Form Without Function? Register Sensitivity in Cefr‐Conditioned LLM and Second‐Language Learner Writing
Bibliographic Data
| ID | 22416398 |
|---|---|
| Authors | Michele Carlo (0009-0007-9011-6531, Graduate School of Foreign Language Education and Research Kansai University, Japan), Michele Del Carlo (0000-0003-0431-662X, Kansai University, corresponding author), Osamu Takeuchi (0000-0002-5195-992X, Graduate School of Foreign Language Education and Research Kansai University, Japan) |
| Year | 2026 |
| Publication date | 2026-07-02 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | International Journal of Applied Linguistics (JOURNAL) |
| Journal identifiers | ISSN: 0802-6106 • E-ISSN: 1473-4192 |
| Publisher | Wiley (PUBLISHER • GB) |
| DOI | 10.1111/ijal.70284 |
| OpenAlex | W7167243873 |
| Language | EN |
| References cited | 23 |
Large language models (LLMs) can generate text that resembles L2 learner writing at requested proficiency levels, but whether such texts adapt to communicative contexts as human learners do remains unclear. This study compared LLM‐generated and human second‐language writing across three task‐defined categories (informal–personal, semi‐formal–expository, formal–argumentative) and four CEFR levels (A2–C1) using Biber's multi‐dimensional analysis (MDA). Six instruction‐tuned models responded to Write & Improve prompts ( n = 592 human; n = 4724 LLM), with EFCAMDAT as a human baseline. Under CEFR‐only prompting, LLM outputs shifted on MDA dimensions tied to informational density and explicitness, but showed weaker, differently directed, or less variable differentiation on dimensions tied to persuasion and elaboration, especially in formal–argumentative tasks. The largest observed gap was at C1, where CEFR descriptors most emphasize register flexibility. LLM outputs were more uniform than human writing and clustered near the center of the human distributional space, showing less variation than authentic learner texts. The broad pattern recurred across all six model families. CEFR‐conditioned generation can reproduce some structural signals of proficiency while showing limited human‐like register differentiation in MDA space, a pattern we interpret as form without function.
Human communication · Persuasion · Writing center · EFL/ESL Teaching and Learning · Innovative Teaching and Learning Methods · Second Language Acquisition and Learning
Register, Genre, and Style
Data Analysis Using Regression and Multilevel/Hierarchical Models
Variation across Speech and Writing
Discourse on the Move
Emmeans
Stanza
Does writing development equal writing quality? A computational investigation of syntactic complexity in L2 learners
Fitting Linear Mixed-Effects Models Using lme4
LmerTest Package
Reflexivity in human-written and ChatGPT-generated English research article abstracts
Does ChatGPT Write Like a Student? Engagement Markers in Argumentative Essays
The Emergence of Complexity, Fluency, and Accuracy in the Oral and Written Production of Five Chinese Learners of English
Theoretical Bases of Communicative Approaches to Second Language Teaching and Testing
| Citation velocity | historical |
|---|---|
| Highly cited | No |