Towards a Front Page for the Digital Humanities #dhthis

In partial response to the issues raised about the publishing model of Journal of Digital Humanities and an apparent lack of transparency, several scholars banded together to create DHThis, an experiment in open sharing of DH-relevant writing online that follows the Slashdot/Reddit model that allows users to vote content up or down and for users to receive karma points. The idea behind such systems is one that trusts in openness and the wisdom of the crowd, while putting in place some rules to discourage gaming the system and to filter useful, interesting, and important content to the top. DHThis is, thus, both a challenge and a potential complement to Digital Humanities Now (DHNow) and The Journal of Digital Humanities (JDH) as curators of online resources (blog posts, tools, CfPs, etc.) relevant to the DH community. Both DHNow (which has already linked to DHThis) and JDH collect, curate, and rebroadcast existing content. In the case of JDH, authors are often invited to revise their work prior to its appearance in the journal.


The scholars responsible for launching DHThis—Adeline Koh, Martin Eve, Roopika Risam, Jesse Stommel, Alex Gil—are all people whose work I respect. They have made it repeatedly clear, too, that this is an experiment and invite anyone who is interested to provide feedback and to help develop this new site. They saw a problem in the existing systems for digital publication and determined to try and solve this problem by providing another space that functions by different rules for community engagement and sharing. I worry, however, that in their eagerness to make a platform available, that the DHThis development team skipped some crucial steps: both the content curation model itself and the CMS used to implement it should be objects of deeper critical consideration before implementation. As a result, Whitney Trettien calls DHThis, "a step backward, both in the pace of the discussion and in terms of identifying tangible solutions." Likewise, Natalia Cecire asks "what structural safeguards will prevent DHthis from becoming a terrifying cesspool of misogyny, racism, and assholery?" 


Rather than look for existing models in some particularly problematic places (Reddit has a terrible history of misogyny, group-think, and outright abuse; Slashdot is similar and is a site devoted entirely to technology news and one we might assume mirrors the IT worker population, with all the problems of that group) as a complement to the existing editor-centered models exemplified by DHNow, JDH, and others, I would rather the first step have been a conversation about what forms openness might take, how we can best encourage community engagement, and what assumptions underly the rules organizing each model. I strongly suspect such conversations happened among the developers (indeed, I would be surprised if they didn’t), but by moving directly to implementation before having a public and thoughtful discussion about the affordances, structures, and assumptions behind these different models, I fear that DHThis may unnecessarily re-create problems with existing models.

For example, the chosen Content Management System (CMS), Pligg, apparently imposes static categories on content. As I write this, the categories are still in flux, but include ones like “Coding/Making”, “Pedagogy”, “Employment”, “Methods”, and “R/C/G/D” (race, class, gender, disability). Other than the RCGD category, the categories reflect the economic and professional pressures around employment, publication, and practical methods/tools that, to many, signal that the digital humanities is far more invested in technology than in the critical inquiry that is a hallmark of the humanities. There’s also a haphazard mix of topic, form, and genre in the existing categories. In the open Google Document where these categories and other site issues are being hashed out, I suggested another category called “Critique”, but am unsatisfied with that label specifically and with the categorization system generally. What Pligg apparently demands for the organization of content is a controlled vocabulary, which I find inherently problematic. A controlled vocabulary for such a potentially diffuse range of content can either attempt to be comprehensive, and so devolve into a near infinite number of categories and sub-categories, or it can attempt to be a useful browsing tool, and so stick to broad generalizations. And although submissions can be marked for multiple categories, there will invariably be some content that does not fit neatly into any combination of overlapping boxes.

The DHThis developers, too, appear to recognize that the categorization scheme is problematic. But the CMS chosen enforces it. Another, more flexible option would to follow more fully the Reddit model by allowing for sub-reddits devoted to whatever topics people find necessary. This model would make it easier for new communities to coalesce, but has the problem that Reddit is a moderated site; each sub-reddit has its own set of moderators who set the rules and topics for “their” sub-reddits and who police the submissions and ensuing discussions for relevance and importance. Given the desire for transparency, collectivity, and openness, a moderator-centered model seems at odds with DHThis’s goals, and so not a proper fit. To this suggestion, the developers answered: “We considered it, but wanted something more open.” And I don't blame them.

Consider, then, the model that organizes content on HASTAC: users may form groups at will, then post their content within those groups (privately or publicly at their discretion). Users can also tag their posts with a much fuller set of categories and with free tags (folksonomies). The categories, however, do not prominently divide the content; they are just one more way to discover related posts. DHThis does allow for free tagging, as well, and I’m confident the categories will expand to be more comprehensive/messy; the site also allows for the creation of new groups, though the groups are not currently featured on the main page. Further, the prominence of categories in DHThis's navigation imply that they are meant as the primary mode of content organization and discovery. 

Because software is a form of writing/making with its own assumptions and affordances, we have a responsibility to critique the software, not just install it to meet our solutionist desires, which leads to another, broader concern I have about DHThis, which Morozov covers at length in To Save Everything, Click Here: The Folly of Technological Solutionism. It seems as if the DHThis team has fallen prey to the process of seeing a problem, wanting to do something about it, and then reaching for a technological solution. The chosen software platform should be evaluated for its limitations, not accepted because it does most of the submission-, voting-, and karma-handling out of the box. 


