User:Yaron K./Editorial

Why Wikimedia should be gender-neutral
In October 2010, the Wikimedia Foundation declared a set of goals for 2015 that included increasing the percentage of Wikipedia's editors who are female to 25%, up from its current value of around 13%. I believe that this policy is mistaken, and should be ended.

Before I get to that, let me state that a topic like this can easily lead to overheated argument, since it deals with inequality - and thus is part of a much larger political debate that touches all of the most inflammatory issues of gender, race and class, in the United States (my home country) and elsewhere. I certainly don't want to get into any of those here - but thankfully I think I don't have to, because Sue Gardner, the executive director of the Wikimedia Foundation, has clarified that the reasoning for this policy is strictly a pragmatic one. In a January 2011 article in the New York Times, which as far as I know is still the definitive account of the policy, she notes that "gender is a huge hot-button issue", but explains the thinking in this way: "The difference between Wikipedia and other editorially created products is that Wikipedians are not professionals, they are only asked to bring what they know. Everyone brings their crumb of information to the table. If they are not at the table, we don't benefit from their crumb."

Now, with all due respect to Ms. Gardner, I would say that this is a rather inaccurate metaphor for the editing on Wikipedia, since a substantial portion of content on Wikipedia is contributed by a relatively small number of editors, who are doing much more than contributing a "crumb" from their own personal set of knowledge - they're doing enormous amounts of research (online and offline), importing outside data via bots, and of course doing copy-editing and the like; this kind of work is not professional only in that it's unpaid. But let's leave that aside - even if the metaphor of Wikipedia as a sort of amateur communal potluck is unconvincing, the larger point can still stand: people tend to contribute about subject matters that they're interested in, and because men make up the dominant majority of editors, the subject matter naturally tends toward the subjects that men are interested in. This argument is easy to understand and non-political: it simply states that it's in Wikipedia's own self-interest to have a greater percentage of women contributing.

It's a reasonable argument to make, but I believe it's incorrect. It makes two important assumptions: that a dearth of writing on women's subjects is a major problem on Wikipedia, and that increasing the percentage of female editors is the most effective way to address this problem - and I believe both are inaccurate.

First, are the smaller amounts of content on women's topics a major problem? The New York Times article cites some examples of the issue:
 * 1) The article on female British writer Pat Barker is substantially shorter than the article on Niko Bellic, the main character in the video game Grand Theft Auto.
 * 2) There are many more articles about characters in the TV show The Simpsons (45) than about Mexican feminist writers (4 when the NYT article was written, though now it's up to 9).
 * 3) Episode reviews for The Sopranos are much more detailed than those for Sex and the City - even though they're both popular, long-running HBO series.
 * 4) Articles about famous contemporary shoe designers Manolo Blahnik and Jimmy Choo are quite short.

But as anyone who has read a significant number of Wikipedia articles could tell you, there are other skews of content - and I think these are a bigger deal:
 * 1) Content about fictional worlds (from film, television, video games, etc.) is more detailed than content about the real world
 * 2) Content about the recent past is more detailed than content about the distant past - and the farther back you go, the sparser it gets
 * 3) On at least the English Wikipedia, content about the non-English-speaking world, and especially about the non-Western world, is less detailed than content about events and people in the English-speaking world.

These skews are so omnipresent that I'm not even sure if examples are needed. I'll just give one to illustrate the second of the three: the 1933 American novel Anthony Adverse is largely forgotten today, but this web page describes it as "one of the biggest selling novels of the 1930s". The English-language Wikipedia, at the time of this writing, has no article on the novel - though there are medium-short articles on its 1936 film adaptation and on on its author, Hervey Allen. Suffice it to say that, if a novel had been written five years ago with the impact of Anthony Adverse, it would have at least one extensive article of its own.

