Legibility Machines: Archival Appraisal and the Genealogies of Use

Abstract

The web is a site of constant breakdown in the form of broken links, failed business models, unsustainable infrastructure, obsolescence and general neglect. Some esti­mate that about a quarter of all links break every 7 years, and even within highly curated regions of the web, such as scholarly publishing, rates of link rot can be as high as 50%. Over the past twenty years web archiving projects at cultural heritage organizations have worked to stem this tide of loss. Yet, we still understand little about the diversity of actors involved in web archiving, and how content is selected for web archives. This is due in large part to the ontological politics of web archives, and how the practice of archiving the web takes place out of sight at the boundaries between human and technical activity.

This dissertation explores appraisal practices in web archives in order to answer two motivating research questions: 1) How is appraisal currently being enacted in web archives? 2) How do definitions of what constitutes a web archive shape the practice of appraisal? In order to answer these questions data was collected from interviews with practicing professionals in web archives, and from a year long ethnographic field study with a large federally funded archive. Method triangulation using the­matic analysis, critical discourse analysis and grounded theory generated a thick and layered description of archival practice. The results of this analysis highlight three fundamental characteristics of appraisal in web archives: time, ontology and use.

The research findings suggest that as expressions of value, appraisal decisions do not simply occur at discrete moments in the life cycle of records. They are instead part of a diverse set of archival processes that repeat and evolve over time. Appraisal in web archives is not bound by a predefined assemblage of actors, technologies and prac­tices. Indeed, artificially limiting our definition of what constitutes a web archive truncates our understanding of how appraisal functions in web archives. Finally, the valuation of web records is inextricably tied to their use in legibility projects, where use is not singular, but part of a genealogy of use, disuse and misuse.

Appraising appraisal along these three axes of time, ontology and use provides in­sight into the web­ memory practices that condition our understanding of the past, and that also work to create our collective present and futures. Explicitly linking appraisal to the many forms of use informs archival studies pedagogy, by establish­ ing the value of records in terms of the processes they participate in, rather than as a static attribute of the records or their immediate context. As machines increasingly become users of web archives the stakes for understanding the values present in web archival platforms could not be higher.