Prejudice masquerading as fact
Angela Saini, author of Inferior: How Science Got Women Wrong, on that memo.
A portion of his argument is indeed based on published science. In particular, there is a school of neuroscience that tries to popularise the notion that male and female brains are distinct. It claims that female brains are typically hardwired for empathy, while male brains are built to analyse systems, such as computers and cars. This all hinges on the idea that autism represents an extreme form of the male brain, caused by exposure to higher than usual testosterone levels in the womb. Yet recent experiments have repeatedly failed to find a direct link between foetal testosterone levels alone and autism.
Indeed, psychological studies show that there are only the tiniest gaps, if any, between the sexes, including areas such as mathematical ability and verbal fluency. Navigating this complicated field for my latest book, Inferior: How Science Got Women Wrong, I was told by a prominent American researcher into sex difference that he no longer refers to brains as sexually dimorphic, because the science simply doesn’t support this. There isn’t a neuroscientist alive who can say with confidence which sex any given brain belongs to.
In short the science in the memo is “flawed,” but the memo is getting lots of support anyway.
What they fail to understand is that there are published scientific papers out there to support every possible opinion, even that black people are intellectually inferior to white people. Getting published doesn’t make an idea true, it only means that someone has managed to get it into print. In evolutionary psychology, theories are sometimes little more than speculation strung together with scant evidence.
There was a time, she points out, when eugenics was considered good science.
Weak scientific evidence and empty theories are still being used to support troubling ideologies. Women are making enormous strides in science and engineering – yet, with some half-cocked hypotheses in their back pockets, male software engineers feel they have the right to tell them they are somehow biologically unsuited to this kind of work.
They forget, perhaps, that many of the world’s original computer programmers were women, including the first: Ada Lovelace. Women began to be marginalised in technology around the time that personal computing took off and become a lucrative industry. Male software engineers forget that discrimination and sexual harassment have driven women out of Silicon Valley, and kept countless more out in the first place.
The myriad historical, cultural and social factors that create inequality are all too easily glossed over when someone reaches for the closest, most convenient biological explanation for what they see. This isn’t just intellectual laziness; this is prejudice masquerading as fact.
It’s also men being assholes.
I have no doubt that there are many sexist and “bro” culture attitudes in the tech industry which do contribute towards the preponderance of males who decide to live with it and which discourage equal or better qualified women. Such attitudes and behaviours do need to be identified and corrected. But I don’t really think the Damore memo is necessarily one of them.
The real point of that memo (and of Larry Summers’ remarks at Harvard) seems to be that when looking for evidence of unfair treatment of any particular group (eg women) it is more convincing to look at what happens for jobs needing near the middle of the distribution of whatever skill or attribute is being rewarded because there the numbers from all groups should be in proportion to the group sizes regardless of even relatively large “gaps, if any, between the sexes” . But even “the tiniest gaps” between averages can lead to substantial differences in the proportions achieving the highest skill levels, so disproportionate representation in those positions is not necessarily evidence of unequal or unfair treatment if even the tiniest gaps might actually exist
Yes, it might be that the memo was intended as a provocation, but it could also just be an expression of frustration with the use of an invalid argument to support a thesis which may well be true. Of course quibbling about whether or not the probably true thesis is being supported by a valid argument may be just something done by “men being assholes”. After all you know, men…logic…assholes they all go together don’t they?
It doesn’t quite work like that, Alan.
Just as education is being ruined by the foolish notion that what is easiest to quantify just must be the most important indicator of whether progress is being made, something very similar happens with recruitment, performance evaluation and promotion.
I can illustrate this with the story of just one man. This actually happened to me over a period of years so, yes, it is anecdote but anecdote can also be fact. I was working for a fairly large international non-profit, one founded almost a century before by a group of women, a group of very forceful women who perceived the chaos, the almost total destruction in some parts, the life-threatening shortages of food across Europe in 1919 differently from the men sitting around that famously shiny table at Versailles. The leader of the team, an articulate woman with all the “right” social connections was threatened with a treason trial for wanting to get in there and actually feed starving children.
