Due to the lack of explicit word boundary markers, L2-Chinese learners have shown some difficulties in Chinese word segmentation. This study aimed to tackle the possible reasons of L2-Chinese learners’ difficulties in word segmentation: L1-biased processing strategy or developing mental representations of Chinese compound words, or both. In an eye-tracking experiment, high-frequency two-character Chinese compound words were used as targets. These compound words were embedded in sentences where their first component characters with prior verbs were manipulated to be either plausible or implausible, while the whole compound words were always plausible. Sentences were presented in character-spaced or word-spaced style. High-proficiency L2-Chinese learners and native Chinese speakers participated. Results revealed non-native-like patterns of L2-Chinese learners: they holistically processed compound words only in the word-spaced condition, while native speakers did so regardless how sentences were presented. The findings indicated that high-proficiency L2-Chinese learners’ difficulty in word segmentation is predominantly caused by their L1-biased processing strategy.

Distributional learning enables listeners to form phonetic categories by extracting statistical regularities from speech input. Younger Cantonese speakers can acquire the Mandarin level-falling (T1–T4) contrast through distributional learning, with bimodal exposure facilitating category formation and unimodal exposure suppressing it, and with fine-grained pitch sensitivity predicting success. However, aging is associated with declines in pitch sensitivity and phonetic boundary formation, which may disrupt this process. This study examined whether Cantonese-speaking older adults exhibit distributional learning of Mandarin T1–T4 and whether individual cognitive factors predict learning success. Sixty-four participants completed a pretest–training–posttest procedure with bimodal or unimodal exposure. While older adults improved in tone discrimination, no group differences emerged. Further analysis showed that those with lower pitch-related auditory memory failed to learn from unimodal input. On the other hand, fine-grained pitch perception abilities did not predict learning outcomes. These results suggest that older adults may rely on alternative learning mechanisms, such as memory-based strategies, when exposed to ambiguous input distributions. The findings indicate a shift from perceptual encoding to memory-driven processing in aging and highlight the limits of passive statistical learning in older adulthood.

This study uses data from the Corpus of Spoken Yiddish in Europe to compare the acoustic correlates of the length contrast in the peripheral vowels of two regions within the Central Yiddish dialect area: central Poland, considered the more conservative variety, and the Transcarpathian Unterland, hypothesized to have diverged from Polish Yiddish in the prewar period. Findings reveal smaller duration differences in [iː] versus [i] and [aː] versus [a] among Unterland speakers compared to speakers from Poland, with a gender effect in the Unterland showing smaller duration distinctions among women. The duration difference in [uː] versus [u] is significantly smaller than the other vowel pairs in both regions, likely reflecting its ambiguous phonemic status. Vowel quality shows no systematic differences between the two regions. The findings point to the possible influence of population mobility, dialect mixing, geopolitics, and multilingualism on vowel systems.

Generative artificial intelligence (GenAI) has been heralded by some as a transformational force in education. It is argued to have the potential to reduce inequality and democratize the learning experience, particularly in the Global South. Others warn of the dangers of techno-solutionism, dehumanization of learners, and a widening digital divide. The reality, as so often, may be more complicated than this juxtaposition suggests. In our study, we investigated the ways in which GenAI can contribute to independent language learning in the context of Pakistan. We were particularly interested in the roles of five variables that have been shown to be particularly salient in this and similar contexts: learners’ Generative Artificial Intelligence-mediated Informal Digital Learning of English (GenAI-IDLE) participation, AI Literacy, Foreign Language Enjoyment (FLE) and Foreign Language Boredom (FLB), and their second language Willingness to Communicate (L2 WTC). Employing a structural equation modelling approach, we surveyed 359 Pakistani English as a foreign language (EFL) learners to investigate their interrelationships between variables. The results demonstrate that EFL learners’ GenAI-IDLE activity directly and positively influences AI literacy and FLE. Students’ AI literacy and FLE play a chain-mediating role in the relationship between GenAI-IDLE participation and L2 WTC. However, FLB lacks predictive power over L2 WTC. We discuss the implications of these results for language learning, in particular in low-resource contexts.

Skilled and impaired language users can fail to understand sentences with noncanonical thematic role ordering. To advance understanding of the cause of the difficulty, we compared noncanonically ordered, object-experiencer-verb (OE) sentences to canonical agentive sentences in otherwise comparable syntactic structures. However, the field has neglected to consider that OE thematic roles may be altered by the comprehender’s interpretation; the aim of the investigation reported here was to address this research gap. We collected participants’ thematic role interpretations for sentences while measuring incremental processing speed as they read the sentences in a self-paced reading paradigm. Fluent English speakers rated the intent of the subjects of OE and agentive active sentences and of the objects of OE and agentive passive sentences. Consistently high ratings of intent identified a cluster containing agentives and an OE eventive subtype with an agent. In contrast, a spread of ratings in a second cluster suggested labile interpretation for a second OE subtype. Splitting the labile group at its active and passive mean, we compared reading time of OE and agentive sentences with a higher-intent, agent interpretation with OE sentences with a low-intent causer/theme interpretation. Whether active or passive, canonicity as determined by the relative prominence ordering of the rated role compared to the necessary OE experiencer or agentive theme, accounted for reading speed at the verb, second noun, and prepositional phrase. We propose that the results support an agent > experiencer > (causer)/theme thematic role hierarchy, suggesting refinements to event structure theory and little v/CAUS morphology.

