Proposal for Feminist Debates in Digital Humanities

Barbara Bordalejo

KU Leuven

Blijde-Inkomstraat 21

Room:03.35

Phone:+32 16 37 21 39

e-mail:

Minority Report: The Myth of Equality in the Digital Humanities:

I have serious reason to believe that the planet from which the little prince came is the asteroid known as B-612.

This asteroid has only once been seen through the telescope. That was by a Turkish astronomer, in 1909.

On making his discovery, the astronomer had presented it to the International Astronomical Congress, in a great demonstration. But he was in Turkish costume, and so nobody would believe what he said.

Grown-ups are like that . . .

Fortunately, however, for the reputation of Asteroid B-612, a Turkish dictator made a law that his subjects, under pain of death, should change to European costume. So in 1920 the astronomer gave his demonstration all over again, dressed with impressive style and elegance. And this time everybody accepted his report. (The Little Prince, Antoine de Saint-Exupéry)

1 Introduction

Digital humanities is a relatively new discipline and its practitioners are constantly pushing its boundaries and modifying its perspectives.[1]The general perception suggests that digital humanities, as a discipline, is balanced in terms of gender. Moreover, we have been told that “Digital Humanities is ‘Nice’” and we have been told why it is so nice (Scheinfeldt). However, these perceptions are easily contradicted. Even the simplest of analysis (such as the ratio of first author gender for accepted papers in the annual Digital Humanities conference) produces evidence undercutting the claims to equality and to the inclusiveness of the field.

The pervasive belief in gender balance in digital humanities prompts the question of how the field might appear if it could be analysed from across a broader set of perspectives. This article analyzes data gathered through a survey of self-described “digital humanists” to understand the current situation of the field. I conclude by making some suggestions which might improve the position of minorities within digital humanities.

2 Background

On May 23rd 2015, Lorna Hughes, then professor of Digital Humanities at University College London, tweeted a link to Gordon Hunt’s article “There is certainly no gender imbalance in digital humanities!”[2] The tweet sparked a range of opinions concerning Hunt’s article. Ben Brumfield, for example, pointed out that the situation in digital humanities was comparatively better than is the case in other fields, such as computer sciences (See Misa 2010 and the data by the US National Center for Education Statistics, wanted to see hard data showing whetherHunt’s article was correct.

My own experience of academia, in general, and digital humanities, in particular, suggests that indeed there is significant imbalance in the digital humanities. However, without data, my perception could (and probably should) be dismissed as just a personal impression.

Figure 1

Indeed, within a few minutes of the start of the twitter discussion, Scott Weingart presented the result of his analysis of the gender ratio of authors accepted for the Alliance of Digital Humanities Organizations conferences between 2010 and 2013, where women represent around 30% of authors. Against this, Hughes argued that she was one of six female digital humanities professors in London alone.[3]However, all these women were also white and native English speakers.

Weingart’s statistics are clear-cut.[4]Despite the prominence of some female scholars, women are underrepresented at the most important gatherings in the field. In an email correspondence with me, Weingart has said that he was interested in using digital humanities methods to research the field itself, a sort of meta-digital humanities research:

...essentially I've just wanted to turn DH methods on ourselves for a number of reasons. What's it like to be scrutinized the same way we scrutinize? What's our own history and contours? And (especially relevant in the context of this graph and these discussions) how can we become better scholarly citizens, and better global citizens?

We tend to inherit problems of representation and equality from academia & tech culture at large, but just because we're no worse than the rest of 'em doesn't mean we shouldn't try to be better. Recognizing problem areas is (I hope) a good start.[5]

The work on the digital humanities acceptances was carried out in collaboration with Nickoal Eichmann. They intend to continue this analysis in the future. Our private exchange prompted a self-reflective post entitled, What’s Counted Counts,[6] in which Weingart acknowledges that other aspects of diversity are much more elusive than gender:

The answer was immediately obvious:ease. I’d attempted to exploreracial and ethnic diversity as well, but it was simply more fraught, complicated, and less amenable to my methods than gender, so I started with gender and figured I’d work my way into the weeds from there.

