Under the Shadow of Babel: How Language Shapes Reasoning in LLMs

TL;DR

This paper investigates how large language models internalize language-specific reasoning biases, revealing that they reflect typological differences and semantic alignment in bilingual causal reasoning tasks, supporting linguistic relativity.

Contribution

Introduces BICAUSE, a bilingual causal reasoning dataset, and provides empirical evidence that LLMs internalize language-specific reasoning biases and shared semantic structures.

Findings

LLMs show typologically aligned attention patterns in causal reasoning.

Models rigidly apply language-specific causal word order, affecting performance.

Semantic representations converge across languages when reasoning succeeds.

Abstract

Language is not only a tool for communication but also a medium for human cognition and reasoning. If, as linguistic relativity suggests, the structure of language shapes cognitive patterns, then large language models (LLMs) trained on human language may also internalize the habitual logical structures embedded in different languages. To examine this hypothesis, we introduce BICAUSE, a structured bilingual dataset for causal reasoning, which includes semantically aligned Chinese and English samples in both forward and reversed causal forms. Our study reveals three key findings: (1) LLMs exhibit typologically aligned attention patterns, focusing more on causes and sentence-initial connectives in Chinese, while showing a more balanced distribution in English. (2) Models internalize language-specific preferences for causal word order and often rigidly apply them to atypical inputs, leading to degraded performance, especially in Chinese. (3) When causal reasoning succeeds, model representations converge toward semantically aligned abstractions across languages, indicating a shared understanding beyond surface form. Overall, these results suggest that LLMs not only mimic surface linguistic forms but also internalize the reasoning biases shaped by language. Rooted in cognitive linguistic theory, this phenomenon is for the first time empirically verified through structural analysis of model internals.