Unimagined Worlds of Digital Humanities and Textual Editing

At what must have been a wonderful conference, Social, Digital, Scholarly Editing, at the University of Saskatchewan this past July, Peter Robinson delivered a paper in which he basically called to end collaboration between digital humanists and textual scholars. A “common error digital humanists make … is this: that digital humanists may usefully collaborate with textual scholars. After twenty years of working with my digital humanist hat on, as a collaborator with textual scholars, I can tell you now: this collaboration is a mistake … We should seek, as soon as we can, to end this collaboration.”

My response here argues that as long as humanities scholarship values texts, and as the notion of text itself continues to change, and as scholars increasingly want to feed more texts into new digital tools, there can be nothing more important than continuing collaboration between digital humanists and textual scholars. I will not make an argument for social editions, access, openness, data, reusability, or other principles of the digital humanities that have been argued elsewhere. My argument for collaboration stems primarily from the fact that because each community has the potential to push the other to pursue research, tools, research agendas, and new ways of thinking about text that neither group can imagine on their own. This collaboration will ideally result in more traditional text editing being taught as part of the digital humanities, and more media studies being taught as part of text editing. That fusion will emerge from collaborative projects.

The gist of Robinson’s engaging and thoughtful paper is that digital humanists who aren’t textual editors shouldn’t be doing textual work (creating online archives, texts, crowd-sourcing transcription, etc) because they think they know something about text editing, but don’t, and therefore aren’t upholding the rigorous standards that text editors rightly and thankfully maintain that makes much work in the humanities possible. Robinson justly complains, for example, about texts stuffed into wikis and other platforms that deliver neither sufficient degree of openness nor quality texts–truly a disservice to the text and its consumers.

Robinson's vision for the future of textual editing, as he has expressed elsewhere (and to some extent in his anti-collaboration paper as well), is to be highly praised, and he might be fairly characterized as one of the most forward thinking textual scholars. This makes his plea—especially how it appears to instigate (propagate?) a DH (Digital Humanities) and TE (Textual Editing) turf war—seem even more bizarre. In an effort to bolster and defend the important role of the editor, he has sacrificed interdisciplinary collaboration and imagination at the Lachmannian altar. Textual scholars should control their tools and their texts, thank you, and will certainly---of course!---avoid the disciplinary blinders that come from traditional training and vocational expectations and impositions.

Needless to say, the DH and TE dichotomy that Robinson employs for rhetorical effect is not rigorous in terms of real people and their projects, as Robinson’s own work illustrates. But I will continue to use his binary categories because I do agree with his implicit assumption that the large majority of the constituents of each of those communities are united by similar a philosophical and value framework, each of which is somewhat opposed to the other. It’s that productive unease that can get them out of well-worn ruts. There are very different skill sets at work. Both sides need each other because of the way traditional graduate training works, namely the necessity of specialization. The kinds of technologies, tools, interfaces, and social changes necessary to enable Robinson’s vision for a robust text environment are not unique to text editing, but have strong parallels in many other DH projects, and not just those with a strong textual component. Just because DH projects do not exhibit (or need) the same textual rigor as a proper TE project would, does not mean they do not make useful advancements toward their mutual goals.

One of Robinson’s biggest beefs is over how DH projects do not appreciate how creating an online archive does not serve as scholarly edition. There is no question that the quality of these kinds of digital archives varies wildly. But is it really the case that these projects make claims for having effectively replaced editorial expertise. Rather, it seems like they aim to provide platforms that allow (potentially abused) freedom to go beyond highly mediated editions. They help provide a reader view a witness or text far more useful for a particular purpose than could a collated average that would normally be the well-regarded output of an editorial effort. But for many textual scholars like Robinson, such resources are too unmediated to be useful. Rather, we need official editions and arguments deliberately prepared by an editor because “without arguments, our archives are inert bags of words and images.”

The view that Robinson reflects here perhaps grows out of a similarly ubiquitous assumption that editorial work is not fundamentally different, even if its tools are and its end products should be more accessible. This is, perhaps, where disciplinary blinders loom large. One could imagine that the future of editorial work lies not only with carefully prepared editions, but also in properly guiding highly specific and situational reading with more accessible granularity than critical editions typically allow. In other words, crucial (digital) editorial work might also serve more as a facilitator of possible texts and interpretations. This kind of editor would be as much concerned with interfaces and metadata as much as the texts themselves. Perhaps editorial success could mean more fluidly mediating an editorial process (guided by sound editorial principles) for readers rather than imposing it. Maybe the success of a future textual scholar will be to make everyone else better textual scholars by helping to highlight texts that particular readers, rather than “objective” editors, need. Robinson would deny these efforts as a useful goal, for “the editor is the editor, and not a 'facilitator'.”

Surely, Robinson's TE position will sound absurd in the DH world; the DH alternative will sound equally absurd in the TE world. But why have such a binary between the roles of editor and facilitator? The conceptions of and uses of text are simply not what they once were in a scholarly world where the best texts could appear only in print. This does not obviate traditional roles and functions of editors, but it certainly creates new roles as well. How to define these roles effectively? Collaborative projects that bring together the new DH explorations of text and TE ideas about using technology effectively with sound editorial processes and products.

