Last week I attended a meeting to discuss changes in tenure and promotion policies at UNC. To its credit, the UNC administration was early to recognize that these polices needed to change with the times, and in 2009 produced a report recommending that tenure committees work to recognize engagement with the public, new (especially digital) forms of scholarly work, and work across disciplinary lines. So far, so good, but the devil is in the details, and currently each department and school is being asked to revise its specific policies to bring them into alignment with the high-level recommendations of the report.
The meeting focused specifically on how humanities departments are addressing this task. Chairs of the various humanities departments, as well as tenured faculty known for their nontraditional work, gathered at the Institute for the Arts and Humanities to compare notes and hopefully stake out some common ground. As the only untenured faculty member in the room, with about 1/30th the professional experience of the average attendee, I mainly stayed in Wise Old Owl mode.
But I did speak up once. The discussion had turned to new forms of scholarship and how to assess their significance or influence. I pointed out that computer scientists seem to have solved this problem. They write software that demonstrates some new algorithm or technique, and often they will make this software available as open source software to be used by other researchers. (This is quite common in the natural language processing community, for example.) Those other researchers may know the work upon which they are building primarily through their use of the software that embodies it, rather than through reading papers. Arguably, then, the software is the primary scholarly product.
So, how do these scientists assess impact in this brave new world in which software code has replaced words on paper as the medium of scholarly communication? Well, usually they write a technical report, post it on the same website that their software can be downloaded from, and kindly request that it be cited by others who publish work for which the software was used. The technical report might get “officially” published in conference proceedings, or it might not: what matters is that it provides a surrogate for the “real” scholarly product, a surrogate that fits neatly into existing networks of dissemination, citation, and evaluation. My point was that we needn’t view the problem of how to evaluate non-traditional forms of scholarly work as one that “consumers” (such as tenure and promotion committees) must solve alone. We can also help “producers” find ways to make their products more easily consumable.
One scholar at the table suggested that this might be viewed as unfair: traditional scholars only have to write an article, while non-traditional scholars have to build something and write an article. Another thought I was being overly conservative: he pointed out that scholars had been writing articles and books to help their digital scholarship “pass” since the 1990s, and it was time that the academy allowed such work to stand on its own.
I didn’t pursue the point further, but this last idea–that digital work needed to “stand on its own”–stuck in my mind. It echoed assertions I’ve seen many times in the “hack vs. yack” DH permathread, that code “speaks for itself.” “And the more I think about it the more I am convinced that this is the wrong way to think about digital scholarship, because no scholarship, no creative and productive work, “stands on its own” or “speaks for itself.”
Two points. First, consider the open source software community. Some digital humanists–often the same ones arguing that digital work speaks for itself–point to the open source community up as model to emulated. Surely, if code speaks for itself anywhere, it must do so there. It’s a community that gets by on little more than “rough consensus and running code,” right? Wrong. Open source software development consists of an enormous amount of communication through mailing lists, IRC, bug tracker comment threads, blog posts, conference presentations, and face-to-face meetups, plus a little bit of coding. The ratio of yack to hack is enormous. Github is full of projects with impressive code that will never go anywhere, because the hackers behind them are unwilling or unable to effectively communicate about them.
Second, the suggestion that digital scholarship should have reached a level where it can “stand on its own” implies that such a level exists, and the traditional book or article are already on it. This is naïve. Books and articles have the stature that they do because they are integrated into and have evolved along with networks of scholarly practice, networks that extend far beyond just what individual scholars do. A key problem is that scholars tend to see scholarly products, whether books or films or collaborative websites, solely as expressive media. They are that, and they are something else as well: they are technologies.
If we view scholarly works simply as ideas fixed (or flowing) in some medium of expression, then our collective inability to shift easily from one medium to another as new expressive possibilities emerge can be puzzling. The puzzle is familiar to anyone who has read histories of technology. Looking retrospectively, it’s easy to identify “clearly superior” technologies that at the time of their invention failed to gain traction. We may then blame conservatism or short-sightedness for this failure, and congratulate our modern selves on not being so hidebound.
Historians of technology tend to take a different view. They usually distinguish between invention and innovation, with the former being just one step in the latter process. New inventions aren’t immediately recognized as useful; they don’t “speak for themselves.” Even something as manifestly desirable (to Southerners at least) as air conditioning wasn’t simply embraced by the sweaty masses. It had to be carefully marketed for its desirability to manifest. And innovation isn’t just about marketing; it’s about creating the conditions under which an invention can be viewed as useful or desirable. Apple didn’t invent MP3 players, but only they were able to create the conditions that compelled masses of people to buy them.
If we view new forms of scholarly work as technologies, it becomes clear that having them “stand alone” is precisely what we do not want to do. Inventions that stand alone wither and die. What we need is more focus on innovation and less on invention. And we ought to note that innovators often find success by taking advantage of existing infrastructure. In the specific case of scholarly communication, that infrastructure includes libraries, funding agencies, and even the villainous publishers. Taking advantage of it requires recognizing how finely adapted many of our traditional tools are to the networks in which they function, and finding ways to emulate that, even if we are accused of pouring new wine into old bottles.
And this is why I’m encouraged to see efforts like The Journal of Digital Humanities that are taking a traditional form (the academic journal) and using it in a new way. I recently received from the editors of the JoDH a request for permission to publish some tweets I contributed to a larger conversation on Twitter about theory and the digital humanities. Traditionalists might scoff at the idea of treating tweets as scholarly publications, while those who dream of burning down the old regime might lament it as a desperate attempt to prop up a dying medium.
I think it’s a clever way to recognize the rich variety of scholarly communication taking place in non-traditional venues, while simultaneously taking advantage of the affordances of a traditional one. It’s not inventive, but it’s innovative. It makes that Twitter conversation a little bit easier to read, a little bit easier to contextualize, a little bit easier for me to point to on my CV, and a little bit easier for a tenure and promotion committee to evaluate. My tweets certainly don’t stand on their own, and fortunately they won’t have to.