“Middle-aged White Guys”: Explanations for Wikipedia’s Diversity Problems

„Welcome to Wikipedia, the free encyclopedia that anyone can edit.“ This inviting welcome message is placed right on top of the English Wikipedia’s main page. Similarly, the vision of the Wikimedia Foundation, the formal non-profit organization behind Wikipedia, reads as follows: „Imagine a World in which every single human being can freely share in the sum of all knowledge.“

Both these lines represent the utopia of digital inclusiveness. ‘Anyone’ should have the possibility to contribute to Wikipedia’s quest for collecting the world’s knowledge. ‘Every single human being’ shall take part in this digital knowledge exchange. In Wikipedia’s early years, critics questioned whether this radical openness allowed for a high-quality encyclopedia to emerge. The main concern was how quality and neutrality of the Wikipedia could be preserved when anyone can change, delete or amend anything at any time (in 2005, for example, the Guardian asked “Can you trust Wikipedia?“).

Responding to these questions, Jim Giles compared in a Nature article (2005) Wikipedia and the renowned Encyclopedia Britannica and found a similar number of errors in both encyclopedias; more recent studies confirm these results with different methodologies (see, for example, Rodrigues 2013). Furthermore, Wikipedia’s quality management became much more sophisticated over the years, for example by introducing “sighted versions” checked by experienced Wikipedians. And even though there are still regularly reports on manipulated or wrong articles in Wikipedia, the end of print encyclopedias nevertheless made it the undisputed winner in the battle of encyclopedias. Today it is hardly possible to make an online search without finding a Wikipedia reference prominently placed in the results list. Wikipedia has effectively become the central directory of world’s knowledge.

After its triumph, the criticism brought forward against Wikipedia changed fundamentally. Pundits such as Noam Cohen in a NYT article see not too much but rather lack of openness and diversity as Wikipedia’s main problem. While anyone can still edit Wikipedia, only a tiny minority is actually doing it. The situation is described on the Wikipedia page on „Systemic Bias“:

„The average Wikipedian on the English Wikipedia is (1) a male, (2) technically inclined, (3) formally educated, (4) an English speaker (native or non-native), (5) aged 15–49, (6) from a majority-Christian country, (7) from a developed nation, (8) from the Northern Hemisphere, and (9) likely employed as a white-collar worker or enrolled as a student rather than being employed as a laborer.“

In the words of Sarah Stierch, who has worked as a Wikipedian-in-residence at the Smithsonian Archives on diversity in Wikipedia: „[I]t’s being written by middle-aged white guys.“ The share of female editors, for example, is dramatically low even compared to other online communities. More women than men use online social networks and also in tech-oriented platforms such as Google+ or Reddit there are 27 and 16 percent female members respectively. The latest Wikipedia editor survey, however, counted only 9 percent women among active Wikipedians – even less than the 13 percent measured in a different study before (see, however, a recent paper by Benjamin Mako Hill and Aaron Shaw, which claims that the opt-in survey tends to undercount women; the corrected numbers are not much higher, though).

What is more, the overall number of active Wikipedians is stagnating for over five years in a row now. Anyone could participate but ever fewer do so. Given that Internet penetration on a worldwide scale is still increasing and that there are still huge gaps in potentially large language versions this stagnation in editor growth is indeed puzzling. Moreover, this stagnation makes it highly unlikely that the current bias in Wikipedia knowledge and editor base will soon be overcome by a stream of new editors. Wikipedia knowledge seems to stay white, western and male.

Figure 1: Number of active Wikipedia editors over time (active = more than 5 edits/month), Source: the Atlantic, http://bit.ly/OJvS3p

At odds with its stagnating community of editors, Wikipedia’s importance as a knowledge ressource is still on the rise. Corporations and politicians alike are often either dissatisfied because there is no Wikipedia article dealing with them or try to sugarcoat extant articles (see, for example, the case of the German Daimler AG). Historically, production of legitimate knowledge has always been a contentious struggle. However, the technological potential of the Internet to make these struggles not only more transparent but also more inclusive has not been realized even by Wikipedia. Technological and formal openness is seemingly not enough for large strata of the society to take part in the endeavor of collective knowledge production. Paradoxically, this makes the inclusive utopia of an open online encyclopedia an interesting case for researching exclusionary dynamics. Drawing upon my own research and a review of extant works on Wikipedia, I thus put together the following, non-exhaustive list of explanations for exclusionary dynamics within Wikipedia:

