This is my take on the “why are women so underrepresented in technology in general, and open source in particular” debate, inspired by several things that have come up lately – primarily the publication of the FLOSSPOLS report of findings & recommendations, a thread on the women@apache mailing list, and a seemingly innocent comment on my Frustrations! post…
A brief technical bio: I’m a woman who only really got interested in computers late in my school career, because they were something I felt I was good with. I had computers around earlier than that, but frankly, I played the games I was given – I wasn’t programming in BASIC, or ‘peek’ing and ‘poke’ing to make my screen do funny things. I’ve worked in both closed-source (in Microsoft), and open-source (httpd-docs/ASF) environments. I’m a fairly active member of several allied organisations – amongst them SAGE-IE, ILUG and Linuxchix. I’m also studying Computer Science at university, as part of a cross-discipline degree (my other subjects are Linguistics & German).
First off, sexism: my experience has been that, with the exception of one of the organisations mentioned above, the majority of the people I interact with on a technical basis are not overtly sexist. Sure, you get morons who think that because I’m a geek, I clearly can’t get a boyfriend, and they’re only too willing to step up to the plate. Or worse, they presume that even though I have a boyfriend, and am clearly unavailable, I’m such a minority that really, they have some kind of bizarre right to harass me. But thankfully, they’re a minority.
In two of the groups I’m a member of – one male dominated, one female dominated, sexist behaviour, where it does occur, is absolutely not tolerated. I can’t think of any occasion where I’ve seen truly sexist behaviour in either of those groups, but I think that’s because it’s so clear that it won’t be tolerated. In the other group – also male dominated – sexism is, at best, ignored. This, frankly, is only minutely better than encouraging it, and thus it flourishes. This practice also leaves the group with a fairly sexist ‘feel’, even when there’s no active sexism ‘right now’.
A much bigger problem, however, is subconscious or covert sexism. This is a bigger problem because it’s more insidious, and less likely to be noticed by others – oftentimes, even those who are being sexist don’t notice it, or would describe their behaviour as something other than sexist. The FLOSSPOLS report shows again and again that women see and experience sexism that men just don’t believe is there. And certainly, it’s possible that the motivations of those perpetuating these problems are entirely innocent – but the net result is an environment that is hostile to women.
The idea of meritocracy comes in here. On the surface, it’s a completely fair, non-sexist, open concept. Anyone can get in, anyone can progress, as long as they’re good enough.
That’s very, very rarely true. Generally, at best, a meritocracy turns very quickly into a merit-and-confidence/pushiness-ocracy. Good work doesn’t win you influence – good work that’s pushed in others’ faces, or at the very least, good work of which others are regularly reminded – wins you influence. And that’s where women fall down. Women are every single bit as capable of good technical work as men. Sure, there are fewer women who have developed and practised their technical skills – but even that doesn’t account for the disparity that’s apparent in the open source world.
There’s no real way of excising the need for confidence – at the end of the day, a contribution has to be made public somehow for it ever to become part of an open source project. But this is definitely a bar that can be lowered, without lowering the standards of merit required by anyone. A meritocracy should be just that – influence earned by merit, decisions made by those who show merit, governance by quality, not confidence.
The final problem with meritocracy is that even after all the noises of “it’s all about the quality of contributions”, women very often aren’t judged on the same basis as men. This is one of the few areas that FLOSSPOLS have looked at where both men and women perceive there to be a problem. People listen or pay attention to women, or don’t, based on the fact that they’re female – not based on the merit or otherwise of their contributions. Call it what you want, that’s not meritocracy.
I think the FLOSSPOLS report is really well worth reading in its entirety, but my notes follow.
Women aren’t passively avoiding open source – they’re actively excluded from it, even if the actions aren’t intended to exclude women. Most open source projects have a “way” – there’s “the Debian way”, “the Apache way”, “the FSF way”, to name but a few. Some of these are formalised, some reside in the minds of the current members of any given community. Most of these “ways” set code above all – which is all very well, but results in software that frankly, is considerably less useful than it could be. It also indirectly disadvantages women, who typically tend to be involved in areas such as documentation. Note that this is often because documentation projects are less sexist – because they have more women anyway! – not because the women choose to document because they’re not capable of coding.
These “ways” also tend to value independent work and discovery – which is all very well, but women typically come to computers later in their careers, and are thus at a disadvantage from the start. Expecting them to catch up on their own, and yet regarding the playing field as level, is rather less than realistic. Fused Silicon and Free Software describes this in a little more detail. This another big problem with the practice (as opposed to the concept) of meritocracy – it’s often not about how good you are, it’s about how well you know the project that’s sitting in front of you. Which, of course, requires either the confidence to ask hundreds of questions, until you get the answers you need, or the perseverance to just keep poking at it – the latter of which is hard to maintain in a hostile environment.
Open source communities often see themselves as ‘apart’ from the rest of society – but they lump women in with ‘society’. Women are either assumed to be male (online), or the subject of intense scrutiny and attention (online & offline), both of which can be horrifically uncomfortable situations to be in. Interestingly, it’s thought that a high proportion of contributors to open source software are on the autistic spectrum – which itself has a disproportionately high number of males.
Aggressive behaviour, while less common amongst those “higher up” in any given community, is nonetheless relatively common in the open source community. This can result in women “failing” before they start, because of lack of knowledge of the true hierarchy, and lack of confidence in their technical abilities.
Women often have difficulties integrating into open source communities because they spend less time on this type of work. This is often not because of lack of interest, but simply because of increased time spent on other “responsibilities”. Men counter this with “that’s just social conditioning” – but neglect to mention that social conditioning is a powerful motivator, in both directions. Again interesting is the note that whilst consumer usage of technology is widening (and thus, more women are likely to be getting early access to technology), women’s involvement in technology in general – both commercial and academic – is decreasing. This is especially pronounced in open source software.
The most interesting recommendation I saw was to sponsor exchange programs or joint projects with parts of the world where coding is not a “male” activity. I think this could be useful for both men and women. Hell, if I could find somewhere where coding was an inherently female activity, I’d be most interested in seeing how some of the men I know would cope!
However, the most useful recommendation I saw was to encourage those in leadership positions to recognise that people are being actively put off, not just failing to participate – and that this is a problem for open source work – now and in the future.