I’ll cut to the chase. My well-intentioned attempts at battling inequality suffer their own sort ofbias: by focusing onmeasurementsof inequality, I bias that which is easily measured. It’s not that gender isn’t complex (seeMiriam Posner’s wonderful recent keynoteon these and related issues), but at least it’s a little easier to measure than race & ethnicity, when all you have available to you is what you can look up on the internet.

Weingart’s post suggested to me that I should try to bring to light those aspects which are not readily available for measurement. The obvious way to do this was through a survey. This made me hesitate: a survey requires people’s willingness to participate. It would have to be designed and to be tested. It would have to ask very personal questions, some of which might be potentially identifying. There were many instances in which it could fail and many reasons that might push it towards that failure.

3. Methodology

3.1 The survey

3.1.1 Design and questions

The survey was modeled on diversity surveys often found as part of the hiring or acceptance process of higher education institutions.[7] There were 28 biographical questions and a final one that allowed comments or suggestions.

The questions covered the following topics:[8]

  1. age,
  2. education,
  3. formal education in DH,
  4. country of birth,
  5. gender,
  6. care-giving situation,
  7. sexual orientation,
  8. ethnic heritage,
  9. cultural background,
  10. native language,
  11. work language,
  12. work institution,
  13. existence of a DH center at work,
  14. attachment to a DH center,
  15. salary source (i.e. soft money, centrally funded),
  16. length of employment,
  17. previous employment,
  18. details of previous employment,
  19. current title,
  20. permanence in current position,
  21. nature of the current position (tenure track or non-tenure track),
  22. country of employment,
  23. visible disability,
  24. invisible disability,
  25. regularity of submission to the DH conference or similar,
  26. number of rejections from the DH conference,
  27. number of acceptances to the DH conference,
  28. perception of self within DH,
  29. additional comments,

Questions 9, 10, 16, 18, and 29 requested a free text answer. Questions 2, 5, 7, 8, 12, 19, 25, and 26 allowed a free-text answer as an alternative to the provided answers. Questions 3, 11, 23, 24, and 28, required further details for one of the possible answers.

The pilot of the survey went live on December 10th 2015. It was tested for five days before the survey was made public.

3.1.2 Gathering of Data

The survey went public on December 15th, 2015. There were 442 respondents in total. Two responses were excluded because the respondents stated specifically that they do not work within digital humanities. Another two responses were excluded because the tone of the answers was inappropriate, suggesting that they were skewed.[9] These suppressions left a total of 438 respondents.

4. Survey Results

4.1 Gender

Of the 438 respondents, 225 (51.36%) identified themselves as female, 2 as Genderqueer/Androgynous (0.45%), 203 as male (46.36%), 1 as transgender (0.22%), 1 as MTF (0.22%), and 5 as Other (1.14%). One person skipped this question. Of those who marked other, 2 presented alternative gender definitions, while 3 (0.68%) stated that they could not see the point of this question or stated that “I find this question intrusive: what has gender got to do with ‘multiculturalism?’" One respondent skipped this question.

Figure 2

It seemed clear from the gathered answers that the vast majority of the respondents identified with traditional binary gender labels.

4.2 Rejections from the Digital Humanities Conference

From Weingart’s statistic analysis, we know that roughly 30% of the accepted presentations in the DH conference have a sole or a lead female author. From this, it follows that some 70% of the presentations are by men or with a man as the lead author. Although Weingart has had access to the data of rejected papers, for ethical reasons, he has not released statistics on rejection. From those respondents who identified themselves as male, 149 answered question 26 (If you submit to the DH conference, how many times have you been rejected?), of those, 94 never had a paper rejected from the conference, 32 had a paper rejected once, 12 twice and 11 more than three times.

Figure 3: Male respondents graph.

In contrast, of 141 female respondents, 104 never had a paper rejected, 19 had one rejection, 11 had two rejections, and seven had more than three rejections. These results seem strange in light of the statistics that come directly from the accepted papers, which clearly show that 70 percent of the papers have a male lead author. However, since the question did not specify the role of the author in the acceptance, it might be that the female respondents are counting those papers in which they are not the lead author. Alternatively, it is possible that regular rejections from the conference might alienate scholars who later decide not to present more submissions. In any case, the data obtained through the survey cannot meaningfully be directly compared with Weingart’s data on equal terms. I present here the results for reference purposes.

