Monday 27 May 2013

On Changing the Rules of Digital Humanities from the Inside

There has been a lot of talk recently about how my field – Digital Humanities – has to change. We are too insular. We’re excluding those who want to partake in it. The structures that have been built within the discipline preclude the type and means of research which we claim to do.  Issues of gender, race, ethnicity, and class raise their heads. There are a few online resources that exist which sum up these feelings: see “Toward an Open DigitalHumanities” google discussion document and, more recently, the Open Thread on “The Digital Humanities as a Historical“Refuge” from Race/Class/Gender/Sexuality/Disability?” over at Postcolonial Digital Humanities.

I’m not denying that there are issues in Digital Humanities. One need only look at the recently published program for DH2013 and cast your eye over the authorship of the accepted papers to see that this year’s Digital Humanities presenting cohort is around 65% male, 35% female. But what I would say, speaking on a personal level and not representing any authority here, is an obvious point which I don’t hear often voiced. Most people “within” Digital Humanities – that is those within the ADHO committee structures, those helping to run the conferences, those helping to allocate student bursaries and prizes, those helping to review papers and manuscripts, and heck, even the cool kids on twitter, are people who want Digital Humanities to be as open and as great as possible. This whole field has been built on the hard work of many academics who have given up their free time to try and entrench the use of computing in humanistic study into an academic field of enquiry, and it wouldn't exist without them, even if the form it exists in is currently imperfect. I would say, from where I sit on various committees, that people want to keep DH growing, and growing healthily. So if there are things wrong with DH, then do give concrete examples, or propose concrete solutions, so they can be taken forward. They're listening - we're listening.

There are things that have really frustrated me within DH, and it is only recently that I’ve started to actively question and pursue them, to get them to be changed. For example, in 2006 I first noticed that the TEI guidelines encouraged the use of ISO5218:2004 to assign sexuality of persons in a document (with attributes being given as 1 for male, 2 for female, 9 for non-applicable, and 0 for unknown). I find this an outmoded and problematic representation of sexuality, which in particular formally assigns women to be secondary to men, and so, in one of the core guidelines in Digital Humanities, we allow and indeed encourage sexist structures to be encoded. I was shocked to hear this – and have often brought it up when discussing entrenched issues in DH about gender balance. In a recent conversation on twitter about this topic, Stephen Ramsay summed up the issue:

James Cummings responded to our tweets, asking why, if it bothered me (and others) so much, hadn’t anyone submitted a feature request to TEI about it? And you know, it had never occurred to me that there would be an easy route to question this sort of stuff. He pointed me to where to submit a request, which I did here.  The discussion which follows is really very interesting – look out for the “you cant possibly be offended!” argument, or the “but we’ve always done it this way!” response. Also look out for very vocal support from Gabriel Bodard, in particular, who helped steer the discussion forward to ensure that at
“the TEI Council meeting in Brown, 2013-04, we agreed to change the datatype of person/@sex, personGrp/@sex and sex/@value from ISO 5218 to data.word, so as to allow the use of locally defined values or alternative published standards to be used in these attributes.” 
Women are secondary in the TEI rules no more! Hurrah! – and all it needed for that to happen was for someone to raise the issue in the correct forum, and explain the issue to those who did not understand it, until they finally did.

I’m Program Chair for DH2014 and issues of diversity and equality are currently on my mind as we discuss and choose plenary speakers for the Lausanne conference. It was recently pointed out to me, though, that the ADHO conference protocols don’t allow issues of diversity to be taken into consideration when choosing plenary speakers, originally saying
“Keynote speakers are decided by the International Program Committee in consultation with the Local Organiser, and should ideally represent a range of disciplines, interests, and geography.”
This isn’t good enough, as it means that you cant say “We’ve got a man to be one of the speakers, how about having a woman for the other one?” without being at risk of being accused of breaching protocol. I’ve recently chased an amendment round the ADHO committee structures, which means the ADHO conference protocols, since last week, state:
"Keynote speakers are decided by the International Program Committee in consultation with the Local Organiser, and should ideally represent a range of complementary disciplines, interests, and geography, with consideration given to issues of gender equality, and economic, ethnic, cultural, and linguistic diversity." 
Perhaps a small deal, focussing on the choice of a couple of speakers a year at our international conference, but pointing to the fact that the ADHO constitution needs to be looked over, to see where we can enshrine issues of gender equality, and other issues of diversity, within our communities. We need to make the rules that people have to abide by. We can make the rules, and we can change the rules. What rules would there help to be?

