This photograph was used to publicize Italy’s failure to pass the legge Fiano (Fiano’s Law) outlawing Fascist and Nazi propaganda in 2017.
Postwar Accountability: The Italian Legacy examines the long reckoning over more than 70 years for violations of International Humanitarian Law against civilian populations in Europe during World War II. Should states be able to claim immunity from prosecution for international crimes, especially in light of the competing demand to protect universal human rights? The current “neofascist wind” gives this question new urgency, as some American scholars have called the reemergence of far-right political agendas across Europe and the United States.
In 2016-17, the directors of Postwar Accountability built a first experimental prototype of a future digital archive about these issues with the support of a Digital Humanities grant. In consultation with technology experts from UC Berkeley, the California Digital Library, and research libraries at Stanford and Harvard Universities, they became familiar with current standards for digital archiving today. Why do institutions elect to build a custom Content Management System (CMS) rather than buy one out of the box? They studied best practices for digitizing large data sets, evaluated Optical Character Recognition (OCR) software, and reviewed security issues, data storage, and many other topics. They gave special consideration to the question of metadata, settling on Encoded Archival Description (EAD). Later they discovered a new metadata schema created specifically for European legal materials, and they may consider this option instead should it prove to be versatile.
Meanwhile, they developed a mobile protocol for producing high-quality digital photographs with the help of a professional photographer. They learned to capture fragile documents under varying lighting conditions, while comparing the advantages of raw and compressed files. They photographed a sample of records, refined the digital images in Adobe Photoshop, and then ran the files through Abbyy Finereader and Tesseract. Because all the results required excessive correction, they concluded that neither program would be scalable for a data set of any substantial size. Subsequently, they arranged to test a promising new ORC app that is being developed in Asia.
Because security is a major concern for this project, an IT advisor recommended that the team use Django (in the Python environment) to build their prototype. Omeka was their first choice, but Django reportedly has much better security. An undergraduate computer-science major was hired to write custom code. Initially, the coder experimented with a simple model in Omeka, but then developed a more robust model in Django. The final product included such standard features as a home page with Welcome and About messages; a menu bar; headers for materials presented chronologically and grouped in the following columns: “Case | Date | Victims | Perpetrators | Trial”; and an active carousel with a slide-through feature that presented documents in an interactive way. The horizontal motion of the carousel was designed to increase visual interest in contrast to the vertical, stationary columns. Materials included legal records, geographic maps, historical photographs, and one film clip with eye-witness testimony. (English translations of Italian and German documents and other annotations were planned as well.) Some of these materials could be discovered by clicking first on the photographs.
Other designs were considered as well, and a second undergraduate with a background in WordPress was hired to develop the front-end. An IT consultant informed the team that it would be possible to integrate WordPress with Django. Nonetheless, after much trial and error the team discovered that this was incorrect. At the minimum the designer probably would have needed to work in HTML, CSS, and maybe JavaScript code, particularly if HTML and CSS files were needed for the interface in order to make the connection. The team reviewed other technical solutions as well but did not find a cost-effective way to integrate the two platforms, particularly since very little documentation existed about how to do so successfully.
This experience highlighted the importance of documentation. Consequently, the directors began to consider open-source software for their future digital archive, such as Blacklight, Hydra, and Ruby On Rail together with Apache Solr (due to the latter’s faceting capability), or a similar combination. Meanwhile they worked on a visual Story Map to introduce the project.
In sum, the opportunity to prototype turned out to be very useful. The Digital Humanities grant made it possible for the directors to evaluate important technical decisions and creative concepts used in digital archiving today. They made real progress in figuring out how to capture a large cache of paper documents and translate them into the digital environment. In addition, they gained valuable insight into the cost and feasibility of designing a new CMS. This experience puts them in a better position to assess and plan the steps they must take in order to succeed in building their archive.