By the way, note that I'm using the word "skews", and not "bias", here. "Bias", especially as in "systemic bias", is a common word used to describe the situation, but I believe it's only partially correct, and needlessly provocative. The third of those three skews - having to do with geographic location - definitely can be attributed in part to where Wikipedia editors live. But the first two, and arguably the third as well, have more to do with simply the nature of information. Fictional worlds are very different from the real world, in that literally anything that happens in them - anything that's mentioned, anyway - can be unambiguously referenced. That's a marked contrast to the real world, where information about what actually happened is often spotty, sometimes contradictory, and sometimes controversial. So it's no wonder that articles about fictional worlds and characters quickly become much more detailed than articles about the real world. And the spottiness and unambiguity of information - especially of online information, which is by far the easiest to research and reference - quickly increases the farther you go backward in time. Even the third one is due, in part, to the difficulty of researching information in one language while writing in another. So I think all three of those problems would exist regardless of who was editing Wikipedia - in other words, the bias of the users or the system is less important than the "bias" of the information itself.

Interestingly, of the examples in the New York Times article attempting to illustrate gender-based bias, two of them can actually be attributed to these other skews as well. For the first example, Niko Bellic is a fictional character, who is popular today, while Pat Barker is a real person, who did most of her writing in the 1980s and 90s (and who was never, as far as I know, that well-known - but that's another story). And the second example actually relates to all three: characters in The Simpsons are fictional, the show is currently still airing, and it's an American TV show (that airs globally); while Mexican feminist writers are real people, lived throughout the last 150 years (or longer), and, of course, write/wrote in Spanish.

This is not to say that there is no major problem with the content that tends to be of interest to women. The examples cited in the New York Times article give a mixed but inconclusive view of that - and unfortunately, as far as I know, there has been no comprehensive study on the issue. But let's assume, for the sake of the argument, there is a big problem - the unfortunate brevity of articles about shoe designers is one indicator of that. Still, I find it telling that, even in examples meant to be about gender, these other skews show up as well; they're so pervasive as to be inescapable.

There's no demographic group that can be recruited to successfully address these other skews - yes, you could bring in older people and people from non-Western countries, and it might have an impact to some extent, but the inherent difficulty of the research would still be there.

Wikimedia is not powerless about these issues, though, and in fact it's working on them - through initiatives like GLAM (galleries, libraries, archives and museums), where professionals from such institutions get involved in adding data, because they have knowledge of, and easy access to, all the relevant source materials. There's no focus on demographics with GLAM - the focus is on the content, and so far it seems to be working very well.

So why not take that same approach, and expand it? If a dearth of female-oriented content is in fact a major problem, why not forget about who's editing Wikipedia, and instead target what's being edited? The Wikimedia Foundation could make a list of 100 or 1,000 or 10,000 topics that need expanding - female-oriented or otherwise - and make it official policy that articles on those topics, in every language Wikipedia, should have their size and quality increased. There are already "WikiProjects" on the various Wikipedias that work toward this kind of thing - it could just be elevated to official policy. This way, we can cut straight to the actual goal - after all, there's no guarantee that a random woman who's convinced to start editing Wikipedia is more interested in writing about Mexican feminists or Manolo Blahnik than about, say, Grand Theft Auto.

Now, you could argue that there's no reason not to do both - to focus on both the content and the editorship. But I would say that there are two important reasons why emphasizing gender distribution is undesirable in itself. The first is a moral one: there's certainly an easy argument to make that, barring any strong reason for it, people shouldn't be judged based on their gender (or other such attributes), but solely on their individual abilities and contributions. When the Wikimedia Foundation declares that it wants to increase the percentage of female editors, it's implicitly saying that every new male editor that arrives is hurting one of its goals, due to his gender - a morally tenuous position at best. And the second is a pragmatic one: editing on Wikipedia overall has been in decline since 2007 (reversing this decline is actually another of the WMF's major goals). Whatever the cause of this decline is, and however serious you consider the problem to be, is it really wise to be assigning different merit to different new editors? One would think we should just be thankful for any new person who decides to start contributing, and that judging people based on anything other than the quality of their contributions is a luxury (if you can call it that) that the WMF can no longer afford.

So, my proposal is: focus on the content, and get out of the judging-by-gender game. Ask people to contribute more of the type of content that you find important, and they may just surprise you - whether or not they happen to be cursed with a Y chromosome.