Women’s History, and . . . Metadata?!

Toward the end of March I had the wonderful opportunity to present at an inaugural conference on the subject of Women’s History in the Digital World (#WHDigWrld on Twitter), at the Albert M. Greenfield Digital Center for the History of Women’s Education (at Bryn Mawr). Not only did this event allow me to meet many people doing wonderful work with digital women’s history, but it also allowed me to bring together people I know from two different projects: HistoricDress.org, and the digital library at Vassar College.

Please encode us with dignity! Early students at Vassar College; photograph courtesy of Vassar College Archives and Special Collections

The event began Friday night, and I arrived just in time for the introductions and the keynote by Laura Mandell. Her talk got us thinking about “Feminist Critique vs. Feminist Production in Digital Humanities.” This overall theme carried us through the entire weekend, reminding us of the need for feminists:

• to be at the table when systems are designed for collecting, encoding, and disseminating information
• to create projects that provide positive models of how to represent less privileged groups and individuals
• to peer review each other’s work
• to provide support for each others’ grant proposals.

The highlight of her talk for many of us, however, was a slide with an example of some of her XML encoding for one of her projects. This is code that holds metadata (“data about data” is the quick definition) about a person, place, text, object – anything, really – and allows information to be understood by a computer. There was a collective gasp from the room when she showed that the TEI (Text Encoding Initiative) preferred code for gender (when encoding a person) was 1 for male, 2 for female.

A tweet from @ProfessMoravec during Laura Mandell’s keynote

I’ve been processing this ever since, thinking about who was at the table when that standard was developed, and why it couldn’t have been m and f, or something else entirely. Eventually it occurred to me that it could be an issue of internationalization, using numbers to avoid one language’s terms. Sure enough, looking into it further, it’s based on an ISO standard, ISO 5218, to be specific. Still, I couldn’t help but keep thinking of different ways to approach this issue – what about other non-serial numbers so it wouldn’t feel as hierarchical? How else could we approach this?

Well, checking in on this again today, a quick search yielded a discussion board conversation on this topic at the SourceForge site for TEI documentation, begun by Melissa Terras. Following through this months-long discussion thread, it appears that in the weeks since #WHDigWrld, the TEI Council has agreed to make a change, allowing the use of alternative or locally defined systems for encoding gender (take a look at the updated specification). Great news, but there’s still much work to be done.

Indeed, Mandell’s XML slide served as an example of how to perform subversive encoding to simultaneously work within current systems and create new systems. Her example pointed out that predominant name authorities, such as the Library of Congress (LOC), sometimes define a woman’s preferred name in the format “Mrs. (insert husband’s name here).” Her slide provided an example of double/triple subversive encoding, including the ISO 5218 and LOC standard terms, but only as alternate terms, following terms and ontologies more appropriate to the given project and to the representation of women as primary figures. It is inspiring to imagine how our projects can meet current standards and interact with other existing projects, yet simultaneously set new standards for like-minded work which could gain traction and someday overtake our current hegemonic standards.

All of this re-opened a door in my mind. I’m in the middle of a Metadata class right now, and up to attending this conference, I had been (for the most part) uncritically accepting the practice of using established authority files for names, places, and subjects, and accepting the preferred terms from such authorities. Disambiguation is the name of the game, and I understand and embrace that. But after seeing Mandell’s example, I am reminded that any “preferred” term must be coming from a preferred group, one that speaks from a position of power and privilege that other groups and other terms cannot attain.

Now that I’m back in school, I’m reminded of how easy it is to just be one of the herd and follow along with what you’re being taught in class. Back when I was getting my first masters, in costume design, my cohort jokingly adopted sheep as our mascot. Of course back then it was more obviously a joke – as a designer you are expected to have a style of your own and not to blend in with the herd. As a librarian, not so much – though I’m lucky that my program at SU seems to be very supportive of discourse. We need to see more examples like Mandell’s, to think more critically about the systems of organization that we are working within, to create alternative systems when necessary, and to stand up and try to make changes in existing standards.

The rest of the weekend continued in this vein, with a wide assortment of wonderful projects from across the country. I’ll highlight the presentations I was able to attend, but please visit the conference website to read more about all the ones that I unfortunately had to miss.

