Fragile Facts in Digital History: Fakes and Errors or Risk and Opportunity?

Digitization has significant implications for the status, availability, and linking of historical facts or data, and the resulting scientific communication: on the one hand, access to many sources has improved through digitization, while on the other hand, poor data quality and stability, inadequate reproducibility of results, and the absence of standards pose a threat to scientific quality. Although the problem has been addressed on various occasions and the need to address the reproducibility and replicability issue is acknowledged, there is a lack of any form of systematization - as is already present in the natural and technical sciences in the form of Standard Operating Procedures (SOP) - or even of considerations for a replication methodology. In digital historical sciences, there is neither a form of error classification nor a systematic examination of known but generally overlooked sources of error.

The section aims to initiate a discussion of the hitherto little-discussed lack of reproducibility and replicability of results and studies in digital historical sciences and to discuss possible solutions.