Philosophy, Science / Tech

The Google Memo: The Economist On Nothing

Most of the debate about James Damore’s memo has focused on its claims about gender, diversity, and affirmative action. Those themes were indeed central to the purpose of the memo. But also important were themes that often got overlooked: reason, open discussion, and classical liberalism. In a way, Damore got some of what he wanted—more discussion about the first set of themes—although he no doubt wished he could keep his job too. Now that there has been so much discussion of those themes, now that the dust has settled after “Googlegate,” it’s a good time to reason through the best arguments on each side of the controversy. Who was right? What can we learn? How can we do better next time there appears to be a clash between the competing values of equality, science, and freedom of speech?

Many of the best arguments on Damore’s side can be found in his own memo.

This may come as a surprise to those who developed their opinion about it, not by reading the memo itself but by absorbing accounts of it in the popular press. The misrepresentation began as soon as the story broke. First to publishing the memo, but without its bibliographic links, Gizmodo’s influential account led to widespread criticism of it for being nothing more than unsourced prejudice. Ignoring not only its research, but also Damore’s many assurances that he values diversity and wishes only to criticize the ways in which it has been pursued at Google, Gizmodo began the tradition of calling his memo an “anti-diversity screed.” It was all downhill from there.

Quillette resisted this momentum, however, by quickly publishing the testimony of four behavioural scientists, who endorsed Damore’s main arguments, at least when it came to their fields (social psychology, evolutionary psychology, and sexual neuroscience). Lee Jussim’s endorsement was strongest, writing that Damore “gets nearly all of the science and its implications exactly right.” More qualified was the assessment of David Schmitt. “In the case of personality traits,” he wrote, “evidence that men and women may have different average levels of certain traits is rather strong.” But Schmitt did not infer from these differences the critiques of Google’s diversity programs that Damore did. Instead, he offered a nuanced endorsement of both affirmative action and open discussion of “some of the real psychological sex differences that might account for variation in men’s and women’s workplace performance.” One can only hope that such subtle and informed assessments will be remembered.

What about the best arguments on the other side? Even if Damore had the support of those scientists, what about the arguments leveled against his memo by others—other scientists, as well as other thinkers more or less aware of the relevant science? There are many considerations relevant to this debate, after all, besides science—political considerations of equality and freedom speech, or rhetorical considerations of the present cultural climate—so the viewpoints of many different sorts of thinkers have been pertinent to it. Surveying them with hindsight, though, we should notice how support for Damore was usually proportional to knowledge of the science of sex differences, evolutionary psychology in particular. Much of the opposition to Damore’s memo, in other words, was either ignorant of the evidence regarding sex differences or openly hostile to the relevant fields.

But that was not always the case. The best argument for Damore’s firing did not come from Google CEO Sundar Pichai, alas, who rationalized his decision by writing that Damore had advanced “harmful gender stereotypes in our workplace.” That’s not a reasonable justification unless a carefully argued memo that accurately reports the current state of evolutionary psychology is thereby pernicious. Pichai’s argument failed, in short, because it was anti-scientific. This was not the case with the best argument for Damore’s firing, however, which came, not from anyone at Google, but instead from The Economist, whose editors imagined themselves as the founder of the company. With this conceit, they wrote “The e-mail Larry Page should have written to James Damore.” Even Jonathan Haidt saw merit in at least one of its arguments, if not the whole “e-mail,” tweeting it with approval.

Upon first glance, it is an impressive document. It reminds me of a speech by Gorgias, one of the greatest of the Greek Sophists. He lived during the classical period, a democratic time when knowing the art of persuasion could make you powerful, or at least help you succeed in court. Young aristocrats flocked to Gorgias, who became wealthy by teaching them this art. In order to attract his customers, he gave demonstrations of his skill. On Nothing was the most dazzling exercise of this skill: Gorgias sets out to prove, believe it or not, that nothing exists. How better to convey your persuasive art than to choose the most implausible conclusion … and make it seem true? Gorgias achieves that result, so long as you listen to the speech recited by a confident orator. Each argument has just enough plausibility to impress a listener, and the speech moves on to the next one before anyone listening has a chance to submit any of them to scrutiny. The cumulative effect is of a massive pile of evidence for his conclusion. And yet, when you read and re-read the text of the speech, analyzing each argument critically, you learn that each is unsound.