Of course, changing rules and guidelines wont make everything change overnight, and I wouldnt like to naively claim they will solve everything, but they are a start. I guess what I’m saying here is that, in general, folks “within” Digital Humanities are doing their best, and open to discussion and improvement, and are not willfully obstructive to those of a different gender, race, or economic class, etc. Criticism is helpful, and if there are things that need changing, or unconscious biases that need rectifying, then point them out, tell us. Tell us where concrete things are that we can act upon. We all want Digital Humanities to be the best it possibly can be, and I, for one, don’t mind changing the rules from the inside, in the time that I remain there. 

28/05/13 Addendum to the original post:  for an ADHO led initiative on diversity see GO::DH. I'd also like to encourage anyone who is interested in discussing change to consider standing for election to one of the ADHO organisations - we always need volunteers who want to roll up their sleeves!


James-C said...

Just to note that the reason the TEI initially adopted this standard is because, where they exist and seem appropriate, the TEI attempts to use international standards. In creation of the TEI scheme we do try to avoid the 'not invented here' syndrome. Thus, although we have ways to do graphics and some basic coordinate-based zones we would still suggest the use of SVG for truly scalable vector graphics, or MathML for mathematics. There is often a 'simple' way to do things in the TEI for basic/straightforward cases and more complicated ways would involve using another appropriate standard. The question raised for the TEI Technical Council was that this ISO standard may not be 'appropriate' any more.

In this case the single-digital representation of sex I think the ISO standard originates from trying to get countries to agree where some already used odd numbers for men and even for women, and other countries where they already used 1 and 2. I've always explained this a bit tongue-in-cheek when looking at this attribute and commented that it could mean that women are twice as good as men -- counting as ascending rather than descending. I always recognised it was a bit problematic, but never thought it was truly offensive. This, however, is not the assumption made by most women I've asked in a very informal straw poll when it came up. (Though the majority, to be honest, weren't really bothered, those that were made the more negative assumption.) That said, because there was strong enough feeling that it could or would be misinterpreted as the TEI recommending a practice that would upset or alienate part of the community then that is more than a good enough reason to change. While we try to use international standards when they exist and are appropriate, here it was determined on re-examination that it really isn't appropriate. Most of the debate was really about _how_ to implement it so that we didn't sacrifice backwards compatibility.

The loosening of datatype enables those who do still wish to normalise people's sex as single digits to allow easy analysis across corpora to continue to do so, but loosens it up so that other values (such as 'male' or 'female' or any of the much more complicated values) are able to be used. This does sacrifice a degree of normalisation and perhaps a non-offensive and inclusive ISO standard will one day exist so that we can re-assess whether to suggest it as a recommendation.

On your larger point about the gender balance in DH more generally, I know that I've always tried to keep a good gender balance on speakers for the Digital.Humanities@Oxford Summer School ( but sometimes one does have to take advantage of what is available in the institution (or submitted to the conference). In this year's DHOxSS bursary competition it turns out that 70% (I think, guessing by name) of recipients happen to be female. (The applicants were judged solely based on their applications and the stated criteria.) In the end one should just judge based on quality of conference abstract or bursary application, general work, etc. and not let gender come into it at all. (Although I'm aware of my privileges as a white male, I do try not to let them influence my evaluations in such situations.) Generally when I teach TEI workshops I notice that there usually are more women than men. I could come up with all sorts of spurious reasons as to why this might be, but I'm not sure that would be helpful. Likewise, I could look at the last couple years and (guessing by name again) try to get some rough statistics as to gender balance in students at the summer school, but that also isn't really indicative since we don't control who attends, all genders of student are welcome!

If there are other parts of the TEI that you or your readers would like to change, please do so by submitting a TEI Bug Report or Feature Request on the TEI SourceForge Site (

Melissa said...

Thanks James - very interesting! and re gender and diversity... I think you are right that we have to be careful that it isnt "our fault" who applies to open calls for things, all we can do is try to be balanced ourselves when we have a choice (hence the plenary speaker thing). Cheers.

Martin Grandjean said...

How to "change the rules" of a field whose structure is not yet clear? (this is a real issue, even if it looks like a trolling). There are many scholars pursuing the digital humanities without even knowing it ... difficult to draw an overall vision.

j4 said...

