5 ms·
Seriously! I have strong opinions about the role of women in our industry and society, as does pretty much everyone. But ignore the gender/culture argument and
by drewrv 9y ago
Seriously! I have strong opinions about the role of women in our industry and society, as does pretty much everyone.
But ignore the gender/culture argument and replace diversity with "Foo". The stated goal of the organization is to increase Foo, from top to bottom, from HR to engineering. What do you think will happen if you publish a letter saying Foo is overrated? What will happen if you make getting Foo harder for them? What will happen if you semi-publicly argue against increasing Foo?
- Cookingboy 9y agoActually that's not what the author of the memo advocated. He did not say Foo was overrated, he said he supports Foo, but he questions the effectiveness of the current company methodology in increasing Foo. He thinks there are other root causes for why Foo is not as high as they want, and he proposed other ways to increase Foo. He was doing a good job by all means, but then people categorized his point of view as anti-Foo when he was simply against the current ways of increasing Foo.
- meowface 9y agoExactly: "pro-diversity" is not necessarily incompatible with "maybe we should reconsider affirmative action-esque initiatives". I disagree with many things in the document, but, by all accounts, the employee had no issues with a diverse workplace. I think the case he was trying to make is that maybe the lack of representation of women in programming and similar fields can't entirely be explained by sexism or discrimination or bias, and that if it's true that women are on average less interested in things like programming, then it doesn't necessarily make sense to make a commitment to, say, "double the amount of women we hire this year" (not a real example; just hypothetical). The pool for women interested in the job may just be lower in numbers even when controlling for all social and cultural factors. Of course, the science on that is still far from conclusive (and many of the "scientific" facts he cited sound dubious), and this absolutely can't be extrapolated to suggest that female programmers who do work at Google or other companies are less interested or are somehow less competent. I don't think the author tried to make that claim at all, nor did he suggest that sexism and bias wasn't an issue contributing to the gender gap or an issue that shouldn't be stamped out, but a lot of people (and the media) seemed to draw those conclusions after reading the document.