@inproceedings{haider-can-2026-arabic,
title = "{A}rabic {B}aby{LM}: Effects of Diglossia and Tokenisation",
author = "Haider, Mozafar and
Can, Burcu",
editor = "Choshen, Leshem and
Cotterell, Ryan and
Gul, Mustafa Omer and
Jumelet, Jaap and
Linzen, Tal and
Mueller, Aaron and
Salhan, Suchir and
Shah, Raj Sanjay and
Warstadt, Alex and
Wilcox, Ethan Gotlieb",
booktitle = "Proceedings of the Second {B}aby{LM} Workshop",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.babylm-main.18/",
pages = "227--239",
ISBN = "979-8-89176-475-0",
abstract = "This paper explores building a baby language model for Arabic in the context of the BabyLM challenge, with a pretraining corpus limited to 10 million words. We focus on two areas: 1. morphologically-aware segmentation and tokenisation methods suitable for Arabic, and 2. the unique challenges of data curation in diglossic linguistic settings, where the language spoken at home and in everyday communication (Dialectal Arabic) differs substantially from the written language used in formal education and most media (Modern Standard Arabic). We evaluate our models using the BabyBabelLM evaluation pipeline. We achieved a significant improvement in the grammaticality scores using different tokenisation and data composition strategies. We discuss these improvements in detail, especially what they reveal about the effect of mixing Modern Standard Arabic and Dialectal Arabic in baby Arabic language models, as well as the challenges they pose for building developmentally plausible models in Arabic."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="haider-can-2026-arabic">
<titleInfo>
<title>Arabic BabyLM: Effects of Diglossia and Tokenisation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mozafar</namePart>
<namePart type="family">Haider</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Burcu</namePart>
<namePart type="family">Can</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 Second BabyLM Workshop</title>
</titleInfo>
<name type="personal">
<namePart type="given">Leshem</namePart>
<namePart type="family">Choshen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ryan</namePart>
<namePart type="family">Cotterell</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mustafa</namePart>
<namePart type="given">Omer</namePart>
<namePart type="family">Gul</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jaap</namePart>
<namePart type="family">Jumelet</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tal</namePart>
<namePart type="family">Linzen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aaron</namePart>
<namePart type="family">Mueller</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Suchir</namePart>
<namePart type="family">Salhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raj</namePart>
<namePart type="given">Sanjay</namePart>
<namePart type="family">Shah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="family">Warstadt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ethan</namePart>
<namePart type="given">Gotlieb</namePart>
<namePart type="family">Wilcox</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">979-8-89176-475-0</identifier>
</relatedItem>
<abstract>This paper explores building a baby language model for Arabic in the context of the BabyLM challenge, with a pretraining corpus limited to 10 million words. We focus on two areas: 1. morphologically-aware segmentation and tokenisation methods suitable for Arabic, and 2. the unique challenges of data curation in diglossic linguistic settings, where the language spoken at home and in everyday communication (Dialectal Arabic) differs substantially from the written language used in formal education and most media (Modern Standard Arabic). We evaluate our models using the BabyBabelLM evaluation pipeline. We achieved a significant improvement in the grammaticality scores using different tokenisation and data composition strategies. We discuss these improvements in detail, especially what they reveal about the effect of mixing Modern Standard Arabic and Dialectal Arabic in baby Arabic language models, as well as the challenges they pose for building developmentally plausible models in Arabic.</abstract>
<identifier type="citekey">haider-can-2026-arabic</identifier>
<location>
<url>https://aclanthology.org/2026.babylm-main.18/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>227</start>
<end>239</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Arabic BabyLM: Effects of Diglossia and Tokenisation
%A Haider, Mozafar
%A Can, Burcu
%Y Choshen, Leshem
%Y Cotterell, Ryan
%Y Gul, Mustafa Omer
%Y Jumelet, Jaap
%Y Linzen, Tal
%Y Mueller, Aaron
%Y Salhan, Suchir
%Y Shah, Raj Sanjay
%Y Warstadt, Alex
%Y Wilcox, Ethan Gotlieb
%S Proceedings of the Second BabyLM Workshop
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-475-0
%F haider-can-2026-arabic
%X This paper explores building a baby language model for Arabic in the context of the BabyLM challenge, with a pretraining corpus limited to 10 million words. We focus on two areas: 1. morphologically-aware segmentation and tokenisation methods suitable for Arabic, and 2. the unique challenges of data curation in diglossic linguistic settings, where the language spoken at home and in everyday communication (Dialectal Arabic) differs substantially from the written language used in formal education and most media (Modern Standard Arabic). We evaluate our models using the BabyBabelLM evaluation pipeline. We achieved a significant improvement in the grammaticality scores using different tokenisation and data composition strategies. We discuss these improvements in detail, especially what they reveal about the effect of mixing Modern Standard Arabic and Dialectal Arabic in baby Arabic language models, as well as the challenges they pose for building developmentally plausible models in Arabic.
%U https://aclanthology.org/2026.babylm-main.18/
%P 227-239
Markdown (Informal)
[Arabic BabyLM: Effects of Diglossia and Tokenisation](https://aclanthology.org/2026.babylm-main.18/) (Haider & Can, BabyLM 2026)
ACL