Language Technology Research and Input from African Linguistics: Prospects and Challenges

The buzzword in technological research today is AI (Artificial Intelligence). AI is characterized by the ability to learn, deep learning, cloud computing and data ingestion. At the center of AI are Machine Learning (ML), Deep Learning (DL) and Natural Language Processing (NLP). Natural Language Processing straddles all of the other two activities of AI. Our concern, in this paper, is to show that the activities of AI can be enriched if more attention is paid to how natural African languages are processed. African languages pose a unique set of problems for AI. Indeed, one of my students has shown that lack of African language input into AI manifests in, often hilarious, auto-translation of African texts by Chat GPT and Google’s neural machine translation technology. In addition, as pointed out to me by a computational linguist, Dr. Tunde Adegbola (personal communication), African languages pose interesting puzzles for Computational Linguistics in the areas of instrumental and quantitative studies of tonality, in the inadequacy of traditional methods of automatic detection of morphemic boundaries, in handling typically African morphological processes such as reduplication and interfixation as well as the commonly manifesting syntactic features of splitting verbs and verb serialisation in African languages. At the application level, the paper explores an attempt by our team to ameliorate a socio-economic difficulty in a Nigerian community using an interdisciplinary outcome. As an agenda-setting paper, it is hoped that the thoughts presented in this paper will open new vistas of collaborative research of an interdisciplinary nature between Linguistics and Computer Science.