The process of generating metadata for these digital objects was one that I found especially challenging. It was not the actual labor of it so much as the sensitivity and precision I tried to bring to this process, which was tempered by the knowledge that this material was someone’s property. By property, I mean much more than simply having physical custody of the object: these documents are the sum total of a community’s history, a record of violence acted upon them, upon their bodies, and being asked to extrapolate from that history “subject terms” or “descriptions” was a process with which I struggled throughout this class. With that in mind, perhaps my process is not one that should necessarily be emulated — perhaps it was too cautious or mentally laborious to be of much use. However, I do think that offering a multitude of approaches to working with this material might potentially benefit the SCL or at the very least that is my hope in detailing my process here.
I began with the “easiest” objects first (though this is not to suggest any of the material was “simple” or “easy”). By this, I mean those documents that were shorter, had clearly identifiable titles, dates, creators, sources, etc., and in their content were focused on a single issue — though as I later discovered, that central “issue” was often bound up with so much more than the explicit content of the document itself. Beginning this way allowed me to ease myself into the process and build up a measure of confidence with the process before moving onto the more challenging documents. I filled in those fields that were contained explicitly in the material itself and did so for every object I was assigned. This meant revisiting my digital objects multiple times, which allowed me to “proof” my work as I went on. After I had exhausted all the information that was readily available from the document itself, I began to intuit those fields that were not as readily available, e.g. “description,” “subject,” “creator,” and “contributor.” Artifacts that did not have titles were prominent among by assigned objects: I negotiated with this fact by creating brief, informative titles that were as grounded in the text of the artifact as possible. My aim was to preserve the language of the material while remaining clear, concise, and descriptive. Subject terms were perhaps the most challenging field to address: remaining consistent across their application (both within my own objects and with my peer’s work); attempting negotiate between brevity and specificity; highlighting those aspects of the artifact that were truly significant while acknowledging that this would mean relegating some issues to the background. I think it would be helpful to have a single set of eyes review all the objects we digitized and look for consistency among the terms to provide a more unified, coherent descriptive language. This process was very labor intensive and often time-consuming. We spent at least a total of four hours of class time working on generating this material and I can personally attest to hours more spent outside of that to make sure that the work I had done in class was up to par.
I’m not entirely sure I ever felt as though I was “done” describing an object — implicit in this process is choosing to foreground some aspects of the artifact and rendering others invisible. A few sentences hardly seemed enough to encompass the disparate, intricate content of some of the objects I was assigned — how can a twenty page journal be summarized in two sentences? This was in my mind the most difficult part of generating metadata: not ever being satisfied, not ever being certain that I was “done” with an artifact. As I type this blog response, I am reviewing my eleven artifacts to ensure that there is a coherent language, style, and tone to my metadata and descriptors. Most helpful in this process was the collaborative nature of our work, the shared spreadsheets, the in-class work, and having a fresh set of eyes to review my metadata. In making this a communal experience, I felt as though I was able to encompass a range of voices and ideas in my metadata that wouldn’t have otherwise been possible. Also helpful were the readings that had been assigned throughout the semester. The ones I would most recommend were Coates’s Between the World and Me, The Fire This Time, and “Metadata for the Masses.” The first helped establish the stakes of the work we have been tasked with, the second helped contextualize the artifacts in light of Watts 1965, and the third helped me flesh out what exactly metadata “meant,” how it is structured, and how it can serve as an organizing principle for a disparate and eclectic grouping of material.
“Metadata for the Masses” — a section of Tagging: People-Powered Metadata for the Social Web — was invaluable in answering the questions posed in this prompt. Tags can serve as associative, descriptive, or organizational tools: they structure links between resources, aggregate and associate related content, they “label” digital artifacts — and it is a labelling process that has the potential to be part of a “bottom-up classification system” (82). In this way (as is the case of “folksonomies”), tags can establish “ informal relationships” between objects based on “usage patterns” rather than a tightly controlled or regimented set of language (82). The benefit of this is a shift in agency from the people to whom these objects “belong” (i.e. those people who have physical or digital ownership of them) to those who are merely “accessing” the material online. This shift in agency is one that I see as being in line with the SCL’s mission to cater not only to academics who use the archive for scholarly research but to the community itself. While this certainly can be construed as a beneficial consequence of allowing “social tagging,” there are naturally caveats that come with this shift in agency: it is never safe to assume that the people who can and would access this material (and thus contribute to the tagging system) are necessarily part of the community or sensitive to the issues endemic in these objects. The internet, as I’m sure everyone is aware, is populated by people who are prejudiced, racially insensitive (to put it mildly), or simply looking to disrupt productive conversations about race in America. I would then suggest a modified social-tagging system in which the language and tags “available” to the public are decided by the SCL but which tag is assigned to which object is left to the people who access the material online. Tagging has been an integral part of our own work on this material, specifically in terms of “descriptive,” “resource,” and “source,” but a tagging system with “opinion” and “play & performance” type tags would work well alongside the “tagging” that has already taken place (67). It would allow for further collaboration between the library, community, and its users as well as give voice to those people who have been implicitly excluded in this metadata-generating process.