As Cardinal Newman observed, in very different context, ten thousand ponies do not make a horse. Gorgias’s trick—let us call it the pony fallacy—is also demonstrated by The Economist’s “e-mail.” It amasses specious arguments, apparently hoping that a multitude of bad-but-seemingly-good arguments will amount to a genuinely good one. That hope seems to have been fulfilled before audiences—those who read the major American news outlets, judging from the unanimity with which they portrayed Damore’s “anti-diversity screed”—who are disposed to believe its conclusion, as well as those who might be disposed to reject its conclusion but lack the patience necessary to treat its arguments one by one. For the only way to expose its sophistry is to pause long enough over each one to evaluate its soundness. That’s likely to become boring—unless, like James Damore, you value reason, enjoy its exercise, and believe that the truth is found that way. Let’s imagine ourselves to be such an audience.

Seven Major Flaws

“Larry Page” begins fairly enough, conceding two of Damore’s main points: first, that there are dozens of differences between the average woman and the average man when it comes to abilities, aptitudes, and interests; second, that the gender gap at Google (four male coders for every female one) does not by itself prove unjust discrimination. But the “e-mail” then proceeds to accuse Damore of motivated reasoning and duplicity of other sorts. It also says that his memo is “derogatory to women in our industry,” and “demonstrated profound prejudice.”

In order to substantiate these serious accusations, The Economist had to expose flaws in his arguments, or provide sound arguments of its own. It endeavors to do both. “Now that we’ve worked out what your memo is really about,” begins this effort, “let’s examine its argument.” It then proceeds to catalogue seven major flaws in the memo. One of those flaws—the last in my order or presentation, second in theirs—introduces ten particular objections. These are each supposed to be gender differences that Damore ignored in his biased memo. Seven flaws, ten objections: by my count, “Larry Page” adduces sixteen arguments. But however you count them, Gorgias would be proud!

1. First Major Flaw

The first of the seven major flaws that “Larry Page” claims to have found in Damore’s memo is that it argues from a gap in outlier gender proportions (on traits) to “differences in men’s and women’s ability to code.” This is wrong on several counts. First of all, as The Economist concedes, this is not what Damore says: “At least that’s what you seem to be doing; you don’t quite say so.” This is not an auspicious beginning: the article is criticizing him not for what he does say but for what he seems to be saying. In other words, it strikes a straw man.

In fact, Damore argues that there are population-level differences between men and women in interests, and because of these differences there is far more interest among the male population in becoming a coder than there is among the female population. The inverse is true for other fields. When it comes to college students in the health professions, social work, education, and psychology, for example, over 75% of them are women. With so much more male interest in computer science—roughly 80% of majors in this field are men—there are more male applicants for coding jobs. Without affirmative action for women in hiring, this proportion persists into the staff of companies who draw their applicants from this pool.

Damore never says women are less able than men to code. Although he does mention abilities once, that is part of his general argument for natural sex differences. For there are some differences between the average man and the average woman in some very specific abilities (spatial rotation, for example). But his argument about Google and coding does not appeal to any specific ability-differences. Instead, he stresses differences of interest, using several specific examples to make this point. “Many of these differences are small and there’s significant overlap between men and women,” he nevertheless adds, “so you can’t say anything about an individual given these population level distributions.”

2. Second Major Flaw

Next, write the editors of The Economist, “you’re ignoring everything else that could explain the gender gap.” This overlooks the context of the memo, which is ironic, because Damore has so frequently been accused of ignoring the context of his speech. He attended a compulsory diversity seminar. At the seminar, according to him, it was said dogmatically that the gender gap at Google was a product entirely of social factors: bias, unfair discrimination and sexual harassment. He knew this claim to be false because he was familiar—perhaps from his graduate studies in biology at Harvard—with the latest science about natural sex differences. So he wrote up what he knew, to challenge the simplistic and dogmatic line of the diversity seminar (which seemed to be that of Google itself).

