The role of the library when computers can read: Critically adopting Handwritten Text Recognition (HTR) technologies to support research

Computational approaches to processing and searching images of historical manuscripts by Handwritten Text Recognition (HTR) is one of the most promising machine learning approaches for academic research in the humanities, having the potential to transform access to our written past for the use of researchers, institutions, and the general public. This chapter surveys the current use of HTR in the library sector, highlighting major tools currently in use and activities being undertaken by academic and research libraries. Using Transkribus as a case study, the chapter provides examples of where libraries have successfully deployed HTR and will focus on emerging issues for incorporating the application and results of HTR into a digitisation workflow (including documentation, results delivery, and sustainability). The chapter will consider how HTR can be best deployed to support researchers, including the need for transparency, training, and data infrastructure. Although HTR technology is now reasonably mature, Academic Libraries need to adopt this machine learning technique in a critical way, signposting the data in a way that explains its creation, and allows its embedding into historical practice, to best support their user communities.