Out-of-the-Box GPTs for Lexicography

  • Gilles-Maurice de Schryver BantUGent & LT3, Ghent University, Ghent, Belgium; and Department of African Languages, University of Pretoria, Pretoria, South Africa (https://orcid.org/0000-0001-7272-9878)
DOI: https://doi.org/10.5788/35-2-2103

Abstract

A detailed analysis is offered of how ChatGPT-3.5 has treated the world's languages for lexicographic purposes, starting with hyper-visible English, moving to semi-visible Portuguese, and ending with low-visible Bantu languages. Case studies for the latter group are region-visible Swahili, nation-visible Xhosa and local-visible Lusoga. This impact-based scale is reflected in the uptake of GenAI technology for lexicography. The results of using out-of-the-box ChatGPT-3.5 are fantastic for English, fake for Portuguese, and simply fail for the Bantu languages. But there is a way out, and the seeds for that are sown in this article. Keywords: generative artificial intelligence (GenAI), large language models (LLMs), generative pre-trained transformer (GPT), out-of-the-box GPTs, lexicography, dictionaries, lexicographers, English, Portuguese, Bantu, Swahili, Xhosa, Lusoga
Published
2025-12-09
How to Cite
de Schryver, G.-M. (2025). Out-of-the-Box GPTs for Lexicography. Lexikos, 35(2), 362-396. https://doi.org/10.5788/35-2-2103
Issue
Vol. 35 No. 2 (2025)
Section
Artikels/Articles

Copyright of all material published in Lexikos will be vested in the Board of Directors of the Woordeboek van die Afrikaanse Taal. Authors are free, however, to use their material elsewhere provided that Lexikos (AFRILEX Series) is acknowledged as the original publication source.

Creative Commons License CC BY 4.0