“I am not denying that there is sexism,” he stresses in the first line of his memo. Nor was he denying that bias, unfair discrimination, and sexual harassment (etc.) were factors in the unequal representation of women among software engineers. “Of course, men and women experience bias, tech, and the workplace differently,” he wrote on its third page, “but it’s far from the whole story.” For the same reason, he was careful to say that the science about sex differences could “in part” explain unequal representation. In this objection, then, the editors at The Economist have committed the part-whole fallacy: Damore’s argument required showing only that natural differences were one part of the story; The Economist faulted him for failing to tell the whole story. That was neither his purpose, nor his obligation.

Damore posted his memo to a “Skeptics” forum in-house. He did not distribute his memo to the whole company, as was widely and falsely reported in the media. He was challenging a false dogma of the company in a forum established for that purpose. By his account, he wished to hear criticisms of his argument in case he was overlooking something. No doubt he also wished to be heard, believing he was right. For if he was, he argued, the efforts to foster diversity at Google were being undermined by their ideological blindness to reality; they could succeed better if they were based on a recognition of natural sex-differences rather than on dogmatic social-constructionism.

For this effort he was often called “anti-diversity,” which he was not. He was for diversity, including race and gender diversity; what he was against was the way that it had been fostered at Google. He argued that this way, when it comes to gender, would remain inefficient and potentially unjust until it is informed by knowledge of the relevant natural differences in male and female populations. “If we can’t have an honest conversation about this,” he wrote in the introduction to his memo, “then we can never truly solve the problem.”

Wanting to solve the problem, Damore’s main target was the shame-culture that inhibited honest discussion of it. His main fear was that Google had supported this shame-culture by creating an ideological echo chamber. His subsequent treatment by the company confirmed this fear. The mainstream media and the academy, who applauded Pichai’s decision and continued to misrepresent a publicly available document, showed the echoes reverberating well beyond this one company. Judging by the public reaction, at least among the intelligentsia in America, the problem is widespread. Damore is important not so much because Google is important—although obviously it is very important—but because he is a canary in the mine.

3. Third Major Flaw

The third major flaw that “Larry Page” finds with Damore’s memo is that, “the gender differences you cite differ between countries and over time.” This is true, but it is not evidence against Damore’s premise. Consider, with Schmitt, an analogy to height. There are gender differences in height. In every culture, men are taller than women. (As always in this field, we are talking about populations—the average man—and not about individuals. It can’t be emphasized enough: some women are taller than some men; some women could be taller than all men; and so on, without contradicting the claim.) This is because men are naturally taller than women. But, someone could object, this difference in height differs between countries and over time. Yes, goes the reply, this is true, but it is no evidence against the claim.

The difference in height between the genders is narrower in countries with poor nutrition, wider in countries with good nutrition. This does not show that the gender difference in height is not natural. On the contrary, it argues that the difference, while natural, exhibits variations in its breadth—narrower in some countries and times, wider in others—due to environmental factors. It is likewise with psychological differences between the genders. The sort of differences mentioned by Damore are cross-cultural. Yes, they vary in breadth from one country to another. But, as Schmitt and his team showed in a watershed study, these differences are widest in countries with the most egalitarian cultures, such as in Scandanavia, and narrowest in regions that are least egalitarian, such as in Africa, the Middle East, and East Asia.11 This is only one study, but it was executed by over one hundred social scientists working in fifty-six countries.

The revolutionary conclusion drawn by Schmitt and his team in the face of this “paradox” is that when people are given the most freedom to be who they want to be, they exhibit their natural interests most. When you pause to think about it, it’s not paradoxical at all. It would only seem so to someone who approached the data in the grip of social-constructionist ideology.

