Mid Sweden University

The purpose of this study has been to investigate how the Swedish national archives have dealt with how authenticity and context are affected when techniques in the field of HTR (Handwritten Text Recognition) have been used against handwritten archival materials within the Swedish National Archives. The HTR-project was carried out with the help of citizen scientists and I have examined how authenticity and context has been safeguarded in the project. 

I've used a qualitative methodology with a case study as a method for this study, with the empirical material consisting of interviews with three employees at two different archival institutions in Sweden, along with a study of documents relating to this project, as well as a literature review relevant to my study.

I have used Terry Eastwood’s ” What is Archival Theory and Why is it Important?” (1994) as a theoretical framwork for analysis. I have applied Eastwood’s thoughts on authenticity and context as two of the characteristics of archives, to the new data that is created through HTR to discuss whether authenticity can be guaranteed and how context can be protected to preserve the interdependency between records within the archive. 

The results of the study show that authenticity have been top of mind during the project at the Swedish national archives, but that there still are questions regarding how to best present the transcribed data to the public. Combining the AI-tools for HTR with citizen science has been a success, even though citizen science has its very own set challenges like distribution of more mundane tasks and keeping citizen scientists motivated throughout the entire project.