Great post! Just a small quibble, since we're trying to be clear about terminology here:

"the use of ISO5218:2004 to assign sexuality of persons in a document"

This standard relates to sex, not sexuality. As far as I know there's no generally recognised standard for descriptions of human sexuality (closest thing I can think of is the Kinsey scale, which is problematic to say the least).

Sex also shouldn't be confused with gender, but that's a whole different can of worms... :-}

Craig Bellamy said...

Thanks for bringing this up and posting this Melissa. And the TEI issues is your strongest point and thanks for pursuing it (admittedly I had a bit of a chuckle and glad it could be amended quickly), but all the 'post-colonial' stuff on that facile 'GO DH list you mention is pretty much academic mashed-potatoes (1990s). Trying to change the world through DH is sot of like trying to change the world through Architecture. Or worse still trying to change the world through cooking. I would really like to stop mashed potatoes from being served at American and Australian and UK Universities to help address the obesity epidemic, but I would be accused of being out of touch with, well, what every a mashed-potato eater would throw at me (possible mashed potatoes).

I like the DH because of its scholarly politics and not because of general in-the-wild politics, which of course has its important place, but has basically swamped vulnerable parts of the of the humanities with mashed potatoes and turned them into servants, at the cost of something more challenging and imaginative (like perspective, power, wisdom, honesty and scholarly significance).

I think I am highly political, but I do this elsewhere (through both the informal and formal political system) and I like the DH (or how I understand parts of it) because of the quality and the self-esteem in the digital, scholarly record that I see and the way it extends and challenges me through significant cultural interpretations of this, not because it looks like the mashed potatoes that I see on TV (and they do look good there). There have been a lot of careers in the humanities built on mashed potatoes since it became the staple diet in the 1960s, but not many on humanities computing. This is where the real broader issues of 'diversity' lay.

It is not that we are elite, it is that we are not elite enough. Is is not that we are exclusive, it is that we are not exclusive enough. And in politics this is a bad thing to say, in scholarship it is a good thing to say. They are 2 different institutions with two very different understandings of merit and cultural contribution. Everything doesn't have to collapse into one; like in a mashed potato empire that is culturally flat but economically pyramidal . One potato , two potato, three potato, four...that's 4 dollars thank you!

There are a lot of demands put on computing professionals everywhere and there are real issues of labor relations in the DH and it is at times, difficult to locate academic merit and the career pathways that may constitute this (and do we really have control over this?). A lot of careers have been built in the humanities since the 1960s in terms of politics-in-the-wild, but hardly any careers in humanities computing. There is no digital humanities, only ways to see the digital humanities as the digital humanities . And I am seeing a big tent with a lot of political mashed potatoes in it at the moment and but much academic computing going on...

Craig Bellamy said...

I blogged a list of favourite projects about 'class, gender, race'. They are more more about content than code, but you get the drift of the argument.

All the best,


Anonymous said...

Apologies if this is sent twice - I received the error 'The characters you entered don't match the word verification. Please try again.'

Also, I notice that comments need to be approved. This is surely at odds with the open, CC-happy, free discussion spirit, etc., of DH, and conversations of equality.

"...with attributes being given as 1 for male, 2 for female, 9 for non-applicable, and 0 for unknown). I find this an outmoded and problematic representation of sexuality, which in particular formally assigns women to be secondary to men"

You do realize you're doing more harm than good to your own movement by espousing this mentality. It re-enforces the belief that feminism has been polluted by damp squibs, non-events, non-important elements of matters, etc., ridiculously argued to be of resonate importance.

By your own logic, if you're an unknown gender -- e.g. trans, gender-queer, merely don't wish to disclose your gender, and so on -- you're systematically encoded as being more important than both men and women.

That said, 'non-applicables' are damned to a serious hell. They're lambasted as wretchedly beyond the pale -- a gulf between them; and men, women, and unknowns.

This is silly. Equality is not won by constructing manners -- and creating rhetoric therefrom -- through which women are supposedly abused in systemically, encoded, and often unseen, ways.

Blogs like this, and published essays -- say, such as 'What Do Girls Dig?' by Bethany Nowviskie -- do nothing by harm to equality-interested DH scholarship. They do this by their own weakness; their lack of rigor, reliance on polemics; and aggrandizing of unimportant aspects of language, code, and statistics.

They wouldn't be accepted in others areas; and only are in this one, as it is so *academically* taboo to criticize equality-interested scholarship. The ivory tower is very forgiving of it.