Stylistic Evolution and LLM Neutrality in Singlish Language

2026-01-10

arXiv.org

AI Summary

This study investigates the diachronic stylistic evolution of Singaporean English (Singlish) over a decade in response to sociotechnological shifts and evaluates the capacity of large language models (LLMs) to generate temporally neutral Singlish text. By constructing a multidimensional stylistic similarity framework that integrates lexical, pragmatic, psycholinguistic, and encoder-based features, the research conducts a cross-year comparative analysis of informal digital texts, offering the first systematic quantification of Singlish’s temporal variation. Findings reveal significant diachronic changes in tone, expressivity, and syntactic structure. Although LLMs can produce surface-plausible Singlish, their outputs retain detectable temporal signals, precluding genuine temporal neutrality—even when guided by tailored prompts or fine-tuning.

Technology Category

• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Singlish is a creole rooted in Singapore's multilingual environment and continues to evolve alongside social and technological change. This study investigates the evolution of Singlish over a decade of informal digital text messages. We propose a stylistic similarity framework that compares lexico-structural, pragmatic, psycholinguistic, and encoder-derived features across years to quantify temporal variation. Our analysis reveals notable diachronic changes in tone, expressivity and sentence construction over the years. Conversely, while some LLMs were able to generate superficially realistic Singlish messages, they do not produce temporally neutral outputs, and residual temporal signals remain detectable despite prompting and fine-tuning. Our findings highlight the dynamic evolution of Singlish, as well as the capabilities and limitations of current LLMs in modeling sociolectal and temporal variations in the colloquial language.

Problem

• Stylistic Evolution
• Singlish
• Temporal Neutrality
• LLM
• Sociolectal Variation

Innovation

• stylistic similarity framework
• diachronic language evolution
• Singlish
• large language models
• temporal neutrality