The concept of configurable texts is hardly foreign to Robinson. One paragraph must be quoted at length because it’s so important: “We [editors] should work towards a future where anyone who wants to be an editor has, fit and ready for their hands, the tools for themselves to make the editions they want. Included in this toolset should be the ability to invite others to contribute, and to allow others to invite themselves to contribute. In this future, parts of the edition, and the materials upon which it stands, may be endlessly and continually repurposed by others. For this to happen, we need far better systems for enabling the discovery and reuse of editorial materials.”

This is spot on. Except for the fact that Robinson seems to think that these kinds of tools should be the exclusive domain of the textual editors. An edition, created by a highly skilled editor, is what everyone else should consume. More problematically, the emphasis and implicit necessity of the edition erects a needless barrier toward open and large-scale text analysis—the kind of project that is like to result in not only better textual analysis, but also textual editions as well. We need access to both texts and editions, but this is a significant change from textual scholarship as it now exists as a discipline that has largely existed to create, through careful arguments and analysis, authoritative versions of texts. What is the challenge in creating configurable texts, whether by an editor or reader?

We know it's not technical. "In the last decades, we have solved almost all the technical problems associated with the making of editions." But this is true in the way that we have already solved all the technical problems with scholarly communication generally. The real problem is that the technical “solutions” have created many other challenges that are far more complex and unclear. We've have only begun to tackle most of these. For many innovative TE and DH projects alike, the problems have been, and will continue to be, fundamentally design problems. We are not even remotely close to solving these, in part because most textual scholars, and indeed most digital humanists, are not trained in information architecture and how to link that kind of infrastructure to graphic and interface design. Even when IT experts and designers are involved in a TE project, they are often not trained enough in textual criticism to really understand what the problems are and how to design effective and sustainable solutions for them, or how to suggest productive comprises between interfaces and existing practices. To be fair, these problems plague DH projects as well, and not many DH projects cannot be highly praised for their interfaces. But the complex, non-linear, and often interdisciplinary nature of many DH projects has forced more thinking about design and interface problems. I don't mean to suggest that DH projects have all the answers and TE projects don't. I mean that they now have better questions. Those questions, coupled with textual editing challenges will result in better textual tools and interfaces for them.

Beyond design, fundamental philosophies at the root of projects are perhaps the most important reason for collaboration. Case in point: Robinson makes an argument against crowd-sourced projects (or social editions), like [Transcribe Bentham](http://blogs.ucl.ac.uk/transcribe-bentham/) because they are financially inefficient. Some unspecified workers in the Philippines could be much more cost effective! He's right, but such criticism seems to have missed the point of early crowd-sourced projects. I'm reasonably confident that the smart people running these project know that they could find a cheaper way to get letters in digital form. The projects were never only about transcribing those letters. Certainly the point wasn't to do it as cheaply as possible, but to begin to figure out how to engage a small army of interested and capable enthusiasts and marshall their efforts toward some useful end. Again, not a technical challenge. Textual scholars may not approve of the final product, but that does not mean there is nothing to learn from these projects. Perhaps more importantly, apart from what these type of social editions have taught us about the mechanisms, software, and results, one could argue that Transcribe Bentham and projects that followed in its footsteps have done more to raise awareness across disciplines of the difficulty and importance of textual editing than most textual scholars themselves have. This is the kind of experimentation usually inherent in DH projects that reflects a philosophy of openness and experimentation and critical inquiry that can help future editors and readers have the appropriate tools they need to read the most appropriate kinds of texts or editions. Yes, they should make better editions, and why DH/TE collaboration remains essential.

It is safe to say that digital humanities projects are likely to remain text-centric, at least until we re-imagine our graduate training. To foster a turf war about who should be allowed to work with text in a public way only encourages more of the problems that Robinson sees with existing projects because the effort of DH projects to make texts, however sloppily they might be created, more available and accessible and reusable is not going away. On the whole, I believe it fair to say that textual scholars have not established much of a reputation for tool building that comes anywhere near the goals Robinson has articulated. DH tools may not fare much better in comparison, but the deliberate discipline boundary blurring and role re-defining that has become almost routine in DH projects has something substantial to offer TE projects that seems unlikely to happen otherwise.

A causal reading of Robinson's paper might mistake it for a rear-guard action to prevent real change in light of the "digital revolution." In my view, it's hard not to see a bit of "Let's take our toys and go home" to it. Yet the paper is in fact thinking fruitfully about the future and the importance of change. Unfortunately, it does so largely within unnecessarily narrow parameters about the editorial roles and responsibilities and disciplinary boundaries. There will always be a place for exquisite editorial work, for standard editions; there is no reason to be insecure about that or draw lines in the sand. But we certainly can go beyond those traditions, and we need to explore those possibilities. That is the nature of scholarship, whether DH or TE.

There are too many possible, even if largely unimagined, worlds where DH and TE projects merge seamlessly together and enhance and move scholarship and methodology forward. We already know that DH projects need more careful attention to methods and arguments of textual production; we know that TE projects need more careful attention to new media and new research methodologies. It’s only in uniting their very different disciplinary perspectives, values, and goals—getting the best of both worlds, one might say—that either group can be fully productive in achieving the goals that each share toward creating individual editions and better reuse of textual materials.