On automating editions: The affordances of Handwritten Text Recognition platforms for scholarly editing

Recent developments in Handwritten Text Recognition (HTR) mean that automated editions – presentational editions generated from both digital images of text, and their corresponding transcriptions created by artificial intelligence – are now available to adopt, adapt, and critique. This paper responds to an absence within scholarly editing literature regarding HTR. HTR is a machine-learning approach that creates accurate transcriptions of images of handwritten documents. We highlight developments in text recognition technology, demonstrating that automated standardised editions are no longer a future possibility, but a reality necessary of consideration within a scholarly editing framework.

We do this via a case study of creating a standardised online edition in the HTR platform Transkribus of the manuscripts of Marjorie Fleming (1803-1811), a Scottish child author who became posthumously famous for her free-thinking and precocious diaries. As well as providing a cost-effective way to generate machine-processable transcripts at scale, Transkribus can now generate digital online editions via its ‘read&search’ platform. This provides an efficient mechanism to share and search digitised texts, bypassing previous procedures and disrupting established processes for data formatting, hosting, and delivery of online editions. However, we show that while read&search can be considered a scholarly digital edition, it needs further development to be encountered as a critical digital edition, providing suggestions for ongoing development. Automating the process of creating scholarly digital editions will encourage others to create them, democratising the digital edition landscape, although we reflect on the ramifications this may have.