Elamur Research⌕
☰
home / papers / 2026 / Towards Linguistically-informed Representations for English as a Second or Foreign Language: Review, Construction and Application

Towards Linguistically-informed Representations for English as a Second or Foreign Language: Review, Construction and Application

Paper · 2026 · Wenxi Li, Xihao Wang and Weiwei Sun

Summary

The paper argues that English produced by non-native speakers (ESFL) should be treated as a coherent linguistic system rather than as broken English. Existing learner resources always anchor ESFL to standard English, so they face a trade-off. Strict anchoring hides learner-specific structure, while loose anchoring damages consistency. The authors review these resources and sort them into three families: target-language-based, multiple-criteria-based and universal-framework-based. They then propose a constructivist alternative in which both ESFL and English sentences are built from one shared inventory of form–meaning constructions. This inventory is implemented in Synchronous Hyperedge Replacement Grammar (SHRG), where each rule pairs a syntactic tree fragment with a semantic graph fragment.

Using this formalism, they create ESFL SemBank, a gold-standard collection of 1643 sentences. Each sentence has aligned syntactic and semantic analyses, derived from earlier silver graphs through a Review–Revise–Rebuild process. In a pilot test of the Linguistic Niche Hypothesis, ESFL shows a rule distribution similar to English but a more transparent link from syntax to meaning. This makes the resource a quantitative tool for second-language acquisition research.

Key points

  • Survey of prior resources:

    • Konan-JIEM uses phrase-structure trees with error node labels.
    • The NOCE corpus gives each token three parallel POS tags.
    • SALLE separates morphological and distributional layers.
    • TLE (UD English-ESL) pairs each learner sentence with a corrected version, both annotated with PTB POS tags and UD dependencies.

  • Source of the silver graphs: Zhao et al. (2020) parsed learner sentences with the ACE parser and the ERG. They then picked, from the K-best graphs, the one closest to the TLE gold tree, which succeeded for 2691 sentences. ESFL SemBank reviews these graphs, revises rejected ones and rebuilds unparsable sentences with selected or newly written SHRG rules.

  • Learner deviations as constructions: omission, insertion and transposition are modelled as constructions that map unusual surface forms onto meanings shared with English.

  • Syntactic comparison: a Chi-Square Test compared 43 non-lexical CFG rules in ESFL and English (χ² = 4.244, p = 0.999). It found no significant difference.

  • Semantic transparency: each SHRG rule was swapped for the most frequent alternative with the same CFG rule. After the swap, ESFL graphs kept more meaning than English ones:

    MeasureESFLEnglish
    Mean Smatch0.9060.875
    SD0.0640.068

    The difference was confirmed by a t-test (T = 14.343, p ≈ 8.8e-41) and a z-test.

Interesting ideas

  • I1. Parallel tags for conflicting evidence: when lexical, morphological and distributional evidence disagree on a token's category, record all three tags side by side instead of forcing one choice; this keeps the conflict visible for later stages.
  • I2. Review–Revise–Rebuild curation: start from automatic silver analyses, accept the good ones, hand-fix the wrong ones, and add new grammar rules only where nothing parses; human effort then goes to the hard residue.
  • I3. Rule-swap stability test: replace each syntax-to-semantics rule with the most common rule sharing the same syntax, then measure how much meaning survives; this quantifies how much interpretation depends on specific constructions versus the general pattern.
  • I4. Picking semantics by syntactic agreement: choose among K-best semantic graphs the one most consistent with an independently produced gold syntax tree, so that a trusted layer disambiguates a weaker one.
  • I5. Deviations as alternative form–meaning pairings: treat forms like inserted prepositions or swapped word order as variant constructions that map onto the same meaning as the standard form, rather than as parse errors.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (tokens): the en tagger could store separate lexical, morphological and distributional tags for noisy or learner input, so later steps can explain which source of evidence they followed.
  • I2 (learn): silver trees agreed by the witness parsers could be triaged the same way, with failures turned into new self-induced repair rules instead of being discarded.
  • I3 (frames): substituting the default VerbNet frame for the chosen one and scoring the difference would show where prag relies on fine lexical choices, and serves as a natural negative control.
  • I4 (tree): gparse and the perceptron parser's best tree could be used to rerank alternative frame or semantic analyses, giving an explainable selection criterion.
  • I5 (frames): frame extraction could include learner-style construction variants (e.g. "contacted with X") that produce the same who-did-what-to-whom frame as the canonical form.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of CambridgeCambridge · United Kingdom243 works · 121,396 citations🎓Peking UniversityBeijing · China244 works · 39,359 citations🎓Minzu University of ChinaBeijing · China5 works · 2 citations

Also on the byline: Bridge University · Nation University.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,263 works

Cite

Cite as Wenxi Li, Xihao Wang and Weiwei Sun. 2026. Towards Linguistically-informed Representations for English as a Second or Foreign Language: Review, Construction and Application. arXiv:2604.09008.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2604.09008 · arXiv 2604.09008