But anyway, decades later this organisation, now huge, was recruiting a new Director of Fundraising and we ended up with Mr X. His experience was in venture capital and he had been the voluntary chair of a small charity in Bristol – on the face of it a perfectly respectable bloke. But look closer. Almost all of our systems – finance, admin, marketing, global communication – were computerised: he had never used a computer, refused to learn, refused to have a computer in his office. He had never worked with women in executive roles: the place was replete with them. We ended up with his de facto deputy director doing his secretarial work and his highly qualified secretary sitting in the corner twiddling her thumbs, hoping that someone would quietly slip her something to do before she went bonkers. There were also a good many other thing he didn’t know about but he either didn’t recognise that or refused to be told. So one can only conclude that his only qualification, appointed as he was by men and without a detailed person spec, was that he, too, was a man.
He didn’t much like me and certainly made my work difficult. I was head of a small multi-disciplinary team, some of our work quite technical but also requiring good people skills. We were producing a high and rising income at very low cost. Not to bore you to death, he eventually manoeuvred me into early retirement, part way through the installation of a bespoke computer system. I wish I had been there the day the systems analyst with whom I’d been working stormed into his office demanding to know why he had got rid of me. One excuse was that I’d never be able to manage the new computer system. The answer? “She fucking designed it, you expletive, exexpletive, expletive.” All this at high volume and heard by several reliable witnesses from whom, obviously, I heard it at second hand.
Don’t worry, I survived!
Alan, the issue I see with the memo, and all arguments based on similar statistics, is that they are all woefully incomplete. When noting an observable phenomenon of some sort, determining the cause of such involves more than proposing a single explanation and then concluding that it must be true, yet this is all memo guy has done. Granted, having decreased aptitude for mathematics / science / etc would certainly result in a dearth of women in those fields, but what the cause is something else?
And so I am reminded of a video involving Neil Degrasse Tyson, where he was asked the question “What’s up with chicks in science?” (Yes, really, and by an adult no less.) Anyway, he answers it in terms of being a different group underrepresented in those fields – black – and notes that social pressure is another plausible cause of disparity, and further that the social pressure needs to be removed before we can re-examine the disparity (if it remains) for a ladybrain cause.
Oh, look, the white male thinks it’s obvious that discrepancies in outcome are best assumed to be the result of discrepancies in ability.
Meanwhile, multinational studies of boys’ and girls’ math exams have shown that the more equal a society is, the tinier that tiny gap in scores becomes. Further backing it up, introducing stereotype threat before an exam significantly lowers the scores of girls (or black and Hispanic kids, in tests on racial rather than sex exam score issues).
*This* white male doesn’t think (and never said) that “discrepancies in outcome are best assumed to be the result of discrepancies in ability”. In fact once discrepancies in outcome exist for any reason, the natural tribalism of humans may tend to exaggerate them and so corrective measures will always be needed on behalf of under-represented groups.
So long as the discrepancy persists, the corrective measures will be needed, BUT persistence of the discrepancy is not itself evidence that the corrective measures have not been strong enough. Such evidence is not provided by the simple statistics of counting people but rather by the more subtle data embodied in experiences that are sometimes dismissed as anecdotes.
Mr. Cooper, please look up “sea lioning”. Prestigious professions are swarming with mediocre-to-incompetent men, prestigious colleges are awash with legacy admissions getting by on “gentleman Cs”…and disparities disappear when the process becomes gender-blind, whether this is symphony orchestras or Harvard admissions (I was an undergrad there during that flap, so this should satisfy your “anecdotal” requirements).
But of course the opposite is also true: if there is even a tiny discriminating effect against women, then the cumulative effect can be large. If (for instance) women are only 10% less likely to get promoted than men, then after just six rounds of promotion you would have nearly twice as many men at the higher rank. And if women are meeting tiny barriers at every step, then they can add up to huge differences in numbers, even without any underlying difference in ability. Given the fact that there is plenty of evidence that there is such discrimination against women, it seems perverse to leap to the conclusion that the outcome must be the result of some inbuilt predisposition.
Here’s a wacky idea – what about waiting until the social pressures actually have been removed before we conclude that women just aren’t as good at X because biology?
In other words…the social pressures – witness this “memo” itself, and the bros gathering to halloo it – are so massive and so powerful that it seems perverse to run around squawking about possible tiny differences between averages.
“Here’s a wacky idea – what about waiting until the social pressures actually have been removed before we conclude that women just aren’t as good at X because biology?” Well, it depends on what the X is. With regard to mental skills it’s not wacky at all (but re being an American Football lineperson it might be).
