Friday, June 12, 2015

Why Women Don't Code. Echidne Answering Comments From Another Reality.

I don't usually write blog posts about anonymous comments elsewhere, but these particular ones are too juicy to miss, because they give us the condensed arguments from anti-feminists as to the reasons why coding is so male-dominated, and thus offer me a chance to respond to those arguments.

The first common argument is that girls just don't like to code.  Here's the current version of it, from Eschaton comments:

Are we really going to pretend that as many women as men want to code? For fuck's sake.
That's worth pondering about.  Note, first, that there was this magical era, about thirty years ago, when computer science students were almost equally likely to be men or women.  Then something changed, and in a fairly short time the field was one for the bros.  This post goes into much more detail about it.

So in a sense it might be true that women don't want to code now, given the work environment they'd meet (with people like the one I quote above probably working there)*.  But there's no evidence that a dislike of coding would somehow be an inherent female characteristic.   If it were, it would have been operating in the early 1980s, too, and I personally suspect that not as many women would be in accounting, say, if the idea of sitting still and staring at numbers or letters wouldn't at all apply to women.

The second common argument is that girls just can't code.  Or:

Right, male and female brains are exactly the same.
Did you know that male and female IQs are the same on average, but that males have more representation on both tails of the bell curve?
Have you considered the implications of this?

This is a bit hilarious, because the extreme upper tail in the IQ distribution seems to be equated with being a coder!  I doubt that, to be honest.  We in the extreme upper tail are not many enough to fill that occupation.

This commentator is saying that guys are smarter (and also more stupid, but that's the other guys, the losers), based on the peculiarity that the results from IQ tests on men show greater variance than the results from IQ tests on women.

I have written about various aspects of the "extreme tail" theory about gender and intelligence many times before. This post summarizes several of the issues:  The gender differences in the extreme tails are getting smaller over time (which suggests that we are not looking at something innate), gender differences in attitudes towards risk (such as likelihood of guessing when the answer is unknown) could create the observed results if men were less risk-averse than women (or played the testing game that way), even if there was no underlying average gender difference in the characteristic under study and so on.

Third, sooner or later this particular conversation turns into a wider one about women and men being innately different In Ways Which Matter (no, men don't have wings and fly and women don't have fins and swim; we are supposed to regard small differences as humongous for the purposes of allocating societal power; and no, differences in aggression etc. do not matter).  This time we get:

OK, let's play make pretend. Yes, women on average are interested in heavy tech, coding, data...
Male and female brains are exactly the same. Oh, except when someone is transgender, then male and female brains are different...

There you go.  As a bonus we get one of the reasons for some trouble between certain feminist strains of thought and certain strains of trans thought.

The research into gender differences in brainz is pretty difficult to do, especially if someone wishes to argue that those differences are both innate and not subject to change.  I've written loads and loads on various studies (start here and here, say), but even if one accepted innately different brains by biological sex (or by gender identity???), that acceptance would not tell us that women can't code or don't want to code.

But it's not the question whether women and men would be equally eager to code if the industry wasn't so broish (and if there was no steering of boys and girls into certain culturally accepted roles) that fascinates me here.  Who knows what would happen in that ideal world?  The point of those comments I analyze seems to be that what exists TODAY cannot be at all improved, because of that assumed innate stuff.  This means that whatever percentage women make out of all coders is the RIGHT percentage.

And just to make everything crystal-clear, the commentator also notes that if there are girls in the industry then they can be found in non-essential roles:

Well, research, data, technical project management are the roles that matter...
I'm sure women are represented in marketing, HR and PR, which are simply less important to the heavy tech companies.

I hope this post didn't infect you with a stereotype threat.

*See also this post about a study concerning the effect of gender stereotypes in getting access to math-heavy fields.