• The perfect start for me on Saturday was a presentation by Patricia Keller about the Sampler Archive Project. Pat was also involved in the Quilt Index, and I’ve enjoyed learning from Pat about both these two projects, which have been pioneers in presenting focused collections of material culture. My team at HistoricDress.org can learn much from Pat’s work on these projects.
• Bridget Baird and Cameron Blevins (a mother/son team!) presented Digital Diaries, Digital Tools: A Comparative Approach to Eighteenth-Century Women’s History. They provided a great introduction to topic modeling, and its pros and cons in application to the diaries of Martha Ballard and Elizabeth Drinker. I was reminded of how the entries in many historic diaries are not unlike the mundane tweets of which we are critical today. They also made me think about how calendars/planners now carry much of the data that diaries did then – I wonder how they will be studied in the future?
• Jen Palmentiero (from Hudson River Valley Heritage), and  Joanna DiPasquale and Laura Streett (from Vassar, where I’m very lucky to have their support and advice for my own project) presented Using Archives and Metadata to Uncover Women’s Lives: Challenges and Opportunities for Scholarship through Archives and Digital Libraries. Their 3 presentations worked perfectly together,  showing the trajectory of tech development to use. Several of their examples demonstrated how technological choices are impacted by factors that are decidedly non-technical, such as funding and institutional culture. Laura’s examples of findings in the digital archive indicated how digital search and discovery can lead to the consideration of different search terms and other practices moving forward.
• Erin Bush, a PhD student at George Mason University, presented Beyond a Reasonable Doubt: Executions of Women in the United States, showing how she is using data science both to help frame her research questions and to try to answer them. Considering the questions she was asking of this particular dataset, she got me thinking about what questions we can ask of objects, given the datasets that are coming available as museum collections are beginning to share APIs for access to data. What questions will a PhD student in 10 years be able to ask of the data coming out of the APIs of the DPLA, or of the projects I’m working on, like HistoricDress.org? What different metadata do we need to be collecting about objects to be able to ask different, better questions? I tried to touch on this briefly in my own presentation, a little later.
• Monica Mercado presented On Equal Terms? The Stakes of Archiving Women’s and LGBT History in the Digital Era, raising important issues around some of the tensions of building digital collections related to gender and sexuality history.
• Mia Ridge presented New Challenges in Digital History: Sharing Women’s History on Wikipedia, a subject I’ve been very interested in, though I regret that I didn’t have the time to participate in the #tooFEW (Feminists Engage Wikipedia) project through THATCamp Feminisms last month. A highlight of her talk, for me, was the simple act of changing a name to red to indicate that there is no dedicated Wikipedia page for that person (making the archival silence visible, as it were), and then the power of changing that red to blue by completing a page for that person.

In our own presentation we ran into some tech difficulties which resulted in me presenting from my laptop screen to a crowded table, but our wonderful moderator, Marla Miller, kept us on track, and we made some great connections in our packed room, with a wonderful (though too brief!) discussion. I shared my work on Fashioning an Education: 150 Years of Vassar Students and What They Wore, which you can also see as a Prezi.  Next, Astrida Schaeffer presented on The New Hampshire Historic Dress Project. Finally, Kiki Smith closed our session by discussing our goals, and our work so far, on HistoricDress.org.  This was a great opportunity for members of the HistoricDress.org team to connect and continue to brainstorm our next steps, and even to grow as we identified new partners among conference goers.

Throughout the weekend there were concerns both for the segregation of women’s history and for the failures of “add women and stir” models. This has had me thinking about how historic costume collections are both victims of sexism and examples of reverse sexism – men’s clothing artifacts are preserved much more rarely than women’s. Of course, this means that costume collections aren’t respected in broader circles, as they are considered feminized and therefore superficial. Digital costume projects have great potential to use artifacts of clothing to fill in some of the narrative of women’s history that is missing, but we have to challenge the stereotypes of superficiality at every turn.

Michelle Moravec summed it up well with the title for her blog post following the event – Was Women’s History in the Digital World the First Berks of Digital Women’s History? Indeed, I hope to be able to attend the 2014 Berkshire Conference of Women Historians, and to see many from #WHDigWrld there, and to return to Bryn Mawr for a 2nd conference on this theme in the future. Many thanks to Jennifer Redmond and her team at the Greenfield Center for bringing us all together.