Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Stylistic Evolution and LLM Neutrality in Singlish Language

Linus Tze En Foo, Weihan Angela Ng, Wenkai Li, Lynnette Hui Xian Ng

TL;DR

This study investigates the evolution of Singlish over a decade of informal digital text messages and proposes a stylistic similarity framework that compares lexico-structural, pragmatic, psycholinguistic, and encoder-derived features across years to quantify temporal variation.

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.

Stylistic Evolution and LLM Neutrality in Singlish Language

TL;DR

This study investigates the evolution of Singlish over a decade of informal digital text messages and proposes a stylistic similarity framework that compares lexico-structural, pragmatic, psycholinguistic, and encoder-derived features across years to quantify temporal variation.

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.
Paper Structure (28 sections, 6 figures, 5 tables)

This paper contains 28 sections, 6 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Contributions:
  3. Related Work
  4. Measuring Stylistic Similarity
  5. Feature Extraction
  6. Handcrafted Features
  7. Encoder-defined Linguistic Features
  8. Machine Learning Classifier Differentiation
  9. Similarity Metric
  10. Temporal Neutrality Metric
  11. Methodology
  12. Dataset
  13. Diachronic analysis of Real-World Data
  14. Generating Singlish Texts with LLMs
  15. Diachronic Alignment of LLM Outputs with CoSEM Data
  16. ...and 13 more sections

Figures (6)

  • Figure 1: Pipeline for our diachronic analysis of CoSEM and LLM-generated Singlish texts.
  • Figure 2: Trend of top LIWC features across categories.
  • Figure 3: Feature Importance of each of the handcrafted features plotted against year-gap. The feature importance is measured by taking the mean (absolute) SHAP values.
  • Figure 4: On CoSEM data, similarity score decreases as the year gap between data increases.
  • Figure 5: Similarity score of generated texts with CoSEM data vary across the years do not have obvious trends.
  • ...and 1 more figures