Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models

[2608.15129] Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models Skip to main content Search Submit Donate Log in Search arXiv Press Enter to search · Advanced search Computer Science > Computation and Language arXiv:2608.15129 (cs) [Submitted on 15 Aug 2026 ( v1 ), last revised 9 Sep 2026 (this version, v2)] Title: Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models Authors: Varvara Arzt , Allan Hanbury , Terra Blevins View a PDF of the paper titled Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models, by Varvara Arzt and 2 other authors View PDF HTML (experimental) Abstract: We systematically compare word order preferences in decoder-only language models across 192 artificial languages and typologically diverse natural languages. On artificial languages, models exhibit a left-branching preference that aligns with neither natural language universals nor human word order learning biases. On natural languages, monolingual models show no clear base word order bias at small scales, but as data grows, a preference for right-branching subject-verb-object (SVO) languages emerges while SOV falls behind despite being the most frequent order cross-linguistically. This SVO advantage extends to multilingual models and correlates with language resource level and data quality rather than word order. Thus, the same architecture exhibits opposite preferences on artificial and natural languages, establishing that word order biases observed in practice are data-driven. Since highly-resourced languages are overwhelmingly SVO, these biases risk gradually reducing word order diversity, particularly in languages that productively use multiple word orders, with the widespread adoption of LLMs. Comments: Accepted to EMNLP 2026 (Main Conference). 22 pages, 12 figures, 7 tables Subjects: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI) Cite as: arXiv:2608.15129 [cs.CL] (or arXiv:2608.15129v2 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.15129 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Varvara Arzt [ view email ] [v1] Sat, 15 Aug 2026 09:03:10 UTC (535 KB) [v2] Wed, 9 Sep 2026 15:25:31 UTC (536 KB) Full-text links: Access Paper: View a PDF of the paper titled Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models, by Varvara Arzt and 2 other authors View PDF HTML (experimental) TeX Source view license Current browse context: cs.CL < prev | next > new | recent | 2026-08 Change to browse by: cs cs.AI References & Citations NASA ADS Google Scholar Semantic Scholar export BibTeX citation Loading... BibTeX formatted citation × loading... Data provided by: Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) We gratefully acknowledge support from our major funders , member institutions , , and all contributors. About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab) Major funding support from

Related documents

• Left-Branching Transformers Excel at Right-Branching Languages: Data Shapes Word Order Preferences in Language Models#611079
• An Application of a Two-Sorted First-Order Language#113097
• Learning to Discover Interesting Mathematics#1076259
• Factors Influencing the Emergence of Dependency Length Minimization in Neural Agent Simulations#973635
• PRISM: Priming Relationships in Syntax and Mathematics#291095