The Language Of Bargaining: Linguistic Effects In LLM Negotiations

Stuti Sinha, Himanshu Kumar, Aryan Raju Mandapati, Rakshit Sakhuja, Dhruv Kumar

Track: long paper (up to 10 pages)

Keywords: large language models, multi-agent negotiation, multilingual evaluation, linguistic framing, strategic reasoning, game theory, Indic languages, language-mediated behavior, bargaining, cross-lingual generalization, ultimatum game, resource exchange, distributive bargaining, integrative bargaining, LLM evaluation

TL;DR: Language choice systematically shapes LLM negotiation behavior. Indic languages reduce agreement stability in competitive games but increase cooperative exploration, showing that English-only evaluation of LLM strategic reasoning is insufficient.

Abstract: Negotiation is a core component of social intelligence, requiring agents to balance strategic reasoning, cooperation, and social norms. Recent work shows that LLMs can engage in multi-turn negotiation, yet nearly all evaluations occur exclusively in English. We systematically isolate language effects across English and three Indic framings (Hindi, Punjabi, Gujarati) by holding game rules, model parameters, and incentives constant for multi-agent simulations for Ultimatum, Buy-Sell, and Resource Exchange games. Our results indicate that language choice is correlated to negotiation outcomes and that this correlation persists across model architectures. Crucially, effects are task-contingent: Indic languages reduce stability in distributive games (where agents compete over a fixed resource) yet are associated with richer exploration in integrative (cooperative) settings. Further, they suggest that findings obtained under English-only conditions may not generalize to other linguistic settings. These findings caution against English-only evaluation of LLMs and suggest that linguistically-grounded evaluation is essential for fair deployment.

Anonymization: This submission has been anonymized for double-blind review via the removal of identifying information such as names, affiliations, and identifying URLs.

Submission Number: 184