• Mirror thesis: To a certain degree, lack of diversity among Wikipedia editors mirrors overall exclusionary dynamics in the society at large. The power to legitimately produce knowledge is unequally distributed and depends on social status as well as on habitus. What the mirror thesis cannot explain though is why the share of female contributors in Wikipedia is even lower than in other online communities and forums.
• Hacker culture thesis: The Wikipedia ideology of free knowledge as well as the use of an open content license has its roots in the free and open source software movement, which is associated with a ‘manly’ hacker culture. And female Wikipedians extensively criticize the rough tone and the overall ‘manly’ atmosphere when asked in interviews and surveys (see, for example, a blogpost by Wikimedia CEO Sue Gardner). However, this explanation poses the question why the share of female programmers in most open source software projects is not only lower than the share of female Wikipedians but even lower than in proprietary software projects (see the slides of a talk by Hanna Wallach on the issue of women and free/open source software).
• Troll thesis: One consequence of radical openness is that, as Wikipedia researcher Joseph Reagle put it in a NYT interview, Wikipedia is also „open to very difficult, high-conflict people, even misogynists“. Even such a minority of trolls can contaminate the discursive climate and drive away particularly female editors, for example by name calling (see Figure 2). To prevent new editors from unfriendly or even insulting encounters, Wikipedia has already launched the „Teahouse“, which shall offer a friendly and welcoming environment for potential future editors.

Figure 2: Names that respondents in the study „Women and Wikimedia 2011″ say they have been called; word size represents frequency (vgl. http://bit.ly/14LdgXL)

• Path dependence thesis: Related to the hacker culture and the troll thesis is the path dependence thesis, which suggests that an initially low share of female editors leads to fewer women joining Wikipedia and vice-versa. As soon as the community is dominated by man and a ‘manly’ tone, it becomes increasingly difficult to change this status quo because it drives away specifically those people that would be able to initiate or support such a change. Early exclusivity may thus be self-reinforcing.
• Leisure thesis: Since women still do most of the unpaid work of reproduction, they have less leisure time than men for contributing to voluntary and unpaid projects such as Wikipedia. Similarly, people in poorer countries have less free time to invest in volunteer work, which could explain lack of diversity in other dimensions such as ethnic background.
• Gender difference thesis: Low participation of women in Wikipedia may be explained with gender differences that root in early socialization processes and also result in different preferences in terms of studying and profession. According to this thesis, the low share of female editors results from the binary conception of gender institutionalized in our society. The problem with this thesis is that Wikipedia is mostly about reading and writing texts, which is not a specifically ‘male’ activity.
• Usability thesis: Before the English Wikipedia recently launched its new „Visual Editor“, contributing required at least a rudimentary understanding of Wikipedias wiki markup language. In addition, the overall usability and design of Wikipedia has not been modernized substantially over the past decade. Taken together, wiki markup and ‘old school’ design emphasize the technological aspects of Wikipedia, which may – together with the gender difference thesis – account for some of the gender bias. It will be most interesting to see whether and how the roll-out of the new visual editor, which allows editing without knowledge of wiki markup language, impacts editor diversity in Wikipedia.
• Bot thesis: Technically versed Wikipedians increasingly delegate recurrent simple and comparably boring tasks such as categorizing or spell checking to automated scripts called “bots” (see also a recent paper by Claudia Müller-Birn, Jim Herbsleb and myself on bots and the rise of algorithmic governance in Wikipedia). The bot thesis argues that bots take over specifically those tasks that previously provided an opportunity for unexperienced Wikipedians to contribute. This way the entry threshold rises and less new editors find their way into the online encyclopedia.

Figure 3: Different explanations for exclusion in Wikipedia (Source: own composition)

In Figure 3 I have plotted the different theses for exclusion in the context of Wikipedia alongside the two axis specificity (how Wikipedia-specific is the explanation?) and technicity (can the explanation be addressed technologically?). In case that each thesis explains at least part of Wikipedia’s diversity problem, this depiction illustrates two things. First, there are more socio-cultural than technological mechanisms of exclusion. The new visual editor may improve the situation but is very likely no sufficient response to the different socio-cultural mechanisms of exclusion. Second, more than half of the explanations are more or less specific to the Wikipedia context. Therefore the community of Wikipedians should not simply point to societal dynamics of exclusion but rather try to actively change its structures and norms.

Increasing diversity in Wikipedia requires not only better technological accessibility but also a transformed online culture. Exclusionary dynamics of online culture have regularly been discussed outside of Wikipedia, as well (see, for example, “The Trouble with Blog Comments“). Maybe it helps to acknowledge that Wikipedia already draws sharp socio-cultural borders, thereby systematically excluding women, people of color and others. Instead of simply propagating openness, a more diverse Wikipedia might thus need deliberate attempts to re-draw this borders to arrive at a culture of reciprocal respect and appreciation.

This article is a shortened and slightly altered version of the German blog post “Wikipedia: Grenzenlose Exklusion?“.

(leonhard)