Figure 4. Female respondents graph.

Question 27 of the survey (How many times more than one paper which lists you as the author or one of the authors has been accepted for the same DH conference?) was answered by 168 male respondents, 72 (42.86%) never had more than one paper accepted for the conference, this means that 96 (51.14%) respondents had two or more papers accepted for the same conference. For this same question, there were 161 female respondents, of which 86 (53.42%) never had more than one paper accepted, while 75 (46.58%) had more than one paper accepted for the same conference.

4.3 General differences based on gender

In general, the differences when using only gender to filter the survey’s results are relatively small. For example, one might have expected a significant difference in question 6 (Are you the primary caregiver of a child,

a disabled or elderly person?), where 36,61% of the female respondents answered “yes,” in contrast with 28,86% of the male respondents. The difference between these is smaller (although still important) than the perceived role of women as the providers of the bulk of the care for younger and older people. For gender-neutral questions, such as question 3 (Is any of your degrees in Digital Humanities?), 10.67% of the female respondents answered yes, while 7.43% of the male respondents gave a positive answer.

Question 7, which dealt with sexual orientation, shows quite a difference between genders. Of the female respondents, 78.73% stated that they are heterosexual, 5.88% gay, 11.31% bisexual and 4.07% choose “Other” (for which the following were specified: asexual, pansexual, queer, heteroflexible and human). By contrast, 87.13% of the male respondents identified as heterosexual, 5.94% as gay, 4.46% as bisexual and 2.48 choose “Other” (these were defined as queer, bisexual tendencies and “do not wish to answer”).

4.4 Rank

Rank is one aspect in which we can see a concrete difference between male and female respondents:

Title / Male % / Female %
Chair or Named Professorship / 2.48 / 1.78
Professor / 8.42 / 5.78
Associate professor / 14.36 / 11.11
Assistant professor / 13.37 / 21.78
Adjunct professor / 0.99 / 2.67
Seasonal lecturer / 0.50 / 0.89
Teaching or Research Assistant / 7.92 / 4.89
Head Librarian / 0.50 / 1.33
Subject Librarian / 1.49 / 2.67
Specialist Librarian / 2.97 / 4.44
Technical Officer / 2.97 / 2.22
Programer / 3.47 / 0.44
Student / 8.42 / 8.44
Other / 32.18 / 31.56
Total / 100 / 100

Analysis of the replies under the category “Other” show that some respondents found it difficult to place themselves into categories that present a different nomenclature in their own country. When possible, the responses were assimilated into categories (See notes for more details). The corrected results can be seen in the following table.[10]

In this table, I have included all the data, including the answers by respondents that identified themselves as “genderqueer,” “transgender,” and “other.” I include this date for this table, as rank is a factor that pertains to an individual and, for this reason, it is important to be able to see all the data. Because I have only include data (conflated) for the non-binary gender option only in this section, this reduces the risk of potential identification in other parts of the survey.

Title / Male / F
emale / Genderqueer /transgender / Other / Total
Chair or Named Professorship / 7 (50%) / 5 (35.7%) / 1 (7.14%) / 1 (7.14%) / 14
Professor[11] / 20 (60%) / 13 (39%) / 33
Associate professor[12] / 34 (55.73%) / 26 (42.62%) / 1 (1.63%) / 61
Assistant professor / 30 (37.5%) / 50 (62.5%) / 80
Adjunct professor / 3 (23.07%) / 8 (61.53) / 2 (15.38%) / 13
Seasonal lecturer / 1 (25%) / 3 (75%) / 4
Teaching/ ResearchAssistant / 16 (57.1 %) / 11 (39%) / 1 (3.57%) / 28
Head Librarian / 1 (25%) / 3 (75%) / 4
Subject Librarian / 3 (33.3%) / 6 (66.6%) / 9
Specialist Librarian / 9 (33.3%) / 17 (66.6%) / 26
Technical Officer / 7 (58.33%) / 5 (41.66%) / 12
Programer / 10 (83.33%) / 2 (16.66%) / 12
Student / 20 (44.44%) / 25 (55.55%) / 45
Post-doctoral researcher[13] / 7 (50%) / 6 (42.85%) / 1 (7.14%) / 14
Director / 2 (22.22%) / 7 (77.77%) / 9
Associate director / 2 (66.6%) / 1 (33.3%) / 3
Project manager / 6 (54.54%) / 5 (45.45%) / 11
Coordinator / 1 (14.28%) / 6 (85.71%) / 7
Researcher / 6 (42.85%) / 8 (57.14%) / 14
Other / 18 / 18[14] / 1 / 36
203 / 225 / 4 / 5 / 436[15]

