Yufeng Liu

Dr. Yufeng Liu

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications



27 July 2026 in PsyArXiv
Verified
Preprint
Open access
YLYufeng Liu
Yufeng Liu
XYXingyu Yang
Xingyu Yang
Word meanings change over time, yet the rate of this change varies systematically within each lexicon. While frequency, age of acquisition, and concreteness have been identified as within-language predictors of semantic change rate, a language-general account of why some concepts resist diachronic drift more persistently than others worldwide has remained elusive. Here, we propose that cross-linguistic semantic stability — the degree to which a concept’s semantic neighborhood is conserved across languages — functions as a fundamental, conceptual-level constraint on diachronic semantic evolution across lexica. We operationalized semantic stability as neighborhood overlap between a target concept’s nearest neighbors in English and the back-translated neighbors of its translation equivalents across 10-to-41 languages, and quantified diachronic semantic change as neighborhood turnover in historical embeddings spanning up to 100 years in English, French, German, and Chinese. Concepts with higher cross-linguistic stability exhibited significantly slower diachronic semantic change within each language and, critically, across lexica when datasets were aggregated cross-linguistically to derive concepts’ change rates. This constraining effect dominated all other predictors for shared concepts; survived controls for well-established psycholinguistic predictors; was structurally confirmed through partial-correlation networks as a direct, unmediated constraint; remained robust under iterative sample expansion of cross-lingual concept sets; and was replicated using historically derived stability estimates. Altogether, our findings reveal that the topological consistency of a concept’s semantic neighborhood across the world’s languages reflects a temporally durable cognitive-communicative constraint that actively governs the pace of meaning change, bridging synchronic cross-linguistic aligned conceptual structure with the diachronic dynamics of lexical evolution across lexica.

7 September 2026 in PsyArXiv
Verified
Preprint
Open access
YLYufeng Liu
Yufeng Liu
SCShifa Chen
Shifa Chen
YZYuqing Zhang
Yuqing Zhang
XYXingyu Yang
Xingyu Yang
BWBrent Wolter
Brent Wolter
Whether word meanings partition reality uniformly across languages or reflect language-specific categorization remains debated. Universalist accounts assume shared conceptual repertoires whereas relativist accounts emphasize language-specific partitions of conceptual space. However, evidence has relied largely on a few semantic domains and behavioral measures, leaving unclear whether and how graded semantic overlap of translation-equivalents affect the neural dynamics of concept processing. Here we combined computational semantic alignment (SA), embedding-based representational similarity analysis, and electroencephalography to test whether graded SA predicts cross-language representational correspondence and neural correlates of implicit semantic processing. Thirty-five Chinese–English bilinguals completed a translation-recognition task involving 160 translation-pairs across 20 semantic domains and 160 unrelated controls. High-SA translations showed stronger cross-language representational correspondence than low-SA translations, whereas controls showed near-zero correspondence. High-SA translations elicited less negative N400 amplitudes and larger N400 translation-priming effects. Moreover, millisecond-by-millisecond comparisons across the 0–1,000 ms epoch detected high–low SA differences within the canonical N400 window for translation-pairs, but no corresponding differences for controls at any time point across the entire epoch, indicating that the N400 finding was detectable without restricting comparisons to a preselected time window and unique to concepts represented by translation-pairs rather than non-translation controls. The SA effects from translation-equivalents were further modulated by second-language proficiency. Overall, by linking cross-linguistic distributional semantic structure to model representational geometry and electrophysiological responses, this work convergingly supports a nuanced relativist account of word meanings, in which bilinguals are neurophysiologically sensitive to SA, with greater sensitivity at higher levels of second-language proficiency.

18 November 2025 in Behavioral Sciences
Verified
Journal Article
Open access
PDF file
CHChunqiao Hu
Chunqiao Hu
SCShifa Chen
Shifa Chen
YLYufeng Liu
Yufeng Liu
Using the Action–Sentence Compatibility Effect (ACE) paradigm, this study investigated whether types of grammatical aspects and L2 proficiency influence embodied simulation during L2 sentence comprehension of English among Chinese learners. Participants judged the semantic plausibility of sentences in progressive or perfective aspect by performing directional actions (toward or away the body) that were either compatible or incompatible with the action direction described. Analysis of the reaction times (RTs) revealed a significant main effect of proficiency, with low-proficiency learners responding more slowly overall. Crucially, we observed a significant three-way interaction between aspect, action–sentence consistency, and proficiency. Simple effects analyses revealed a qualitative reversal: advanced learners exhibited a significant ACE only for sentences in the progressive aspect, indicating grammatically guided simulation sensitive to ongoing actions, whereas we found no ACE for perfective sentences, consistent with their focus on event completion rather than on-going action processes. In contrast, low-proficiency learners showed a significant ACE for the perfective aspect, suggesting a reliance on lexically triggered simulation, while they showed no simulation effect for the progressive aspect due to shallow morphosyntactic processing and L1 transfer. These findings support a proficiency-dependent dual-pathway model of L2 embodiment: advanced learners engage in direct mapping from grammar to simulation, whereas low-proficiency learners rely on an indirect, lexically mediated route. In summary, our findings demonstrate that the embodiment of grammatical meaning in L2 acquisition is not automatic, but is developmentally modulated, evolving from a lexically dependent to grammar-dependent simulation as proficiency increases. Furthermore, these results call for future research to explore the pedagogical applications of grammar-focused embodied instruction and to examine this dual-pathway model across other linguistic structures and L2 populations.