Transforming the Story of the Arabic Book, 700–1800: Knowledge, Information Technology, and the Arabic Book

The premodern Arabic textual tradition between 700 and 1800 is one of the most prolific in human history, with the number of surviving works surpassing the extant production of any other pre-print culture except perhaps imperial China or Sanskritic India. These texts can now be studied in completely new ways using pioneering digital technology that can identify textual units displaying what one scholar has termed ‘extended textual similarity’. This means not only word-for-word matching, which can be detected by existing techniques and has been computationally studied, but also textual overlap obscured by transformations such as paraphrase and translation. By identifying, documenting, and interpreting this kind of transformational reuse across thousands of Arabic books, KITAB-Transform’s team of eight historians, linguists and computer scientists will uncover hitherto obscured connections between texts and shed light on the writerly attitudes and practices that created the Arabic written tradition and continue to shape Islamicate societies’ reservoirs of memory today. The project relies on a vast corpus of digitised texts that continues to grow thanks to an OCR pipeline developed by team members. KITAB-Transform will produce interdisciplinary monographs and edited volumes, journal articles, blog posts, language models, and data sets, as well as a web application through which less technically inclined scholars can access and work with our data, supported by a robust outreach programme. Our work will make an important contribution to Arabic book history and help redress a significant imbalance in the development of large language models, which favour English and modern social media but are comparatively impoverished for historical languages.