And of course adverse social pressures still need to be removed even if we do eventually conclude that women aren’t as good at X on average (even in the case of the football linepeople, where I am sure qualified females *do* exist and should not be deterred by social pressures).
In my opinion (already expressed above) it may always be necessary to have compensating practices such as affirmative action in order to counter a natural tendency of disparities to grow without reason. There are good reasons for this regardless of whether or not there is some level of “justified” disparity but they need to be correctly explained to those affected by them and the mere fact of a disparity is not proof that it is an unfair one.
Giving a wrong argument in support of a correct action is often not productive – and can even be counterproductive by reducing trust in the correct argument when it is finally presented.
But who says it is? Who ever says it is? It gets wearying having this supposed claim rejected over and over when it’s not in fact a claim that gets made.
I don’t think anybody says employee percentages should correspond exactly to population percentages. I think the claim is that if there is a large disparity that is likely to be the outcome of hiring practices or workplace culture or both rather than just the percentage who apply and are competent.
The outcome of every study conducted with a modicum of rigor shows that any prevailing wide gender disparity goes immediately away upon blinding. So, yes: in this case a disparity of the magnitude seen in Silicon Valley is proof that it’s artificially imposed and maintained (especially when early CS was dominated by women before it became lucrative and prestigious). I do realize that large-scale facts won’t dissuade people fixated on minutiae as a way of diluting the importance and consequences of such biasing.
I think the claim that “if there is a large disparity that is likely to be the outcome of hiring practices or workplace culture or both rather than just the percentage who apply and are competent” is easily confused in weak minds like mine with the claim that “the mere fact of a disparity is proof that it is an unfair one”.
Indeed the only distinction I see there is the inclusion of “likely to be” which converts the claim from one of proof to one of strong likelihood – which for the total effect is still not something I agree with. What I would say myself (after providing some fuller explanation) is “if there is a large disparity, that is almost certain to include the outcome of discriminatory workplace culture rather than just the percentage who apply and are competent” – which expresses a *stronger* level of certainty in only a *partial* effect (and does not suggest any deliberately selective hiring practices). Of course, we know from other evidence that the disparity sometimes *is* entirely the result of hiring practices and/or workplace culture, but any effort to convince the beneficiary of that is hampered rather than helped by including unsound arguments along with the sound ones.
The only distinction? The word “large” is another distinction, and it’s the same one I was pointing out. Speaking of sound and unsound arguments.
I believe we’ve entered a phase with Mr. Cooper that would make his fellow sea-lions proud.
Then, Alan, you are not thinking scientifically.
Whichever side of the argument a given published paper may come down on and many are ambiguous we are talking about very small differences. So if you set up an experiment which finds a difference between men and women in a particular aptitude then you need to repeat it several times and with large samples. Suppose that you do this 17 times and the difference is always the same one and always statistically significant – practicing scientists will tell you that this itself is almost impossible – what then?
Well your first task would be to ask someone professionally competent to see whether there is something in the design of your test which is creating the effect and test the same variable in a different way.
Then you’d be asking yourself why something like a 1.3% or whatever difference accounts for a 70-plus difference in recruitment or promotion. Logic, let alone the scientific method, tells you that something else is going on which makes Ophelia’s wording entirely reasonable.
And I do hope that, after all that, you’ll stil remember that these are averages which would show up if you converted your data into a scatter diagram. Twenty published papers all with the same result tell you absolutely nothing about the person sitting before you asking for a job, be that person man or woman.
Well, I didn’t see the word “large” as a distinction because I might as well have said “the mere fact of even a *large* disparity is not proof that it is an unfair one”. It is quite possible for one population to have a *lower* average value of some parameter and still be overwhelmingly over-represented at the very top (and bottom!) end.
I am virtually certain that there *is* a bias against women in tech culture which explains at least a large part of the discrepancy – but I believe that more for other reasons (including both personal experience and things like the blinding experiments referred to by Athena Andreadis) than for the mere fact that the discrepancy exists (or is large). And I think that even Ophelia’s version of the “if there is..then it is likely…” statement is counterproductive to the goal of convincing people.
But apparently I am irritating some of you so I won’t comment further unless asked to by Ms Benson.
You’re assuming that the rest of us think there’s a bias against women in tech culture which explains [some of] the discrepancy solely because it’s a discrepancy and not for other reasons of the sort you mention? Why would you assume that? And you must be assuming it, because why else point out that there are additional reasons?
