Data-Driven Research for Film History: Exploring the Jean Desmet Collection

Abstract

The article discusses the results of the research project “Data-Driven Film History: A Demonstrator of EYE's Jean Desmet Collection” (2014–15). With a team of media scholars, archivists, and software developers, we conducted data-driven research to explore the affordances and pitfalls of digital methods for mining the resources of film archives and researching (early) cinema history. Working with the collection of cinema owner and distributor Jean Desmet, preserved at the EYE Filmmuseum, we used digital tools for the specific purpose of visualizing distribution networks and color patterns in film programs in the Netherlands during silent cinema's transitional years. The article discusses the lessons we learned from the project, both as scholars and as archivists. It opens with a brief introduction to the Desmet Collection and then considers our motivations, on one hand, for exploring digital research methods and, on the other, for choosing a particular set of tools. Subsequently, it discusses the mapping tool we developed as part of the project, reflecting in particular on the metadata issues it raised. In the final part of the article, we recount our experience of experimenting with tools for color analysis, paying particular attention to how the color patterns they generate might ultimately be combined with a data mapping and what sort of research possibilities this might in turn open up. The article demonstrates, among other things, how crucial it is for researchers to understand the provenance, affordances, and limitations of the data at their disposal and to make the discrepancies between them visible and transparent for others–including archivists.