Research Projects

The rapid integration of generative artificial intelligence tools like ChatGPT into educational contexts has created what can be termed an “implementation-before-investigation” phenomenon, where pedagogical practice has outpaced empirical research on learning outcomes. This study addresses the fundamental question of whether generative AI supports genuine second language (L2) writing development or merely enhances performance whilst technological support is available. Through a rigorous three-phase mixed-methods design, this project investigates both learner development and teacher preparation across multiple linguistic contexts. Phase 1 employs a quasi-experimental study with 200 L2 English learners using a matrix design that alternates between AI-supported and independent writing tasks. Crucially, pre- and post-benchmark assessments conducted without AI support will measure authentic development using established quantitative metrics (error-free T-units, lexical diversity) and qualitative measures (coherence, cohesion). Systematic interventions will introduce evidence-based strategies including prompt engineering, critical evaluation of AI output, and AI-assisted revision workflows. Phase 2 develops and validates a pedagogical framework through training 300 student teachers across English and modern languages programmes (French, Japanese, Korean). This cross-linguistic approach enables investigation of framework transferability across different orthographic systems and pedagogical traditions. Partial Least Squares Structural Equation Modelling will examine relationships between teacher beliefs, confidence, perceived AI usefulness, and pedagogical intentions. Phase 3 validates findings through classroom implementation, following student teachers into practicum and early in-service contexts. Simplified assessment tools from Phase 1 will monitor authentic student development whilst PLS-SEM modelling validates predictive relationships between teacher preparation and observed pedagogical behaviours. This research directly addresses critical gaps in our understanding of the impact of AI on language education by differentiating between AI-enhanced performance and sustained learning development. The validated pedagogical framework will provide scalable, evidence-based guidance for responsible AI integration, with cross-linguistic validation ensuring applicability across Hong Kong’s multilingual educational landscape and beyond. Outcomes will inform teacher education curricula, institutional policies, and assessment practices, contributing to principled approaches to AI integration during this transformative period in language education.

This project builds on the applicant’s current GRF study (#18601423) which investigates tone-vowel interaction in Cantonese and Mandarin. This line of work challenges the traditional ‘source and filter’ assumption that tone and segments are independent of each other. In this proposal, we seek to extend our current work on the tongue to another articulator – the larynx, using ultrasound imaging. This used to receive little attention in the past due to the low temporal resolution of ultrasound scanners, but is now emerging in phonetics due to much higher frame rates. To date, there are only a handful of published studies looking at larynx height with ultrasound (notably by Marc Brunelle at U of Ottawa) on Southeast Asian languages, leaving behind a huge gap in the literature on Chinese languages which also have rich tonal inventories. This project seeks to answer the following questions: - How do L1 and L2 speakers differ in tone production (esp. ‘tonal coarticulation’) - How are the movements of the tongue and the larynx temporally coordinated during tone production? - Does larynx height predict individual differences in tone-vowel interaction? We plan to test native Cantonese and Mandarin speakers, who also speak the other language as L2. Twenty participants will be recruited for each L1 group. They will produce Cantonese and Mandarin (either as L1 or L2) disyllables, and have their voice, as well as tongue and larynx movement recorded using ultrasound imaging. Our findings will enhance our understanding of articulatory planning and execution with respect to tone.

Cremation has been the most popular means of body disposition in Hong Kong; however, this practice raises concerns about pollution, public health, and land shortages, thereby threatening the sustainability of the city’s development. Green burial practices (GBP), such as using biodegradable materials and scattering cremated remains at green burial sites (e.g., gardens or at sea), offer environmentally friendly alternatives that can address the urgent demand for land resources and ease the financial burden on families. Since 2021, the SAR government has launched a campaign to promote GBP. The promotion involves textual and multimodal discourses, including webpages, videos, posters, and online articles. Prior research on the popularisation of GBP has tended to focus on public opinions toward these practices. However, crucial arguments, and linguistic and audio-visual elements that constitute the advertising discourse, have been overlooked, and the effectiveness of such discourse has remained unmeasured. This study perceives the green burials campaign as a form of strategic public communication, where the discourses aim to reformulate addressees’ beliefs, values, attitudes, and behaviours. Adopting a mixed-methods approach, the research will examine textual and multimodal discourses. Specifically, corpus-assisted discourse analysis, with the assistance of large-language models, will be conducted to reveal the discourses/themes and metadiscourse features in the textual data. Moreover, multimodal discourse features of the webpages, videos, and psoters related to GBP will be revealed and examined. These analyses elucidate how linguistic and semiotic patterns, as well as key themes/discourses, are utilised and constructed to (re)contextualise and (re)frame GBP in ways that are compatible with traditional Chinese funeral cultures and to influence addressees’ view towards GBP. To supplement the discourse studies, a survey will be distributed to investigate addressees’ perceptions of the promotional discourse and to measure its effectiveness. The significance of this project is three-fold. First, it contributes to the growing literature on death concepts, funeral culture, and environmental sustainability by explicating how GBP can be made compatible with traditional funeral cultures and help enhance environmental and socio-economic sustainability. Second, it elucidates the dynamic relationships between metadiscourse and contradictory ideologies (i.e., traditionalism and environmentalism) in advertising discourse. Third, it offers practical implications for enhancing the green burial promotion in terms of language style, content organisation, arguments, and the use of audio-visual resources.

