The detection of authors and scribes of manuscripts, based on philological and palaeographic criteria, has a strong research tradition in (Slavic) historical linguistics. Traditional philology relies on time-intensive methods like contextual research, close reading, transcription, and critical editing to study old manuscripts. In contrast, modern image and pattern recognition can automatically identify handwriting styles and graphic patterns. This project compares both approaches, exploring their strengths and how they can be combined. Manuscripts by unknown scribes offer an ideal testing ground for integrating computational and linguistic methods. It’s crucial to distinguish between the scribe (the manuscript’s physical writer) and the author (the original source of the text). Handwriting features belong to the scribe; linguistic features reflect the author. We present the assistance tool LiViTo and our case study to discuss the application of the tool and potential for author/scribe detection in a mixed methods approach.
Meyer et al. (Wed,) studied this question.