@inproceedings{oketch-etal-2026-cla,
title = "{T}-{CLA}: Taxonomy of Cultural-Linguistic Artifacts for Modeling Misalignment in {NLP}",
author = "Oketch, Kezia and
Lalor, John P. and
Zhang, Denghui and
Abbasi, Ahmed",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2693/",
pages = "61267--61284",
ISBN = "979-8-89176-459-0",
abstract = "We introduce T-CLA, a taxonomy for analyzing cultural-linguistic misalignment in NLP systems. We define misalignment as a mismatch between sociolinguistic variation in user language and model representations, and argue that such variation is systematically structured. We construct T-CLA through a bottom-up analysis of 8,680 Swahili health-related responses from Kenyan speakers, identifying 170 artifacts spanning code-mixing, Sheng usage, tribal lexicons, and loanwords. We validate the taxonomy through an annotation study with 155 participants and 1,560 annotations, showing that these artifacts are broadly recognizable but unevenly detectable across speakers. Using T-CLA as an analytic lens, we find that artifacts with lower detection salience are also associated with larger model errors and fairness disparities across tribal affiliations. T-CLA provides a generalizable approach for diagnosing cultural-linguistic misalignment and understanding model error from socially embedded linguistic variation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="oketch-etal-2026-cla">
<titleInfo>
<title>T-CLA: Taxonomy of Cultural-Linguistic Artifacts for Modeling Misalignment in NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Kezia</namePart>
<namePart type="family">Oketch</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">John</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Lalor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Denghui</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ahmed</namePart>
<namePart type="family">Abbasi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>We introduce T-CLA, a taxonomy for analyzing cultural-linguistic misalignment in NLP systems. We define misalignment as a mismatch between sociolinguistic variation in user language and model representations, and argue that such variation is systematically structured. We construct T-CLA through a bottom-up analysis of 8,680 Swahili health-related responses from Kenyan speakers, identifying 170 artifacts spanning code-mixing, Sheng usage, tribal lexicons, and loanwords. We validate the taxonomy through an annotation study with 155 participants and 1,560 annotations, showing that these artifacts are broadly recognizable but unevenly detectable across speakers. Using T-CLA as an analytic lens, we find that artifacts with lower detection salience are also associated with larger model errors and fairness disparities across tribal affiliations. T-CLA provides a generalizable approach for diagnosing cultural-linguistic misalignment and understanding model error from socially embedded linguistic variation.</abstract>
<identifier type="citekey">oketch-etal-2026-cla</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2693/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>61267</start>
<end>61284</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T T-CLA: Taxonomy of Cultural-Linguistic Artifacts for Modeling Misalignment in NLP
%A Oketch, Kezia
%A Lalor, John P.
%A Zhang, Denghui
%A Abbasi, Ahmed
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F oketch-etal-2026-cla
%X We introduce T-CLA, a taxonomy for analyzing cultural-linguistic misalignment in NLP systems. We define misalignment as a mismatch between sociolinguistic variation in user language and model representations, and argue that such variation is systematically structured. We construct T-CLA through a bottom-up analysis of 8,680 Swahili health-related responses from Kenyan speakers, identifying 170 artifacts spanning code-mixing, Sheng usage, tribal lexicons, and loanwords. We validate the taxonomy through an annotation study with 155 participants and 1,560 annotations, showing that these artifacts are broadly recognizable but unevenly detectable across speakers. Using T-CLA as an analytic lens, we find that artifacts with lower detection salience are also associated with larger model errors and fairness disparities across tribal affiliations. T-CLA provides a generalizable approach for diagnosing cultural-linguistic misalignment and understanding model error from socially embedded linguistic variation.
%U https://aclanthology.org/2026.emnlp-main.2693/
%P 61267-61284
Markdown (Informal)
[T-CLA: Taxonomy of Cultural-Linguistic Artifacts for Modeling Misalignment in NLP](https://aclanthology.org/2026.emnlp-main.2693/) (Oketch et al., EMNLP 2026)
ACL