For over half a century, Hong Kong’s visual identity has been shaped by the extraordinary density of urban texts, where Chinese and English signs created a distinctive streetscape that symbolized the city’s vitality and cross-cultural character. These texts form a holistic urban discourse that shapes social interactions and cultural meanings. Although recent regulations have led to the removal of many large-scale signs, Hong Kong remains a text-rich environment where multiple languages and scripts continue to shape public spaces, urban discourses, and cultural identity in evolving forms reflecting social and technological change. Previous studies have focused mainly on contemporary practices or isolated periods, often overlooking the evolution of urban textscapes amid shifting sociocultural and geopolitical realities, and the impact of physical carriers (from three-dimensional signage to mobile displays) on the city’s semiotic ecosystem. The interplay among text display, urban density, and architectural styles remains undertheorized, as does the spectrum of translation practices mediating between literal linguistic conversion and broader cultural acculturation. This project examines public signs in Hong Kong as a holistic system of everyday discourse, situating developments from 2015 to 2027 within a broader historical and sociopolitical context. Through fieldwork across diverse neighborhoods, from text-rich vernacular areas to text-poor financial districts, I will investigate the interaction of signage with color schemes, architectural details, and other semiotic resources in constructing urban visual identity. The research theorizes anew China-Western convergence by conceptualizing Hong Kong as a “city of translations”, where language and semiotic affordances mediate cultural interaction across China, the West, and beyond. Archival research and theoretical analysis of collected data will support an innovative framework for urban languages and semiotics, and the role of translation in shaping civic participation. An interactive digital platform will be established to serve as both a research corpus and a public engagement tool, which will enable continuous documentation of Hong Kong’s evolving linguistic landscape and fostering community participation in cultural preservation and creative imagination. This project will advance our understanding of urban spaces as mediators of cultural exchange, providing insights for language policy, heritage conservation, and urban planning. By conceptualizing Hong Kong as a city of translations, it will contribute new theoretical perspectives to sociolinguistics, translation studies, and cultural studies, and innovate our understanding of multilingual urban environments amid rapid technological and social change.

Health concepts have gained global importance, especially highlighted by the COVID-19 pandemic, with distinct perspectives across age groups significantly impacting public health strategies. In Hong Kong, traditional Chinese medicine popular among the elderly emphasizes natural remedies and a balanced lifestyle, whereas the youth are more inclined towards Western practices that advocate for active living and regular medical check-ups. These divergent views extend to illness management and public health behaviors, such as attitudes towards wearing masks. Our study investigates these health concept disparities between two key age groups, 15-25 and over 65, in Hong Kong. By exploring these generational and cultural differences, we aim to understand how changes in social, historical, and technological contexts have influenced health beliefs. This knowledge is crucial for developing targeted public health initiatives and educational programs tailored to the unique East-West blend of health practices in Hong Kong. Utilizing the Digital platform for collecting Online Language Data (DOLD) developed at the Education University of Hong Kong, we will collect and analyze personal health narratives from these age groups. Participants will share their experiences and perspectives on health, which will be transformed into textual data for corpus-assisted discourse analysis. This analysis will employ tools like AntConc to perform keyword analysis and explore how these keywords relate to health perceptions within the narratives. We will employ an enhanced narrative analysis framework that considers both the form and function of narratives within our corpus, using a mixed-methods approach to quantitatively code and qualitatively interpret how different age groups in Hong Kong conceptualize health, thereby informing more effective health communication strategies. The quantitative aspect of our research will examine correlations between health concept variations and demographic factors such as age, gender, and education level. This approach not only highlights the linguistic framing of health concepts but also underscores the extralinguistic factors affecting health views across different life stages. By adopting a constructionist approach and focusing on the distinct health perspectives of younger and older residents in Hong Kong, our study addresses a significant research gap. It aims to enhance understanding of health within these demographic groups, thereby improving health communication, informing policy, and optimizing educational outreach in Hong Kong’s culturally diverse environment. The outcomes are expected to provide valuable insights that will help refine health strategies to better meet the needs of these varied groups.