I realize that the site is an experiment or, to use a writing metaphor, a very rough draft. In response to a blog post I wrote about surveillance, Alan Liu recently encouraged us to think about the issues of “openness”, “public”, and “media” as necessary “intervening objects of inquiry”. And Josh Honn tweets of #dhthis:

Given that, in HASTAC, we already have a successful model for humanistic online community-building and content sharing, and that Slashdot/Reddit models of knowledge organization and dissemination have a history of re-inscribing existing power structures, then, if we do need a new site or a new tool (which is not altogether clear) to address these problems (which should also be clearly defined and discussed), then it seems to me that we should start with a more thorough examination of the models for sharing, discovering, and discovering content, of what it means to be “open” and “transparent”, what constitutes a “collective” or a “community”, of whether a solution lies in software or elsewhere, and, if a(nother) technological solution is necessary, what the best1 tools are.2

I applaud the DHThis team for their courage to publicly experiment, possibly fail, and invite collaboration and comment in the process. In that vein, I suggest we take the existing DHThis site3 not as a solution, but as another significant contribution to an ongoing conversation. If we accept it as such an offering, we also might consider the relationship between more traditional modes of discussion and the hacking ethos prevalent in DH; it suggests a dialectical approach to building in which we should expect multiple attempts at tools not to “solve a problem”, but instead to prompt further conversation through concrete examples that can make clear what we have yet fully to examine. As Alex Gil notes:

We rolled out fairly quickly as you can tell. Jesse and Martin worked hard and around the clock the platform ready, but we all knew that we were coming out in our pijamas. The reason for doing that was so we can jump right in into the part where we build this with the community (and by community we simply mean those who want to play here). No hiding. Platforms come and go depending on participation, and this one is no exception. If folks find it useful it will stick and serve a purpose. 

We're not ignoring the conversations that have happened around this kind of platform. We just figured we can talk while we build.

I’m of the camp that feels the more important an issue, the more we should yack before we hack, even as we recognize hacking as speech of another sort.4  So, I accept the generous DHThis invitation to make my own small contribution of what I would like to see, the questions I have, and what sort of platform I would like to work on and use.

I am fine with trusting an elected and rotating moderator board with clearly debated and drawn up agreements to do good work selecting content, managing comment trolls, and curating connections. To me, that is open and transparent, as well as a long-standing practice wherein these questions have been considered. (Though this raises the question: how is this different from DHNow?) And let’s think through the possibilities of the platform. What if we had random selections of submissions, various text analysis algorithms (parallax pace Drucker) for exploration, along with moderated comments to deal with trolls? Should everything be archived and stored for the record, with notices clearly posted to such an effect? What about a real name policy for all users, but one wherein names are only visible to other registered and confirmed users, and perhaps hidden by default, but easily discoverable? Or maybe only to the moderating board? I don’t know, but our modes of self identification and behavior policing (rules are only rules if they are enforced) must be examined and debated.

Maybe we should force SSL and encrypt storage to discourage casual government snooping. Or even store everything on offshore servers to protect all members from any governmental, institutional, or political silencing. Moderators may need access to all PMs, perhaps only after authorized by a vote, to police abusive behavior. At the same time, perhaps the traffic analytics should be public so that we can see the reach of the site and evaluate it fairly. 

Instead of ranking items based on a simple binary evaluation, why not weight based on some metric around newness, importance, commenting frequency, and quality? There’s no need to restrict ourselves to one (or even two) ways of exploring and sorting the submissions. Karma could be based on one’s level of engagement and network measurements. I would also like to see part of the curation of content is in linking things together as related in some way or another. A dynamic and interactive exploratory interface that allows us to move through the networks of content, in a versioned, wikified sort of way, sounds like fun. Indeed, we should attend to the initial conditions we set for the organic growth and mapping of communities. We should discuss what new ways of sorting, ranking, and evaluating content are possible.

As with all communities, divisions will arise and the potential for abuse will exist. Maybe we need a clear set of rules, which should include language to encourage full social justice and respectful conduct, language that is similarly debated and ratified. The nice things about building a platform derived from such discussions is that we can agree upon and neutrally enforce, through the code itself, many aspects of our policies.

If we don’t pause now and think through the media we use, we’re doomed to repeat the dramas and dissatisfactions that lead people to leave Twitter and to worry (appropriately) so much about silencing of all types, about inclusiveness balanced with respect, and about our evolving modes of communication, evaluation, and reputation. How do we (can we?) include people who don’t feel the need to spend so much time on Twitter or reading blog posts? We do have other work to get done, after all, and should also be concerned about the ways uncompensated labor tends, like a gas, to expand and fill all available space.

Because what we’re discussing and what DHThis offers is the development and expression of a coherent community or, at least, a loosely overlapping set of communities. And the critiques and worries stem from our experiences with existing communities, governments, and other forms of structure and power. We should be deliberative and slow when building new ones. 



1 Where “best” is a combination of technological capabilities, an ability to meet the demands of critical inquiry, and an evaluation of the existing work in this space.

2 A few other questions that occur to me: Is “popularity” of content an important facet for discovery? Does karma tied to user profiles imply mistrust? Is there room for serendipity? What does it mean to curate? I’m sure others can come up with much better questions. The point is, I’m more interested in the conversation we could be having than the current incarnation of DHThis.

3 We should get this site into a web archive, stat!   

4 When we frame hacking, coding, and the whole maker ethos as expressions of speech, we bring it clearly into the purview of the humanities. Certainly, such work is far closer to speech than corporate money, imho.  

Last modified Tue, 17 Jun, 2014 at 8:48