The Economist did not name its source for this argument, but in the days after the Damore story broke, a post by Suzanne Sadedin making just this argument got a lot of attention. However, her argument was quickly refuted—using the logic of the height analogy—in a post by “Artir” that was less celebrated but more rigorous—

The distinctions and inferences at this level of debate are quite difficult to assimilate without serious attention and study. Not surprisingly, then, the simpler interpretations of data have gained wider currency, leading to pseudo-scientific journalism that asserts confidently that Damore’s science is completely bogus—an example of motivated reasoning, if anything is. So it is ironic to read “Larry Page” accusing Damore of the same.

 

4. Fourth Major Flaw

Another major flaw that the editors of The Economist find with Damore’s memo—the fourth in my reckoning—concerns his use of sex differences to make his argument about Google: “they don’t even support your argument, because you don’t seem to understand what makes a great software engineer.” This was the best point made by Yonatan Zunger in a blog post that circulated rapidly in the days after the Damore story broke:

That post was mendacious in too many ways to enumerate here (my critique of it should still be buried somewhere in the comments section). But there is even a fallacy hidden in its best point. Everything Zunger says about the difference between a great engineer and a merely competent one could be true (and I suspect it is), without affecting Damore’s argument.

To become a great engineer, you have to first be a competent one. That is to say, you must be hired as an entry-level coder before you can rise to become a project manager. Thus, if the differences Damore cites would predict the gap at the entry level, this would act as a bottleneck toward the higher-level project manager jobs, affecting those proportions too. The women and men hired for the entry level jobs will (without affirmative action) have the same abilities and interests for the entry-level work: that’s why they were hired, they’re all excellent and motivated. If the gender gap widens at the higher ranks, that too will need to be explained. Part of it could be due to bias, discrimination, and harassment; but part of it could also be due, as many behavioural scientists have argued, to population-level differences between men and women. Damore alludes to these arguments, particularly the argument that men are more driven to achieve status, but he doesn’t cite any sources. There is no shortage of research, however, corroborating his allusion.

5. Fifth Major Flaw

A fifth major flaw with Damore’s memo, according to “Larry Page,” is that “you clearly don’t understand our company, and so fail to understand what we are trying to do when we hire.” Nothing more is said. As stated, it is a meaningless objection. As such, there can be no meaningful reply.

6. Sixth Major Flaw

On to the sixth major flaw, therefore, which is that “even if you are right that more men than women are well-suited to the job of software engineer at Google, you are wrong that taking steps to recruit more women is inherently unfair to men.” Normative conclusions about gender policies are far trickier than empirical conclusions of behavioural science, because normative arguments presume more than descriptive ones do. As for his own normative presuppositions, Damore does not hide them: “I consider myself a classical liberal and strongly value individualism and reason.” He also says very clearly what he takes to be unfair: “treat people as individuals, not as just another member of their group.”

Google is already treating people as members of groups (e.g., men or women), so, if Damore cites good science about the natural characteristics of these groups, company policy should reflect a knowledge of these characteristics. There is nothing untoward in someone posing this challenge to a company’s practices: “You are basing your claims to fairness on a false, social-constructionist ideology; I have shown that the ideology is false; the basis of the fairness argument is now gone; what, if any, is the new basis of fairness?” There may be a new basis, supporting better policies, and that is what Damore tried to start a rational conversation about.

As it happens, there is a good basis for a fairness argument—marketability of their products to women—but it arises later in the “e-mail,” and will be addressed then. In the meantime, let us note that whatever are the facts of psychology and behaviour revealed by the best current science, no normative conclusions follow from them without further argumentation (and some normative premises). Damore is suggesting that such normative discussions happen in light of the relevant scientific evidence. Instead he was treated as though he had violated a sacred value, confirming what he was ultimately claiming about the company’s ideological monoculture.

7. Seventh Major Flaw

A seventh, and final, major flaw in the argument of the memo is supposed to be that it focuses on some gender differences (to shore up its conclusion) without looking at other gender differences (that would undermine it). Before proceeding, let’s remember that this is a memo by an amateur and not a book by an expert, so it can’t reasonably be expected to say everything. Nevertheless, when considering the soundness of its argument, it is legitimate to doubt that its selection of evidence is biased. What, then, are The Economist’s grounds for asserting that Damore has been biased. What, in short, has he ignored?