Aren’t the additional reasons what we’ve been pointing out all along?
Also. Wait a damn minute here. I didn’t read this carefully enough the first time.
What the hell? I make a deliberately limited claim and you simply announce that “weak minds” will read it as an absurdly absolute one? On the basis of what?!
All along you’ve been attacking the absurdly absolute claim, so I tell you that’s not the claim, and say what the claim is, and your response is just that “weak minds” will read it as the one you started with. I hope you can see what an unreasonable move that is.
I will take those direct questions as an invitation to respond just once on that issue only.
Re #17: I am not assuming anything about what you think, and by agreeing with many arguments that justify affirmative action and workplace culture mitigation I did not intend to be seen as “pointing them out”. But I still think that combining them with language that can easily be (mis?)interpreted as a weaker argument undermines their effectiveness.
R e#18: The word “if” can easily be misinterpreted (esp by techies) as the beginning of an implication.
When I say “If A , B” such an interpretation might be that A by itself logically entails B where the real intent might have been something else.
I don’t know what to say about that. If the word “if” can easily be misinterpreted I don’t know how we can discuss anything at all.
Wow, this discussion has really gotten long. But…one thing I notice about Alan Cooper is that he seems to think we are not making the arguments we should be making, but in fact, those are the arguments we have been making all along, not only on this discussion of Mr. Damore, but in all the years I’ve been reading this blog. The arguments from evidence have been rife…So I think he’s strawmanning, not sea lioning. That’s just my observation, though.
On the anecdotal side: I was once an intern in a wetland restoration job that I really did want to become permanent. As an intern, I was the top of the list for hiring when a job came open. At the end of my internship, there was a job open, and I put in my application. Within a few days, I received a letter explaining that the job had been withdrawn; they weren’t doing any hiring at the time. So I moved on, and found another job (in air quality, in which I had no training or experience, but it was primarily clerical, so why would it matter to anyone but me?) Within a week of my leaving the internship, the job was re-posted. I found out who they hired. They hired someone with a bachelor’s degree and 2 months of experience for a job requiring a master’s degree and 2 years of experience (both of which I had). There was one thing I did not have that the new hiree did have: a Y chromosome. Or, if you prefer, a penis. I will admit, I never actually saw either the chromosome or the penis, and so have to rely on secondary sexual characteristics, as well as knowing the young man (since the 2 months of experience had been on the project I had led as an intern).
Lest you say, oh, yes, but one incident. At this office, another intern (male) had come up for permanent status at a time when there were zero jobs available, and a hiring freeze. In spite of that, he was made permanent. They created a job out of thin air, wrote up a lot (and I do mean a lot) of paperwork to justify the position, and made sure he got his job. Another intern, a female, came up for permanence about eight months before I did. She was informed there was no job available. She left and went back to school. A week later, they hired someone (a male) to fill the position that did not exist. A person who did not have nearly as many qualifications as the woman that was told to go away. I realized after a while working there that there were almost no permanent employees that were female, and those that were had been there before the current (at that time) administration had taken over. In an office of 15 full time employees, there was 1 female. It wasn’t because females didn’t want the job; we were applying. We were being hired as interns, led to hope for a permanent position, then dropped on our ass as soon as our internship came to an end. In this way, they made it look like they had a fair hiring policy, because anyone walking in off the street would have seen about 30-40% women, and not known they were all part time and temporary.
This is again just an anecdote, I realize, but sometimes that is where you begin looking for patterns. And everyone there was whispering about the fact that they didn’t hire women, because everyone knew it was true.
God damn, that’s infuriating.
They’re not mutually exclusive. In fact, I’d describe it as the triple combo of sealioning, strawmanning, and concern trolling. “Pardon me, can we have a polite fact-based discussion of why Damore’s argument may be a correct response to this really bad argument I have imagined that feminists make? I, of course, know that you aren’t making that argument, and I completely agree with you, but I am concerned that your real argument might be confused for a bad one.”
“You’re not making this silly argument, but let’s talk about what would happen if you were.”
What Screechy said in comment 23. File Mr. Cooper’s analysis under “dissecting gnats and swallowing camels” — or the saltier Hellenic equivalent: “The world’s on fire and some are busy combing their pubic hair.”