Renaissance Women, Text Encoding and the Digital Humanities: An Interview with Julia Flanders

Julia Flanders is Director of the exemplary Brown University Women Writers Project and Associate Director for Textbase Development at the Brown University Scholarly Technology Group. With those projects and as Editor in Chief of the Digital Humanities Quarterly, due to launch in 2007, Julia is a key figure in humanities computing and text encoding initiatives. Academic Commons recently caught up with her to talk about her various projects.
Academic Commons:  You’ve been involved with the Brown Women Writers Project since 1992. What are the most important developments for WWP in the past several years? What’s ahead long-term for the WWP and projects like it?

Julia Flanders: In a sense, all of the important developments we’ve had in progress lately have come to fruition this year. The project released a new version of Women Writers Online this past summer, with much faster searching and a new interface. We’re now using the Philologic search engine (from the University of Chicago) which provides a lot of very interesting new functionality, particularly things like text analysis tools which we haven’t been able to offer before. Most importantly, since this is open-source XML software, it’s easier than before to experiment with interface ideas; we’ll be launching a “sandbox area” this year in which we can offer some unusual interface tools for people to play with.

This winter, we’re finishing up the WWP’s Guide to Scholarly Text Encoding, which will be published online in 2007; it will provide in-depth guidance for non-technical scholars who want to learn more about text encoding and the Text Encoding Initiative (TEI), or just understand what it is and why it’s interesting. We also just learned that we’ve received funding from the NEH to offer a two-year series of workshops and seminars on scholarly text encoding, aimed at humanities faculty. This will draw on the Guide and give us an opportunity to reach a wider audience. We’re also pursuing a few other new projects: we’re seeking funding for a project which would explore the implications of providing annotation tools for readers of Women Writers Online. Annotation is a familiar idea but we think that in practice it might take our readers in a number of experimental and challenging directions: for instance, developing curricular materials that are linked directly to the texts, or writing hypertextual critical essays that consist of annotation sequences. There are a number of hard questions to address, particularly concerning issues of peer review and the technical longevity of the annotation system, but I think we need to treat these as challenges rather than obstacles.

Long-term developments for Women Writers Online include experimenting with more ways for readers to see and explore textual pattern, through visualization tools. As text collections scale up, familiar narrative reading processes become harder to apply, at least as the first stage of research; it’s helpful to have ways of seeing the whole collection and grasping its patterns as well as focusing in on individual texts. A lot of interesting work is being done in this area: the NORA project has been developing tools for data mining and visualization, and the TAPoR project in Canada has been creating a portal for text analysis that among other things offers experimental tools that can be used through TAPoR or incorporated into local project interfaces. Some of this work may benefit projects like the WWP directly and some of it may inspire further development; we’d like to take advantage of open-source efforts like these and test them out on the WWP collection.

Academic Commons: What’s your sense of how faculty are using digital resources like WWP in their research? What kinds of changes are happening in their work, and what kind of obstacles are they facing?

Julia Flanders: At the moment, I think they’re using digital collections in much the same way as they use collections of printed books: to find documents they’re interested in and to read them. Searching helps to speed up this process; online access makes it more effortless and exposes readers to a wider range of material. But habits of reading are not yet changing very much.

The biggest obstacle is the granularization of online resources, and the lack of cross-collection analysis functions. This is a problem partly because of funding and intellectual property issues, but also because it is something fundamental about the incunabular stage of electronic publishing we’re still in. Different projects are experimenting–appropriately!–with different kinds of markup, different approaches to representing materials in digital form. Those differences pose challenges for integrated searching, but they also represent important explorations into digital modeling. Tools like the Open Archives Initiative (OAI) are making it increasingly possible to find items across digital collections, but I think the more detailed analysis functions will have to wait until a further stage in the history of electronic publishing.

Academic Commons: As you note, we might see the early electronic editions as a sort of incunabula, a transitional form of text as we move toward something more stable. Is that a reasonable way of seeing the developments in digital texts, and, if so, what’s the most important aspect of the new form of text?

Julia Flanders: We are at an incunabular stage in the emergence of electronic editions, though I don’t think what we’re moving toward is more stable than what we have now. Over the past decade, I think some important fundamental practices have emerged: the use of XML (and probably some form of TEI) for the transcriptions of the text; the use of page images, linked to those transcriptions, to provide additional information; and, depending on the kind of edition, the printing of base texts with several parallel versions aligned together, providing a single text with variant readings encoded in the text stream.

Using this basic framework, a great many different surfaces can be produced, and I think that’s where you now see the greatest variation: in the behavior of the interface, the ways readers are invited to prod at the text, the kinds of information they are invited to manipulate or inspect. There’s still experimentation and research going on with respect to what’s under the hood (i.e. the encoding of the text), but I think the basic ideas there are pretty solid.

Academic Commons: You’ve been offering Text Encoding Initiative workshops for faculty. If I’m a humanities professor in a liberal arts college why might I want to learn this stuff? How is familiarity with text encoding standards–a somewhat arcane subject, we might agree–going to change my scholarship and teaching unless I’m heavily invested in preserving old texts?

Julia Flanders: I think there are a number of reasons why people take these workshops. It used to be that we saw faculty who had gotten involved in a digital project–either as a project advisor, or as the founder or editor–and wanted to understand how the encoding worked. For faculty in this position, there’s a clear motivation, even if they aren’t planning to do any of the encoding themselves: the editorial decisions they make about how the text will be represented all require some understanding of encoding (not at the technical level but at the conceptual level).

But more recently, we have started to see participants who have no particular project in hand but simply want to understand text encoding because it is methodologically so central to modern editing and, by extension, to modern textual scholarship. I doubt they would describe themselves as being interested in “text encoding standards” any more than they would have described themselves fifty years ago as being interested in “editing standards.” What they want to know–and what these workshops emphasize–is how text encoding works as a representational system: what it lets you say about texts, what assumptions it makes about how texts work, how it fits in with current scholarly practices. Given how many digital resources now are based on XML-encoded texts, some understanding of these technologies and methods is as important as understanding scholarly editing (at a basic level) would be for someone using a scholarly edition in their research. Even faculty who have no special interest in preserving old texts nonetheless use these materials in research and teaching: digital resources like ProQuest’s Literature Online and Early English Books Online collections are in some ways the modern equivalent of the Norton Anthology, the Oxford Classics and similar print sources. Faculty can’t instill a critical approach to texts in their students if they have no idea how the very sources they’re using are produced.

Academic Commons: You’re the Editor in Chief of the new Digital Humanities Quarterly. What’s the schedule for DHQ and what do you hope to accomplish?

Julia Flanders: DHQ will launch in 2007. What we’d like to accomplish is a gradual but persistent experimentation with the scholarly journal form: publishing peer-reviewed, high-quality articles on digital humanities, while offering a new range of ways to read the field and explore connections between articles. We’re also planning to offer some additional publication modes: editorials, reviews, blog entries, and interactive media pieces, plus the opportunity for reader commentary and discussion, so that the journal can represent the same kind of intellectual give-and-take that makes conferences so engaging. What will be less visible, but perhaps will have a greater impact over the long term, are some of the ways we hope to challenge traditional journal publishing assumptions. For instance, authors will retain ownership of materials published in DHQ. The journal will also be open-access, to help expose the field of digital humanities research to a broader audience. We’re hoping, above all, that the journal will help foster greater cross-pollination between digital humanities and the traditional humanities disciplines.