This study presents a conceptual replication of Jacob et al.’s (2018) comparison of L2 early-stage processing of derived vs. inflected words. Previous studies on this issue focused predominantly on L2 learners from morphologically complex, alphabetic L1s, and generally showed L2 decompositional processing of derived but not inflected words. This replication study examined whether the previous claim for a qualitative difference in L2 early-stage processing of derived and inflected words could generalize to L2 English learners from a morphologically isolating, logographic L1, i.e., Chinese learners of L2 English. Results from a masked priming lexical decision task showed qualitatively the same magnitude of priming in the derivational, inflectional, and form control conditions for Chinese learners of English, suggesting reliance on surface form information in the early-stage processing of both derived and inflected words. Results of the current study add to the literature on L2 early-stage processing of derived vs. inflected words.

This retrospective cohort study examined the relationship between a continuous measurement of bilingual engagement (operationalized as language entropy) and cognitive aging in regional minority language speakers. We drew Frisian–Dutch bilinguals (n = 7,448) and Low Saxon–Dutch bilinguals (n = 10,114) from the Lifelines Cohort Study and included participants aged 20–80, enabling an adult lifespan perspective. Cognitive functioning was measured using the Cogstate Brief Battery, which assesses processing speed, attention, working memory and recognition memory. We did not observe a robust relationship between bilingual engagement and cognitive functioning. Our results suggest that bilingual engagement does not play a key role in processing speed, attention, working memory and recognition memory performance in Frisian–Dutch and Low Saxon–Dutch bilinguals. Implications for the bilingual engagement measurement and potential investigations into regional minority language bilingualism and cognition are discussed.

The adoption of corpus technology in school classroom settings remains limited, largely due to insufficient technological pedagogical content knowledge (TPACK) training for pedagogical corpus use. To address this gap, we investigated how teacher education in corpus-based language pedagogy (CBLP), a subdomain of TPACK for corpus technology tailored to language teachers, influenced student TESOL teachers’ self-efficacy for independent language learning and teaching. Employing a mixed-methods approach, including a CBLP training intervention (n = 120), survey data (n = 96), and interviews (n = 8) with student teachers at a university in Hong Kong SAR, China, the research validates a theoretical model through confirmatory factor analysis and structural equation modelling. Results demonstrate that corpus literacy (CL) is foundational for effective CBLP implementation and development of independent learning self-efficacy, which in turn fosters innovative, resource-rich instructional strategies. CBLP also enhances teachers’ self-efficacy for student engagement, fostering more interactive and motivating classrooms. These findings emphasise the value of embedding CL and CBLP within TESOL teacher-education programmes to prepare future language teachers for self-efficacy within dynamic, technology-enhanced classrooms.

Location mentions in local news are crucial for examining issues like spatial inequalities, news deserts and the impact of media ownership on news diversity. However, while geoparsing – extracting and resolving location mentions – has advanced through statistical and deep learning methods, its use in local media studies remains limited and fragmented due to technical challenges and a lack of practical frameworks. To address these challenges, we identify key considerations for successful geoparsing and review spatially oriented local media studies, finding over-reliance on limited geospatial vocabularies, limited toponym disambiguation and inadequate validation of methods. These findings underscore the need for adaptable and robust solutions, and recent advancements in fine-tuned large language models (LLMs) for geoparsing offer a promising direction by simplifying technical implementation and excelling at understanding contextual nuances. However, their application to U.K. local media – marked by fine-grained geographies and colloquial place names – remains underexplored due to the absence of benchmark datasets. This gap hinders researchers’ ability to evaluate and refine geoparsing methods for this domain. To address this, we introduce the Local Media UK Geoparsing (LMUK-Geo) dataset, a hand-annotated corpus of U.K. local news articles designed to support the development and evaluation of geoparsing pipelines. We also propose an LLM-driven approach for toponym disambiguation that replaces fine-tuning with accessible prompt engineering. Using LMUK-Geo, we benchmark our approach against a fine-tuned method. Both perform well on the novel dataset: the fine-tuned model excels in minimising coordinate-error distances, while the prompt-based method offers a scalable alternative for district-level classification, particularly when relying on predictions agreed upon by multiple models. Our contributions establish a foundation for geoparsing local media, advancing methodological frameworks and practical tools to enable systematic and comparative research.