Lost in the Tower of Babel: The Adverse Effects of Incidental Multilingualism in LLMs

TL;DR

This paper critiques the current incidental multilingualism paradigm in LLMs, highlighting its flaws and proposing a shift towards intentional multilingual design for fairness and robustness.

Contribution

It provides an empirical analysis of language support and response accuracy in LLMs and advocates for multilingualism by design as a core research goal.

Findings

LLMs often support languages differently than they respond in.

Simple language-change attacks can reveal hidden language assumptions.

Current models exhibit brittle and unequal multilingual behavior.

Abstract

This paper argues that contemporary multilingual NLP has converged on a fragile and misleading paradigm of incidental multilingualism. Today's LLMs appear multilingual largely because they are trained on massive, uneven web corpora, not because multilingual or multicultural competence has been treated as a core design objective. We contend that this paradigm systematically produces unequal, brittle, and opaque behavior across languages, with severe consequences in real-world and agentic deployments where models must reason, plan, and act across multiple linguistic contexts. We report a focused empirical study of two practical questions: which languages models self-report as supported and which languages they actually respond in across multilingual prompts. We additionally demonstrate how even a simple language-change attack can surface these failures and expose hidden assumptions about language in LLM-based systems. To address this, we call for a shift toward multilingualism by design: a research agenda that treats equitable multilingual performance, cultural grounding, and cross-lingual behavioral understanding as first-class goals in all aspects of the model pipeline.