本項目根植於博物館面向包容和無障礙的巨變之中。每個展品標籤、語音描述、目錄、新聞稿和社交媒體貼文不再僅僅是文字，而是創造豐富個人體驗的媒介。香港的博物館已經接納這種變革的理念，但博物館文本創作者對於如何撰寫能夠引起多元觀眾共鳴的中英雙語文本仍存有疑惑。問題還包括缺少對有效的語言工具的瞭解和缺乏培訓材料以培養新一代博物館撰稿人。 這些實際問題與英語和中文應用語言學研究中的兩個缺陷有關：首先，在博物館語境下，對語言資源的研究幾乎未考慮觀眾接受度。其次，對包容的、無障礙的中文語言資源的關注極其有限。這些尚未解決的缺陷給本項目提供了研究機會和社會影響力，包括為香港的博物館提供包容和無障礙文本所需的語言工具和見解。 這項調查將專注於一種博物館文本體裁，即展品標籤，因其對參觀者的體驗至關重要。我們將進行一項接受度的研究，以了解多元觀眾的心理——是甚麼激發了他們的好奇心和參與度——透過調查揭示他們如何看待和接受英語和中文的展品標籤。隨後，一項語料庫輔助話語研究（CADS）將旨在確定與多元參觀者偏好相吻合的英語和中文語言資源。最終，接受度研究和CADS將使我們能夠確定哪些語言資源最適合觀眾偏好。我們還將支持中文語言在博物館中的有效使用，推進有關包容與無障礙的博物館文本的知識。 本項目的成果將包括一套多功能的培訓材料，具有各種益處。短期內，一系列定製的工作坊將提升香港博物館專業人員撰寫具有包容性的文本的技能。我們的合作夥伴將從寫作指南的改進中獲益，而大學教育者和學生將獲得增強寫作訓練資源。中期內，我們將透過在線課程與本地和國際博物館分享知識。由此帶來的整合包容策略將增強訪客的博物館體驗。最終，這將促使更廣泛的，語言更包容的社會轉變，從而使博物館參觀者和專業人員從中受益。

This project seeks to investigate the relationship between tongue movement and tone in speech production – two parts of articulation formerly considered independent from each other. We look at consonant-vowel coordination in Cantonese and Mandarin, two languages with respectively six and four lexical tones, under different tone and speech rate conditions. Both acoustic (formant) and articulatory (high temporal resolution ultrasound tongue imaging) data will be collected for analysis, followed by analysis-by-synthesis using VocalTractLab. Our findings will shed new lights on (i) our understanding of speech production, (ii) individual differences in articulatory control, and demonstrate (iii) the use of articulatory synthesis as a convenient tool for hypothesis-testing in articulation research.

Empirical corpus-based studies have demonstrated many positive outcomes in learners’ development of various language skills. However, frontline language teachers in primary and secondary schools are unfamiliar with corpus technology, mainly due to the lack of technological, pedagogical and content knowledge (TPACK) training in corpus technology. To address this knowledge gap, we have recently developed a corpus-based language pedagogy (CBLP) that blends language pedagogy with corpus technology. This proposal aims to frame training in corpus technology for student teachers within the TPACK framework to foster CBLP for effective teaching using corpus technology. This research will also provide a theoretical model by investigating how student teachers in Hong Kong and Chinese Mainland receive TPACK training in corpus technology, and how this can influence their self-efficacies for independent language learning and teaching.

The number of people being diagnosed with autism is increasing globally. Healthcare professionals define and categorise autism as a type of mental disorder, and individuals with autism (IWAs) are often perceived as having disabilities. Thus, autism is generally medicalised as an illness. The news media plays a crucial role in shaping the public’s understanding of autism and IWAs. Prior research has suggested that the negative representation and medicalisation of autism in newspapers can cause IWAs to be stigmatised and their mental health to deteriorate.

Despite negative impacts brought by medicalisation, very little research has systematically examined the linguistic construction of the medicalisation in the press. Without this knowledge, we cannot understand whether and how autism and autistic identity is stigmatised in the discourse. Other research gaps are also identified. First, very little research has focused on the Chinese speaking region in which more than 10 million people have been diagnosed with autism. Second, nearly all of the previous studies have used content analyses to examine somewhat small news corpora. The language use in the news portrayal of autism has been overlooked. Moreover, the studies discussed the impacts of the news portrayal about autism without exploring the perceptions of IWAs.

Guided by a social constructionist perspective, the proposed study investigates the news representation of autism in Hong Kong with consideration of the perception of IWAs. Based on a self-compiled corpus containing more than 18000 news articles related to autism, the focal analysis will be a corpus-assisted discourse analysis that combines corpus linguistics and critical discourse analysis to examine how autism is portrayed and medicalised linguistically through the analyses of keywords, word frequency, collocates, N-grams and concordances. Diachronic analyses will be conducted to understand changes of the representations and medicalisation over time (2000-2022). Then semi-structured individual interviews will be conducted with IWAs to explore their perceptions of autism, their autistic identity and the news portrayal of autism. The results of the analyses will be compared empirically to discuss the potential impacts of the news on IWAs and generate implications.

The significance of this project is three-fold. First, it elucidates how portrayals of autism and the medicalisation of autism in Chinese newspapers operate and may stigmatise IWAs from a discourse-analytical perspective. Second, it enhances the public’s understanding of IWAs by eliciting and investigating their voices. Third, it provides practical implications for journalists, researchers, educators, healthcare practitioners, and lawmakers.