This table shows that the first three categories (Chair, Professor and Associate Professor) are occupied by men in higher numbers. This is not particularly surprising given the general bias in academia towards male scholars and the way in which women and minorities tend to experience university careers differently from male colleagues (Johnsrud & Des Jarlais).

Fourteen respondents have a chair or named professorship. Five of those (35.71%) are female, one is genderqueer/transgender (7.14%), one chose “Other” (7.14%) and the rest (50%) are male. Nore is it surprising that all of the named professorships, including the ones that were outside the global north and Australia, are held by respondents who described themselves as caucasian in question 8. When asked to specify their native language twelve of the fourteen listed English as their first language. For more details about ethnic diversity see section 4.5 below.

From a total of 33 full professors, only 13 (39%) are female. Of the associate professors, 26 (41%) are female. But at the level of assistant professor, we see a sharp reversal in the numbers, where 62.5% are female. This could be explained in different ways. Perhaps the discipline is already changing: there are so many more women at the beginning of the tenure-track system that a more even distribution of gender ratio at higher levels in the future appears possible. This would be good news. It is conceivable that change might be coming to the field and that the superior number of female assistant professors will eventually translate into a higher number of female associate, full professors, and chairs. However, it is also entirely possible that these women have not yet been displaced by the system; that, being at the beginning of their careers, they might end up taking much longer to move from assistant to associate professor, or from associate to full professor (Towsend 2013), or might not manage to move at all.

Categories such as “adjunct professor” or “seasonal lecturer” received too few responses to produce significant results, although both of them, as “director” or “coordinator,” present a much higher percentage of female respondents.

The categories of jobs within libraries (head, subject, and specialist librarian) were selected by 39 respondents. Women hold 66.66% of the library jobs, as opposed to men. Similar to the “named chairs” and “professors,” the head librarians are caucasian and all of them use their native language at work. Of the subject librarians, seven were caucasian while two were mixed. Only one of them uses a language different from his/her native language for work.

Again, there is a higher number of respondents (26) who are specialist librarians, of those 24 (92%) are caucasian. Only three (11%) of them do not use their native language at work.

4.5 Ethnic Background and Other Factors[16]

Of the 438 respondents, 342, that is 79.35%, described themselves as caucasian. There is no substantial difference between male (80.5%) and female (79.64%) respondents and, although one could have expected a high percentage of caucasians in the field, the preponderance of caucasians might be related to the way in which the survey was distributed (through social media, like Twitter and Facebook, as well as via distribution lists, such as Humanist, RedHD, and GO::DH). But it also reflects a homogeneity that characterizes the different digital humanities gatherings, particularly the Digital Humanities conference (sponsored by the Alliance of Digital Humanities Organizations).

When this is viewed in combination with other factors we see that of the 33 full professors, 29 are caucasian and 4 (12%) described themselves as mixed. One person refused to respond, since “ethnicity is a trivial accident of birth and not connected to the truly great heritage I enjoy from the other earlier humans who were not my direct ancestors.” This type of comment suggests a lack of awareness of the experience of minorities: ethnicity is only trivial for those who can pass as part of the majority group in which they move, that is, for those who are not part of a visible minority. Individuals of ethnic minorities know very well that is nearly impossible to disconnect oneself from one’s ethnicity (Essed 1991, 1994) class, or gender (Crenshaw 1991).

Among the full professors, of the 13 female respondents, two are of mixed race. Of the 21 male respondents, two are of mixed race, lowering the percentage of male, mixed race full professors. Within the same group, one respondent is bisexual, and two are gay or lesbian. Again, at this level, the dominance of heterosexual caucasians is remarkable. Although a variety of languages were represented in this group, English was the dominant one (16 native speakers), followed by Spanish (5 speakers). For scholars at the professorial level, only 20% work in environments in which their native language is not the main working language.