EnTaCs: Analyzing the Relationship Between Sentiment and Language Choice in English-Tamil Code-Switching

Paul Bontempo

cs.CLMar 27, 2026

EnTaCs: Analyzing the Relationship Between Sentiment and Language Choice in English-Tamil Code-Switching

Authors: Paul Bontempo

Organizations: University of Colorado, Boulder Depts. of Computer Science and Linguistics

Abstract

This paper investigates the relationship between utterance sentiment and language choice in English-Tamil code-switched text, using methods from machine learning and statistical modelling. We apply a fine-tuned XLM-RoBERTa model for token-level language identification on 35,650 romanized YouTube comments from the DravidianCodeMix dataset, producing per-utterance measurements of English proportion and language switch frequency. Linear regression analysis reveals that positive utterances exhibit significantly greater English proportion (34.3%) than negative utterances (24.8%), and mixed-sentiment utterances show the highest language switch frequency when controlling for utterance length. These findings support the hypothesis that emotional content demonstrably influences language choice in multilingual code-switching settings, due to socio-linguistic associations of prestige and identity with embedded and matrix languages.

Figures & tables

Sentiment	n	EN%	TA%	Switches
Positive	19,871	34.3	61.7	1.99
Mixed_feelings	3,855	28.5	67.4	2.50
Negative	4,479	24.8	72.1	2.00

Table 1: Dataset statistics by sentiment.

Class	F1
English ( en )	0.932
Tamil ( ta )	0.638
Ambiguous ( na )	0.963
Macro avg	0.844
Accuracy	0.933

Table 2: Token-level language identification results on the validation set (epoch 5).

Figure 1: Distribution of English proportion by sentiment class. White diamonds indicate group means.

Figure 2: Distribution of language switch count per utterance by sentiment class. White diamonds indicate group means.

Code-Switching

Code-switching studies how people combine languages in speech and text and how AI systems can recognize, understand, and generate these mixtures.

See all 22 papers

Latent Mechanisms of Language Control in Multilingual Language Models
Aug 31, 2026Ryo Mitsuhashi, Sabri Boughorbel, Majd HawaslyMultilingual Language Models Multilingual Language Model Evaluation
Princeton University New Jersey, USA · Prince Sattam bin Abdulaziz University Riyadh, Saudi Arabia · QCRI, Hamad Bin Khalifa University, Doha, Qatar
A Cross-lingual Comparison of Human and Classification Model Entrainment Behavior in Code-switched Speech Settings
Jul 28, 2026Debasmita Bhattacharya, Siying Ding, Alayna Nguyen +1Code-Switching Interpretable ML
Department of Computer Science, Columbia University · Instagram
Loanword or Switch? The Annotation Boundary, Not the Model, Drives Kazakh-Russian Code-Switching Identification
Aug 1, 2026Bogdan SavelyevCode-Switching Language Identification
Independent Researcher Kazakhstan

Aug 31, 2026J/K move · Enter open · S save

Latent Mechanisms of Language Control in Multilingual Language Models

Ryo Mitsuhashi, Sabri Boughorbel, Majd Hawasly

Princeton University New Jersey, USA · Prince Sattam bin Abdulaziz University Riyadh, Saudi Arabia · QCRI, Hamad Bin Khalifa University, Doha, Qatar

Multilingual Language Models Multilingual Language Model Evaluation Code-Switching LLM Interpretability Causal Interventions in Language Models

Open paper AskPDF arXiv

Multilingual large language models can exhibit unintended code-switching -- unnecessarily alternating between languages during generation. We present a comparative study of three methods that identify language-controlling latents in cross-layer transcoders: activation value-based selection (ValSel), activation frequency-based selection (FreqSel), and LLM-generated latent annotation-based selection (AnnSel). To evaluate the efficacy of these methods in identifying language-controlling latents, we introduce two multilingual benchmarks that exhibit code-switching for fine-grained analysis of language steering across seven languages. Through targeted intervention experiments on Gemma-2-2B and Qwen3-4B, we find that all three methods effectively manipulate generation language, with FreqSel achieving the strongest overall performance, while AnnSel offering interpretable latent selection through explicit language annotations. A knock-out analysis suggests the methods select non-overlapping but each-functional latent subsets, indicating redundancy rather than a single canonical language direction. Code and data can be found at https://github.com/rm-3284/Latent-Mechanism-Multilingual.

Abstract

Code-Switching

Explore similar work

Latent Mechanisms of Language Control in Multilingual Language Models

A Cross-lingual Comparison of Human and Classification Model Entrainment Behavior in Code-switched Speech Settings

Loanword or Switch? The Annotation Boundary, Not the Model, Drives Kazakh-Russian Code-Switching Identification