How Conservative are Language Models? Adapting to the Introduction of Gender-Neutral Pronouns

2022/04/11 by Stephanie Brandl, Brandl, Stephanie, Ruixiang Cui +3
Computer Science · Social Sciences · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Gender Studies in Language #Text Readability and Simplification

paper · pdf · doi:10.48550/arxiv.2204.10281

openalex publication_date 2022/04/11 · openalex created_date 2022/05/08 · openalex updated_date 2026/07/28

Abstract

Gender-neutral pronouns have recently been introduced in many languages to a) include non-binary people and b) as a generic singular. Recent results from psycholinguistics suggest that gender-neutral pronouns (in Swedish) are not associated with human processing difficulties. This, we show, is in sharp contrast with automated processing. We show that gender-neutral pronouns in Danish, English, and Swedish are associated with higher perplexity, more dispersed attention patterns, and worse downstream performance. We argue that such conservativity in language models may limit widespread adoption of gender-neutral pronouns and must therefore be resolved.

Related

• What about em? How Commercial Machine Translation Fails to Handle (Neo-)Pronouns
• MISGENDERED: Limits of Large Language Models in Understanding Pronouns
• Welcome to the Modern World of Pronouns: Identity-Inclusive Natural Language Processing beyond Gender
• Pronouns Beyond the Binary: The Change of Attitudes and Use Over Time
• Don't Forget About Pronouns: Removing Gender Bias in Language Models Without Losing Factual Gender Information