Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2605.09156 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910256525737984 |
|---|---|
| author | Chatterjee, Ahan Schöffel, Matthias Aßenmacher, Matthias Wiedner, Marinus Arias, Esteban Garces |
| author_facet | Chatterjee, Ahan Schöffel, Matthias Aßenmacher, Matthias Wiedner, Marinus Arias, Esteban Garces |
| contents | The diachronic evolution from Latin to the Romance languages involved a restructuring of the grammatical gender system from a tripartite configuration (masculine, feminine, neuter) to a bipartite one (masculine, feminine) in most Romance languages. In this work, we introduce an interpretable deep learning framework to investigate this phenomenon at both lexical and contextual levels. First, we show that conventional tokenization strategies are insufficiently robust for this low-resource historical setting, and that our proposed tokenizer improves performance over these baselines. At the lexical level, we evaluate the contribution of morphological features to gender prediction. At the contextual level, we quantify the contributions of different part-of-speech categories to grammatical gender prediction. Together, these analyses characterize the distribution of gender information between the lemma and its sentential context. We make our codebase, datasets, and results publicly available at \href{https://github.com/ahan-2000/Lost-in-Translation-}{https://github.com/ahan-2000/Lost-in-Translation-}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_09156 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Lost in Translation? Exploring the Shift in Grammatical Gender from Latin to Occitan Chatterjee, Ahan Schöffel, Matthias Aßenmacher, Matthias Wiedner, Marinus Arias, Esteban Garces Computation and Language Artificial Intelligence The diachronic evolution from Latin to the Romance languages involved a restructuring of the grammatical gender system from a tripartite configuration (masculine, feminine, neuter) to a bipartite one (masculine, feminine) in most Romance languages. In this work, we introduce an interpretable deep learning framework to investigate this phenomenon at both lexical and contextual levels. First, we show that conventional tokenization strategies are insufficiently robust for this low-resource historical setting, and that our proposed tokenizer improves performance over these baselines. At the lexical level, we evaluate the contribution of morphological features to gender prediction. At the contextual level, we quantify the contributions of different part-of-speech categories to grammatical gender prediction. Together, these analyses characterize the distribution of gender information between the lemma and its sentential context. We make our codebase, datasets, and results publicly available at \href{https://github.com/ahan-2000/Lost-in-Translation-}{https://github.com/ahan-2000/Lost-in-Translation-}. |
| title | Lost in Translation? Exploring the Shift in Grammatical Gender from Latin to Occitan |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2605.09156 |