As I see it

Posted Aug 1, 2009 10:54 UTC (Sat) by man_ls (guest, #15091)
In reply to: Patterns of denial by alankila
Parent article: OSCON keynote: Standing out in the crowd

It is somewhat offensive to say "terminal estrogen poisoning" (just as with the male version), so you have to be ready to bring proof that the alleged poisoning is really causing trouble. Estrogens are much milder hormones than testosterone, so the context has to be chosen carefully: surely it would not come as a surprise if I told you that my prebirth classes were a bit too estrogen-laden. But, as it turns out testosterone poisoning does not have a scientific basis (apparently aggression is more related to some androgens), and "poisoning" is a loaded term anyway; so the testosterone version is also a pejorative term and it should be used with similar care. If you see nix's original context he was sarcastically referring to random flaming on some extreme "macho" mailing lists, so there should be no reason for offense. How's that for being wikipedantic!

The problem is that too many people were "laying traps" such as yours, instead of pondering what was being discussed. This has led to 170+ posts carefully winding around the issue of female participation in free software projects. This behavior is not typical here on LWN: presumably it would not have happened if we were discussing about kernel pointers or free software licenses, and that is a pity.

to post comments

As I see it

Posted Aug 1, 2009 19:48 UTC (Sat) by alankila (guest, #47141) [Link] (21 responses)

Anyway, I'm not going to discuss biology on this topic.

The problem here is that there is not enough data for me to make any kind of reliable conclusion. One thing that irks me is that the discussion goes like this:

- industry has 20 % female participation
- oss has less than 2 % participation.

=> OSS has a problem, we should have 20 %, too. Let me offer one potential cause for why not: different incentives. OSS work is mostly unpaid, industry work is not.

I'll back this up, like everybody else, with an anecdote. My girlfriend is a programmer, but she is not an OSS programmer. Programming is just a day job for her, she doesn't really like coding and doesn't derive much pleasure from it. So she's not interested in spending her spare time on it.

As I see it

Posted Aug 1, 2009 21:16 UTC (Sat) by mjg59 (subscriber, #23239) [Link] (13 responses)

That's not a plausible model. You need an explanation for why women in the computing industry are less likely to also have it as a hobby than men are.

As I see it

Posted Aug 1, 2009 21:31 UTC (Sat) by nix (subscriber, #2304) [Link] (3 responses)

Also, OSS work is not mostly unpaid, and hasn't been for years (despite
strange outliers like me).

As I see it

Posted Aug 2, 2009 12:44 UTC (Sun) by alankila (guest, #47141) [Link] (2 responses)

Let's consider the word "mostly". Is it true that over 50 % of OSS contributors are paid programmers now? I definitely have missed that change. I thought most of us were just a freelance army...

As I see it

Posted Aug 2, 2009 19:55 UTC (Sun) by nix (subscriber, #2304) [Link] (1 responses)

It's been true for many years that >>50% of code is written by people
employed to do free software development. Whether >50% of *developers* are
paid for it is, as you point out, a somewhat different matter: you'd
expect things to take longer to reach that point.

As I see it

Posted Aug 2, 2009 23:06 UTC (Sun) by Baylink (guest, #755) [Link]

> It's been true for many years that >>50% of code is written by people

Again; your source?

As I see it

Posted Aug 2, 2009 12:42 UTC (Sun) by alankila (guest, #47141) [Link] (8 responses)

Different priorities in life? I thought it's quite self-evident that women and men don't enjoy quite the same things.

As I see it

Posted Aug 2, 2009 19:54 UTC (Sun) by nix (subscriber, #2304) [Link] (7 responses)

It is fallacious to extrapolate generalities about both genders to
anything about any specific members of either gender. The sexes are much,
much more similar than they are different.

As I see it

Posted Aug 2, 2009 20:29 UTC (Sun) by alankila (guest, #47141) [Link] (6 responses)

... and as you can see, I was making a quite generic statement about men and women ...

As I see it

Posted Aug 2, 2009 21:36 UTC (Sun) by nix (subscriber, #2304) [Link] (5 responses)

"Women and men don't enjoy the same things" would explain these ratios
only if *almost no* women enjoyed software development. As the much more
even gender ratios in proprietary software development make clear, this is
not the case. Some other explanation is necessary.

As I see it

Posted Aug 2, 2009 21:40 UTC (Sun) by Baylink (guest, #755) [Link] (4 responses)

Do you have some justification for assuming that a majority of people -- even in IT, or more specifically programming -- do what they do because they like it, rather than merely because it pays quite well?

Certainly those are not mutually exclusive, but we wouldn't have the phrase "golden handcuffs" if that were not a consideration.

As I see it

Posted Aug 2, 2009 22:20 UTC (Sun) by nix (subscriber, #2304) [Link] (3 responses)

In computing in the UK there was a long period when it was considered
the 'default employment choice' for young males with no particular
interests, because it paid well. The universities gorged on this and by
2000 some had single yearly intakes so large that they had to split
lectures into several groups to fit them into the lectures theatres.

