The Influence of Linguistic Instructions on Personality Trait Manifestations in Large Language Models

DOI

10.2991/978-2-38476-444-0_23How to use a DOI?

Keywords

Large language models; Big Five personality; Cross-cultural psychology; Linguistic relativity; AI ethics

Abstract

The rapid deployment of multilingual Artificial Intelligence (AI) systems necessitates rigorous assessment of linguistic factors in machine behavior patterns. Existing studies have primarily focused on monolingual personality expression in large language models (LLMs), leaving critical gaps in understanding how bilingual instruction sets modulate cognitive architectures. This study employs a bilingual instruction controlled experiment to systematically investigate differences in the Big Five personality trait dimensions of six mainstream LLMs when processing Chinese and English instructions. Using the revised Big Five Inventory (BFI-2), representative Chinese and Western models—including GLM-4 and GPT-4—were evaluated. Paired-sample t-tests and effect size analyses were conducted to compare results. While no statistically significant differences were found between Chinese and English instructions across the five dimensions (p > 0.05), moderate effect trends emerged in Neuroticism (d = 0.51) and Openness (d = 0.60). Notably, English instructions significantly increased Neuroticism scores (Delta = 3.17), while Chinese instructions enhanced Openness (Delta = 3.83). These findings suggest that linguistic framing induces parameter activation patterns distinct to structural features of each language, with implications for culturally-aware AI training protocols.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Xichen Fang
PY  - 2025
DA  - 2025/07/10
TI  - The Influence of Linguistic Instructions on Personality Trait Manifestations in Large Language Models
BT  - Proceedings of the 3rd International Conference on Language and Cultural Communication (ICLCC 2025)
PB  - Atlantis Press
SP  - 191
EP  - 197
SN  - 2352-5398
UR  - https://doi.org/10.2991/978-2-38476-444-0_23
DO  - 10.2991/978-2-38476-444-0_23
ID  - Fang2025
ER  -