@misc{li2026linguistically,
  title = {{Towards Linguistically-informed Representations for English as a Second or Foreign Language: Review, Construction and Application}},
  author = {Wenxi Li, Xihao Wang and Weiwei Sun},
  year = {2026},
  doi = {10.48550/arxiv.2604.09008},
  eprint = {2604.09008},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2604.09008}
}

References

36 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Abend O, Rappoport A. Universal Conceptual Cognitive Annotation (UCCA). In: Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) Sofia, Bulgaria: Association for Computational Linguistics; 2013. p. 228–238. https://www.aclweb.org/anthology/P13-1023. ↗
  2. Adjemian C. On the nature of interlanguage systems. Language learning. 1976;26(2):297– 320. find
  3. Banarescu L, Bonial C, Cai S, Georgescu M, Griffitt K, Hermjakob U, et al. Abstract meaning representation for sembanking. In: Proceedings of the 7th linguistic annotation workshop and interoperability with discourse; 2013. p. 178–186.
  4. Berzak Y, Kenney J, Spadine C, Wang JX, Lam L, Mori KS, et al. Universal Dependencies for Learner English. In: Erk K, Smith NA, editors. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) Berlin, Germany: Association for Computational Linguistics; 2016. p. 737–746. https://aclanthology.org/P16-1070. ↗
  5. Chomsky N. Lectures on Government and Binding: The Pisa Lectures. Mouton de Gruyter; 1981. find
  6. Chomsky N. Knowledge of Language: Its Nature, Origin, and Use. Byrne D, Kölbel M, editors, Prager; 1986. find
  7. Copestake A. Slacker Semantics: Why Superficiality, Dependency and Avoidance of Commitment can be the Right Way to Go. In: Proceedings of the 12th Conference of the European Chapter of the ACL (EACL 2009) Athens, Greece: Association for Computational Linguistics; 2009. p. 1–9. http://www.aclweb.org/anthology/ E09-1001. ↗
  8. Corder SP. Error analysis and interlanguage, vol. 198. London: Oxford university Press; 1982. find
  9. Croft W. Radical construction grammar: Syntactic theory in typological perspective. Oxford University Press; 2001. find
  10. Davidson S, Yamada A, Fernandez Mira P, Carando A, Sanchez Gutierrez CH, Sagae K. Developing NLP Tools with a New Corpus of Learner Spanish. In: Calzolari N, Béchet F, Blache P, Choukri K, Cieri C, Declerck T, et al., editors. Proceedings of the Twelfth Language Resources and Evaluation Conference Marseille, France: European Language Resources Association; 2020. p. 7238–7243. https://aclanthology.org/2020.lrec-1.894. ↗
  11. Dıaz-Negrillo A, Meurers D, Valera S, Wunsch H. Towards interlanguage POS annotation for effective learner corpora in SLA and FLT. In: Language Forum, vol. 36; 2010. p. 139–154. find
  12. Dickinson M, Ragheb M. On Grammaticality in the Syntactic Annotation of Learner Language. In: Meyers A, Rehbein I, Zinsmeister H, editors. Proceedings of the 9th Linguistic Annotation Workshop Denver, Colorado, USA: Association for Computational Linguistics; 2015. p. 158–167. https://aclanthology.org/W15-1619. ↗
  13. Flickinger D. On building a more efficient grammar by exploiting types. Natural Language Engineering. 2000;6(1):15–28. https://doi.org/10.1017/S1351324900002370. ↗
  14. Flickinger D, Bender EM, Oepen S. Towards an Encyclopedia of Compositional Semantics: Documenting the Interface of the English Resource Grammar. In: Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14) Reykjavik, Iceland: European Language Resources Association (ELRA); 2014. p. 875–881. http://www.lrec-conf.org/proceedings/lrec2014/pdf/562 Paper. pdf. ↗
  15. Flickinger D, Zhang Y, Kordoni V. DeepBank: A dynamically annotated treebank of the Wall Street Journal. In: Proceedings of the 11th International Workshop on Treebanks and Linguistic Theories; 2012. p. 85–96. find
  16. Goldberg AE. Constructions: A new theoretical approach to language. Trends in cognitive sciences. 2003;7(5):219–224. ↗
  17. Goldberg AE. Argument structure constructions versus lexical rules or derivational verb templates. Mind & Language. 2013;28(4):435–465. ↗
  18. Hoffmann T, Trousdale G. The Oxford handbook of construction grammar. Oxford University Press; 2013. find
  19. Kyle K, Eguchi M, Miller A, Sither T. A Dependency Treebank of Spoken Second Language English. In: Kochmar E, Burstein J, Horbach A, Laarmann-Quante R, Madnani N, Tack A, et al., editors. Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022) Seattle, Washington: Association for Computational Linguistics; 2022. p. 39–45. https://aclanthology.org/2022.bea-1.7. ↗