Then the dotcom crash hit. That supply was choked off, and has not
recovered.

I suspect that anyone who's worked in the field across the dotcom crunch
is either doing it out of desperation because they have no other
experience, or is doing it because they're hooked. (Some people who joined
later may have done so for the pay, but it's been pretty stagnant since
then in my corner of the field in any case.)

(And in any case, are women more or less likely to do things because of
money than men? Again you're derailing, changing the subject from 'why is
the gender gap so large' to 'women are Just Naturally Different', when no
differences in ability or inclinations of more than a few percent have
ever been documented in any properly controlled study I've ever heard of.
Again, we are not a very sexually dimorphic species as mammals go: any
argument for the absence of female free software developers that is based
around assumptions of radically differing abilities or inclinations
between the sexes is probably incorrect and should be considered only as
the last option.)

As I see it

Posted Aug 2, 2009 22:29 UTC (Sun) by Baylink (guest, #755) [Link] (2 responses)

Wow.

Again, no I'm not.

The fundamental issue here, as you are playing it, is "the majority of men in FOSS development behave in sexist ways, and much change their behavior".

(If that's *not* the argument you're making, please tell me now.)

That is an extraordinary assertion (that 80% of 20,000 someodd people need to change the way they behave), and requres, as I have noted before, extraordinary proof.

At the least, it requires you to show your work as you go, and you continue not to. I've kept my mouth shut on this for a couple of days, since as a couple of people noted, it could be construed that I was sucking up all the oxygen in the room... and lo and behold, several other poeple (appear to me to) agree with my fundamental position.

My argument was precisely that I see no gender difference in "doing things solely for the money", even though you try to paint what I said the opposite way. In this case, the person "derailing" is you: the argument is yours: that all developers are doing it for love, not money, and I don't believe that a whit.

As I see it

Posted Aug 2, 2009 23:45 UTC (Sun) by mjg59 (subscriber, #23239) [Link] (1 responses)

The fundamental issue here, as you are playing it, is "the majority of men in FOSS development behave in sexist ways, and much change their behavior".

That's not the case. There's no reason to believe that the majority of people in the open source community are sexist. The concern is that there's a minority who engage in sexist behaviour and a larger body (perhaps a majority) who either don't recognise this behaviour or who aren't concerned about it. To deny that this behaviour exists is unrealistic. Anecdotes don't provide statistics, but when they're describing verifiable events they do prove whether or not something has occured. So really, what you're disagreeing with is the scale of the problem. And that's fine. I don't have hard statistical data to back my belief that it's fairly significant. You don't have hard statistical data to back up your belief that it isn't.

That's disappointing to some extent, but it makes little difference. The changes in behaviour that would help here are simple things like not using sexually loaded language, not making off-colour jokes in a technical environment, not turning a conference into a sexualised environment by hiring women in short skirts purely to hand out flyers. These things cost approximately nothing, but doing them removes a great deal of the perceived sexist environment and means that anyone engaging in more flagrantly sexist behaviour is more obvious.

Will that magically get the number of women involved in free software up to 20%? No, of course not. But if there is a barrier to reaching that figure, then the onus should be on the people who want to maintain that barrier to justify it.

But let's get back to where this thread started. "Women and men don't enjoy the same things" would explain these ratios only if *almost no* women enjoyed software development, followed by a discussion of whether people do things for love or money. That's not the point. Let's say that 10% of men involved in the software industry enjoy it enough to get involved in free software in their spare time. If the same figure were true of women then we'd have the same 20%/80% split (approx) that we see in the commercial world. Given that the actual figure is more like 2%/98%, if differing levels of innate interest are the reason then the figure for women has to be around 0.8%. Or, to put it another way, women have to be over ten times less interested in engaging in free software development than men. This seems an awfully large figure to assign to nature. Of course, if 10% is an overestimate (which it is) then the 0.8% figure drops in the same way. At which point it becomes pretty obvious that "Women and men don't enjoy the same things" would explain these ratios only if *almost no* women enjoyed software development is true.

So "Men and women are just interested in different things" doesn't seem to hold. You're being accused of derailing because you've managed to turn a discussion of sexism in free software into an argument about whether or not people write software because they're paid to do it or not. That might well be an interesting discussion to have, but there's no realistic way that it's relevant to this issue. And by changing the topic you imply that the original topic isn't the important thing here. Which isn't a great way to reassure people that their concerns are being taken seriously, which in turn isn't a great way to convince them that they'll be able to fit in. Which is where we came in.

As I see it

Posted Aug 3, 2009 20:18 UTC (Mon) by nix (subscriber, #2304) [Link]

An accurate description. I cannot imagine what misreading of what I had
said could lead to anyone interpreting it as '80% of men are sexist'. That
is plainly ludicrous and nobody has proposed it. (It is also tangential
and thus yet *another* bloody derailing.)

Sidestepping the issue

Posted Aug 1, 2009 23:21 UTC (Sat) by man_ls (guest, #15091) [Link] (6 responses)

My girlfriend is a programmer, but she is not an OSS programmer.

