Designing tools for a ‘Ajami Digital Philology: first results and further perspectives

Guest lecture with Oreen Yousuf (PhD student, Uppsala University).

'Ajami is the general term referring to the collective one-thousand-year-old manuscript tradition of texts in very different African languages written in the Arabic script. As thousands of ‘Ajami manuscripts were produced and copied across Africa, to foster research on this so far neglected part of Africa’s cultural heritage, there is an urgent need to digitize the manuscripts, create automatic recognition models for the diverse languages involved, and to perform natural language processing tasks on the digitized texts. The development of these digital tools will allow to more robustly analyse the diachronic and synchronic aspects of the languages in which the texts are written and shed fresh light on the cultural landscapes in which both the manuscripts and the texts were produced. In this presentation I will fist provide an introductory background of Ajami manuscript cultures, describe some prevalent orthographic characteristics of Ajami texts, and then delve into my work on Handwritten Text Recognition (HTR) for ‘Ajami manuscripts which is going to set the technical foundation for the development of a full-fledged ‘Ajami digital philology.