Lucia Donatelli

Project title

Language and Analogical Reasoning

Research question

How do composition and analogy interact during human language acquisition? How is the abstraction involved in these processes represented in humans, and can it be replicated in machines?

Project description

Lucia Donatelli investigates the relationship between analogical reasoning and language, with a particular interest in how analogy and composition underpin compositional generalisation — the capacity to systematically recombine familiar concepts, as encoded in linguistic structure, into novel ones of a differing linguistic form. Her work seeks to illuminate the mechanisms by which speakers extend and repurpose what they already know to express something entirely new.

Selected publications

• Qin, Y., Varghese, D., Dahlgren Lindstr¨om, A., Donatelli, L., Misra, K., Kim, N. 2025. Visionand-Language Training Helps Deploy Taxonomic Knowledge but Does Not Fundamentally Alter It. Accepted, Neural Information Processing Systems (NeurIPS).
• Tao, S., Donatelli, L., & Hanh, M. 2024. More frequent verbs are associated with more diverse valency frames: Efficient principles at the lexicon-grammar interface. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Association for Computational Linguistics.
• Otto, C., Yang, X., Groschwitz, J., Koller, A. & Donatelli, L. 2024. A Corpus of German Abstract Meaning Representation. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Association for Computational Linguistics.