Computing Detailed Colexifications with Missing Data Information from the CLICS⁴ Collection

2026/02/18 by David Snee, Johann-Mattis List · 1 voice
Computer Science · Social Sciences · #Language and cultural evolution #Natural Language Processing Techniques #Topic Modeling

paper · doi:10.15475/calcip.2026.1.2

openalex publication_date 2026/02/18 · openalex created_date 2026/02/19 · openalex updated_date 2026/07/15

Abstract

CLICS⁴ offers a refined structural representation of cross-linguisitic colexification patterns but retains an implicit representation of missing data. This obscures whether the lack of a colexification in a language for paired concepts is due to its true absence in the language, or due to missing data on the concept or word form level. We introduce straightforward workflow that can be applied to individual dataset from CLICS⁴ to identify cases of colexification via a three-way attestation scheme. Our approach captures the presence or absence of a colexification in CLICS⁴, but it also explicitly encodes the presence or absence of data at the level of the original questionnaire, or the individual language, elicited with the help of the questionnaire.

Discussions

• New contribution to our Blog / Journal on Computer-Assisted Language Comparison in Practice, this time by our doctoral student David Snee and myself: Computing Detailed Colexifications with Missing D [bsky, 2 points, 0 comments]

Related

• Inference of Partial Colexifications from Multilingual Wordlists
• The Database of Cross-Linguistic Colexifications, reproducible analysis of cross-linguistic polysemies
• Advancing the Database of Cross-Linguistic Colexifications with New Workflows and Data
• Partial Colexifications Improve Concept Embeddings
• Colexifications for Bootstrapping Cross-lingual Datasets: The Case of Phonology, Concreteness, and Affectiveness