October 28, 2014

Women In Computing

My grandmother was, for a time, a newspaper woman. She was a rare bird in that role, and it must have helped that she brought a college degree to a job where, for men, a high school diploma was plenty. And I expect she worked cheap.

My mother was, for a time, a newspaper woman. She wasn’t such a rare bird: her boss was a woman, and there were plenty of women on the paper in all sorts of roles. “Don’t every let them know you can type, Patsy” was her boss’s sage advice.

After first wave and second wave feminism had hit the beach and after the sixties had opened the professions (and Chicago’s Berghof restaurant) to women, that particular fight seemed to be settled. I’ve worked in science, engineering, and technology since I left college, and while women were seldom the majority in the room, they were seldom scarce. That has certainly been the case in hypertext research, the corner of computing I know best: women have always made up a substantial plurality of the audience, about half the program committee, and perhaps a little more than half of the top jobs.

In short, I’ve pretty much assumed that the work of our generation was incremental, a matter of getting from 40% to 50% and smoothing out the welds. Tough work, sure, and not to be minimized, but this was also work not to be compared to the early days of suffrage and barricade.

Well, not so fast. Women now make up almost 50% of medical students and more than 40% of physical science students, but only 17% of computer science majors. I find Bob Martin’s diagnosis unconvincing and his final anecdote, intended to be demonstrate redemption, is deeply wrong-headed.

When I get a program to work, I feel like I've slain the beast and I'm bringing home meat. She responded by saying that she felt that she had nurtured something into being.

Bob Martin, of all people, should know in his bones that these are two different kinds of programming, that both are needed, and that neither has anything at all to do with sex or gender. Few have written better, and no one has written more, about the interconnection of debugging and design. Debugging, in my experience, always demands intensity and sometimes requires rage. When rage slips into design, you’re in terrible trouble: you detest your pointy-haired management, you despise your ignorant users, you loathe the code, and either your work won’t communicate what you really think (in which case it is a dishonest lie and everyone will know it), or it will. To be intelligible is to be found out.

The conventional wisdom formerly dictated that you build separate teams with separate responsibilities: designers and coders and testers. Of course, that always leads to wrangles over who’s in charge and how much they get paid, and all that rage shifts from the bugs to the company, to those sons and daughters of bitches who won’t let you do your work properly and who are wrecking the product. Martin himself was instrumental in showing us a better way, demonstrating how refactoring (relentless refactoring forsooth) could slay the beast and nurture the design. These a guns to which we should stick.

Stacey Mason skillfully argues the case for a plague on all houses without asserting the false equivalence that is the worst nonsense of #gamergate.

I’m disappointed with 4chan for a harassment campaign of such incredible scale, but I’m disappointed with social justice because I expected more from us than name-calling, mocking, and immaturity.

One aspect of the computer science numbers to which I think we need to pay attention is the context of the Great Recession. Lots of people in school are desperately worried about jobs. Medicine no longer promises the prospect of luxurious sinecures, but you’ve got to figure that you can get a job. Physical science opens lots of doors. A generation ago, you could hope to get a job at some startup like Microsoft and come away, a decade later, with a few million bucks. A generation ago, Route 128 was dotted with computer companies. They’re all gone now, gone with the hope of making a lot of money by being Employee 37, doing your job and not getting fired.

Chemists and physicists can learn a ton of programming if they want to, and they can wind up in science or industry, in research or finance or software. I went that route, and while it’s probably harder to do nowadays, I bet you could still do it. I think it’s much harder for someone with a CS degree to do research in physics or chemistry.

So CS enrollments are way down, across the board. I don’t know why these forces would exert differential pressure on men and women. I don’t know that they do. If they do, I don’t know how one would demonstrate the effect.

But the notion that dumb jokes and boors are responsible for driving women from computing doesn’t ring true, not to my ear. I don’t really know many men or many women who would be scared off by bad jokes and innuendo. I admit I don’t know a lot of college kids, but it seems unlikely that kids today have less gumption than we had. The women I’ve known in computing and in science wouldn’t be deterred, and in point of fact they weren’t.

Nor were my mother and my grandmother, and surely, back then, those reporters were a plenty hostile to their intrusions. First wave, second wave: I guess we’re still storming that beach.