  20. Lee J, Li K, Leung H. L1-L2 Parallel Dependency Treebank as Learner Corpus. In: Miyao Y, Sagae K, editors. Proceedings of the 15th International Conference on Parsing Technologies Pisa, Italy: Association for Computational Linguistics; 2017. p. 44–49. https://aclanthology.org/W17-6306. ↗
  21. Li W, Wang X, Sun W. Compositional Syntactico-SemBanking for English as a Second or Foreign Language. In: Che W, Nabende J, Shutova E, Pilehvar MT, editors. Findings of the Association for Computational Linguistics: ACL 2025 Vienna, Austria: Association for Computational Linguistics; 2025. p. 24395–24406. https://aclanthology.org/2025.findings-acl.1252/. ↗
  22. Lin Z, Duan Y, Zhao Y, Sun W, Wan X. Semantic Role Labeling for Learner Chinese: the Importance of Syntactic Parsing and L2-L1 Parallel Data. In: Riloff E, Chiang D, Hockenmaier J, Tsujii J, editors. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing Brussels, Belgium: Association for Computational Linguistics; 2018. p. 3793–3802. https://aclanthology.org/D18-1414. ↗
  23. Lupyan G, Dale R. Language structure is partly determined by social structure. PloS one. 2010;5(1):e8559. find
  24. Nagata R, Sakaguchi K. Phrase Structure Annotation and Parsing for Learner English. In: Erk K, Smith NA, editors. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) Berlin, Germany: Association for Computational Linguistics; 2016. p. 1837–1847. https://aclanthology.org/P16-1173. ↗
  25. Nagata R, Whittaker E, Sheinman V. Creating a manually error-tagged and shallow-parsed learner corpus. In: Lin D, Matsumoto Y, Mihalcea R, editors. Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies Portland, Oregon, USA: Association for Computational Linguistics; 2011. p. 1210–1219. https://aclanthology.org/P11-1121. ↗
  26. Nemser W. Language contact and foreign language acquisition. Languages in Contact and Contrast Essays in Contact Linguistics. 1991;p. 345–364. find
  27. Nivre J, de Marneffe MC, Ginter F, Goldberg Y, Hajič J, Manning CD, et al. Universal Dependencies v1: A Multilingual Treebank Collection. In: Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16) Portorož, Slovenia: European Language Resources Association (ELRA); 2016. p. 1659–1666. https://www.aclweb.org/anthology/L16-1262. ↗
  28. Oepen S, Lønning JT. Discriminant-Based MRS Banking. In: Proceedings of the Fifth International Conference on Language Resources and Evaluation (LREC’06) Genoa, Italy: European Language Resources Association (ELRA); 2006. http://www.lrec-conf.org/proceedings/lrec2006/pdf/364 pdf.pdf. ↗
  29. Partee B. Compositionality. In: Varieties of Formal Semantics: Proceedings of the 4th Amsterdam Colloquium, vol. 3 Foris Dordrecht; 1984. p. 281–311. find
  30. Pollard C, Sag IA. Head-driven phrase structure grammar. University of Chicago Press; 1994. find
  31. Ragheb M, Dickinson M. Developing a Corpus of Syntactically-Annotated Learner Language for English. In: Proceedings of the 13th International Workshop on Treebanks and Linguistic Theories Tübingen, Germany: University of Tübingen; 2014. p. 292–300. https://doi.org/10.5281/zenodo.10054513. ↗
  32. Sagae K, Davis E, Lavie A, MacWhinney B, Wintner S. High-accuracy Annotation and Parsing of CHILDES Transcripts. In: Buttery P, Villavicencio A, Korhonen A, editors. Proceedings of the Workshop on Cognitive Aspects of Computational Language Acquisition Prague, Czech Republic: Association for Computational Linguistics; 2007. p. 25–32. https://aclanthology.org/W07-0604. ↗
  33. Sagae K, MacWhinney B, Lavie A. Adding Syntactic Annotations to Transcripts of Parent-Child Dialogs. In: Lino MT, Xavier MF, Ferreira F, Costa R, Silva R, editors. Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC’04) Lisbon, Portugal: European Language Resources Association (ELRA); 2004. http://www.lrec-conf.org/proceedings/lrec2004/pdf/749.pdf. ↗
  34. Santorini B. Part-of-speech tagging guidelines for the penn treebank project (3rd revision). Philadelphia, PA: The University of Pennsylvania; 1990.
  35. Selinker L. INTERLANGUAGE. International Review of Applied Linguistics in Language Teaching. 1972;10(1-4):209–232. https://doi.org/10.1515/iral.1972.10.1-4. 209, https://doi.org/doi:10.1515/iral.1972.10.1-4.209. ↗
  36. Zhao Y, Sun W, Cao J, Wan X. Semantic Parsing for English as a Second Language. In: Jurafsky D, Chai J, Schluter N, Tetreault J, editors. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics Online: Association for Computational Linguistics; 2020. p. 6783–6794. https://aclanthology.org/2020. acl-main.606. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/li-2026-towards-linguistically-informed-representations-english/