2022). Regarding Objective 5 (To determine the efficacies of a four-week L2 vocabulary learning programme in enhancing cognition, especially in older adults), the pre–post comparisons revealed significant improvements in a range of cognitive domains, including episodic memory (verbal list learning test), semantic retrieval (picture naming), and matrix reasoning (Raven’s SPM). However, no interaction between Group (YL/OL/OC) and PrePost (pre/post) was found. The result suggested that the L2 vocabulary learning programme could potentially lead to cognitive enhancement, but due to the absence of the Group by PrePost interaction, it could not be determined whether such longitudinal cognitive improvements presented primarily training effects or genuine learning-induced gains. | |||||||||||||||||||||||||||
| Summary of objectives addressed: |
|
|||||||||||||||||||||||||||
| Research Outcome | ||||||||||||||||||||||||||||
| Major findings and research outcome: | At the time of the report, two journal articles were published based on the data collected in the present project (in Frontiers in Human Neuroscience and Languages), along with 8 conference abstracts/presentations in international conferences and a local conference (PolyU Academy for Interdisciplinary Research Conference 2023). In particular, based on the data collected in PHASE I, we obtained new empirical support for the recruitment of episodic and semantic memory in foreign language learning (Fong et al., 2022). Based on the data collected in PHASE II, we have obtained a novel finding about the role of the parahippocampal gyrus in lexical consolidation. Currently, two manuscripts are being prepared to publish the results arising from PHASE II (as highlighted in Section 5.3 and above); they will be submitted to Q1 international journals for publication within the next year. This delay was mainly because of the pandemic and that an MRI machine was available for use only in the middle of the project. The project team has made the best use of the invaluable opportunity offered by this project, to obtain longitudinal neuroimaging data from both the learner groups and the control group. These included diffusion-weighted imaging (DWI) data, which were collected both pre- and post-learning. The preliminary result based on DWI, regarding the association of vocabulary and grammar learning outcomes with three white matter structures (corpus callosum, arcuate fasciculus, and uncinate fasciculus), has been submitted to the Society for the Neurobiology of Language 16th Annual Meeting (October 24-26, 2024, Brisbane, Australia). The findings will also be written up for journal publications in due course. Another research related to second language learning has also been conducted. Specifically, code-switching is a common habit of life-long bilinguals. Previous research suggested that the comprehension of switching in language was cognitively taxing because it required the listener/reader to jump from one mental lexicon to another in order to process the words. However, our study found that prolonged reading time was only found in non-habitual use of code-switching. This finding adds to the knowledge on the organisation of mental lexicon, and eventually, on how learning a new language would affect it. The finding of this experiment was initially reported in the International Symposium of Bilingualism and later published in Languages (Hui et al., 2022). | |||||||||||||||||||||||||||
| Potential for further development of the research and the proposed course of action: |
The present research will be extended along three tracks. Firstly, upon further detailed analysis, the team will focus in disseminating the research findings as a series of publications in Q1 journals. In particular, we will prioritize to publish the novel findings about the roles of the hippocampus in lexical consolidation during foreign language learning. Apart from that, benefitted from the additional acquisition of diffusion MRI data in the present phase, a follow-up paper will incorporate the findings of both the structural and diffusion MRI data, to give a comprehensive account on the anatomical correlates of foreign language learning. Secondly, having established two education-matched younger and older beginners of Italian, the team now is now pushing forward the investigation on the longer-term neurocognitive changes of language learning. A follow-up research project based on an Italian booster programme was thus proposed and it has been funded by Departmental GRF. Thirdly, another project funded by the Faculty of Humanities, HKPolyU, focusing on artificial language learning, has been kickstarted. Although this research will zoom in on the linguistic functions of the cerebellum, it will further shed light on the role of medial temporal lobe in language learning. | |||||||||||||||||||||||||||
| Layman's Summary of Completion Report: | The medial temporal lobe, especially the hippocampus, is well-known for its role in formation of declarative memory, including both episodic and semantic memory. Surprisingly, the empirical evidence supporting its critical role in foreign language learning (FLL) has remained scarce. Indeed, phonological awareness, roughly our ability to discriminate sounds of a language, as well as phonological short-term memory, have usually been considered the most important predictors of foreign language learning success, rather than our episodic memory functions. However, it is plausible that older adults would especially recruit declarative memory in learning a new language, to take advantage of their crystallized knowledge accumulated over their lives. The present project aimed to evaluate the roles of medial temporal lobe and declarative memory functions in learning a foreign language (Italian), especially in older adults. Results revealed that the anatomical measures of brain regions implicated episodic and semantic memory functions act as a consistent predictor of learning success in older adults, with pattern of cognitive association as convergent evidence. In addition, functional magnetic resonance imaging showed that the parahippocampal gyrus is especially active during the acquisition of new vocabularies. Our results support various prominent accounts of language learning, including the complementary learning systems theory. | |||||||||||||||||||||||||||
| Research Output | ||||||||||||||||||||||||||||
| Peer-reviewed journal publication(s) arising directly from this research project : (* denotes the corresponding author) |
|
|||||||||||||||||||||||||||
| Recognized international conference(s) in which paper(s) related to this research project was/were delivered : |
|
|||||||||||||||||||||||||||
| Other impact (e.g. award of patents or prizes, collaboration with other research institutions, technology transfer, etc.): |
||||||||||||||||||||||||||||
| SCREEN ID: SCRRM00542 |