A veritable red herring. I work directly with 30+ developers, men and women; none of them is a free software developer but me, and I'm just a small time contributor to a few random projects. What does this show? Nothing, it's an unrelated issue.

The problem here is that there is not enough data for me to make any kind of reliable conclusion.

Let us see if we can find any data -- wait, the announcement that started this discussion contains a lot of hard data, and from women already involved in free software development. Don't you think that solving the problems perceived by these women (sexism, lack of respect, death threats) would be a first step, and that we can try to make it more comfortable at least for those that choose to join us?

Do we really need to play with statistics? Imagine that we were discussing this link:

Alleging a string of racist incidents and management indifference, a group of black Halifax firefighters files a human rights complaint.

And when talking about why there are so few black firemen in Halifax, I said: "I don't know, maybe black people don't enjoy fighting fire". Don't you think that we would be sidestepping the real issue?

Sidestepping the issue

Posted Aug 2, 2009 13:04 UTC (Sun) by alankila (guest, #47141) [Link] (5 responses)

I know anecdotes are worthless. That's why I ironically used one, and copiously even flagged it as such. I disagree on one point: I refuse to believe that it didn't support the point I was making.

Let's assume OSS work would be just like any other work: same background level of sexism, lack of respect and even death threats. My point was that even if you fixed all that to same level as industry in general, I doubt we would have 20 % women. The anecdote I used was there to distract you about the more important point which is that incentives in OSS work are different in industry work. And that matters when you have two groups which seem to have different priorities in life. You will get different participation rates.

I will not go into black firefighters in this topic. I know you are implying that I have a blind spot just because we are discussing women and not blacks. However, I insist that my contribution is not worthless.

Let's assume we do all the things that would "fix" our community, and in 10 years we find that we have 10 % women instead of 20 % as is industry average even then. We will still be talking about subtle sexism and structural misogyny and things like that, if we just keep on staring at "20 %" and "10 %" and insist that this is evidence of a problem.

Exacerbating the problem

Posted Aug 2, 2009 13:54 UTC (Sun) by xoddam (subscriber, #2322) [Link] (3 responses)

Anecdotes are only worthless if your interest is in assessing trends and nothing else.

We aren't particularly interested in assessing trends here. Obviously most of us are caring, sharing men and a small but very much appreciated minority are caring, sharing women.

The interesting statistical question is, why is the female minority so small. And a very plausible answer is, because some men behave badly. The question of *how many* men are behaving badly isn't half as interesting as how many women there are, or how many there might be if the men did not misbehave.

Any problem of sexist behaviour in our community is not a major trend. It consists a few isolated examples of totally-unacceptable behaviour, of which anecdotes are an existence proof.

Oh, and another problem consists in this kind of thread, which demonstrates that as long as no-one mentions sexism everyone is cool, but as soon as it is adduced a small minority of posters make a very loud noise trying to insist that no problem exists, or that if it does it is statistically insignifigant, or that it's not as bad as problem Q "in the real world", or that women have the upper hand these days, or that even if a problem does exist in this community, the poster isn't responsible, etc.

When you do that, you don't make the problem go away, you exacerbate it.

You are defending the indefensible.

Please don't.

Exacerbating the problem

Posted Aug 2, 2009 15:07 UTC (Sun) by alankila (guest, #47141) [Link]

My mistake. I thought it was fair to discuss it. I didn't realize everyone needs to be on the program and just nod their heads in agreement.

Anyway, I grow tired of this, so you get your wish.

Exacerbating the problem

Posted Aug 2, 2009 23:10 UTC (Sun) by Baylink (guest, #755) [Link] (1 responses)

> Any problem of sexist behaviour in our community is not a major trend.

This is precisely the opposite of the argument which has been being made here by you, njs, nix, man_ls, and others, all week long.

Care to clarify?

Exacerbating the problem

Posted Aug 2, 2009 23:45 UTC (Sun) by xoddam (subscriber, #2322) [Link]

Unfortunately, a few people are arseholes.

Others, also a minority, quibble and "derail" in a defensive manner, questioning the validity of the whole discussion. The quibbling and derailing *of itself* contributes to an atmosphere of hostility, the existence of which the quibblers contest.

Most people are neither arseholes nor quibblers, therefore this is not a general trend. But it's still a problem.

Clear now?

Sidestepping the issue

Posted Aug 2, 2009 17:30 UTC (Sun) by man_ls (guest, #15091) [Link]

Let's assume we do all the things that would "fix" our community, and in 10 years we find that we have 10 % women instead of 20 % as is industry average even then.

Let us not just assume it; let us do it. Let us fix our community so that we don't harbor sexism anymore. Once women feel comfortable within us (or at least as comfortable as men) we may continue debating the other issues: interests, abilities and whatever you like. Until then, it is like discussing whether aspirin or paracetamol are best for skull fractures: you may have a point but it's better to take care of the bigger issue first.