7.1 First Objection

“Men score higher on measures of anger and lower on co-operation and self-discipline,” write the editors, before adding: “If it had been the other way round, I’m betting you would have cited these differences as indicating lack of suitability for the job of coder.” It’s no critique to speculate what Damore would have done if “it had been the other way round.” There’s no basis for that speculation except the imputation of “motivated reasoning” that persists throughout the “e-mail.” But the imputation of motivated reasoning is legitimate only after the argument of Damore has been shown to be unsound. Since it has been shown to have withstood all the criticisms thus far, the memo has not been shown to be motivated reasoning, so the “e-mail” has not earned the right to that imputation. Furthermore, it is never explained why those features upon which men score lower are specifically relevant to coding. If there are careers where those features are central—say clinical psychology—behavioural science would predict a deficit of men. It does, and there is.

7.2 Second Objection

For his second such objection, “Larry Page” writes: “You lean on measures of interest and personality, rather than ability and achievement, presumably because the latter don’t support your hypothesis.” There’s no need to presume anything. The differences in interest are major and, yes, they support Damore’s hypothesis. That’s what you do when you present an argument, you present the evidence that supports it. For every hypothesis there are infinitely many facts that do not support it. For obvious reasons, you don’t lean on them. But if there is evidence among them that contradicts your hypothesis, and you deliberately ignore it, you are not arguing in good faith. Has Damore done that? Well, “Larry Page” has yet to show that.

7.3 Third Objection

The third gender difference that Damore ignored is quite general. “In many countries girls now do better in pretty much every subject at school than boys,” begin the editors, but “if it had been the other way around I’m sure you wouldn’t have neglected to mention that fact.” Again, this is speculation; and yet the editors are “sure.” Furthermore, it is a problem that girls in many countries are doing better in pretty much every subject than boys are. Notice: “in pretty much every subject.” That would be a just development only if girls were naturally “better” (more ability, more interest, or both) than boys in pretty much every subject. But they’re not. Girls are “better” at some things naturally (as a population), whereas boys are “better” at others (as a population). That’s Damore’s point, substantiated by the science summarized in the meta-meta-analysis at Heterodox Academy.19

7.4 Fourth Objection

The fourth gender difference that Damore ignored is quite specific. “The sole published comparison of competency in coding I am aware of,” write the editors, “found that women were more likely than men to have their GitHub contributions accepted—but if they were project outsiders, this was true only if their gender was concealed.” First of all, the difference was very small: 78% of blind female submissions were accepted, while the figure for male submissions was 74%. When their gender was revealed, however, the percentage of female submissions accepted dropped to 74%, while the male figure remained the same. This hardly seems statistically significant—although I am no statistician—least of all when it occurs in only one study. There are individual studies in this field to suit everyone’s prejudice. Sober thinkers thus rely, whenever possible, on meta-analyses, which make the Heterodox Academy post so valuable. “Based on the meta-analyses we reviewed above,” it concludes, “Damore seems to be correct that there are ‘population level differences in distributions’ of traits that are likely to be relevant for understanding gender gaps at Google and other tech firms.” At the end of his “e-mail,” “Larry Page” urges Damore to read this meta-meta-analysis, oddly unaware that it corroborates his case.

7.5 Fifth Objection

Next, The Economist introduces a fifth gender difference that Damore ignored: “There is plenty of evidence that women in Silicon Valley suffer sexism.” Except that the first line of his memo says “I … am not denying that sexism exists.” In any case, this is a red-herring. Damore simply argued that sexism (bias, discrimination, harassment, etc.) cannot be the only explanation of the gender gap at Google (as was asserted dogmatically at the diversity seminar to which he was responding) because there are also natural differences.

7.6 Sixth Objection

Before introducing the sixth gender difference that Damore ignored, “Larry Page” grants that there are natural sex differences in animals. “With humans,” he nonetheless advises the Harvard graduate in evolutionary biology, “you must take great care before concluding that any specific difference is innate, since our societies are so much more complex and varied than those of other animals.” Yes, indeed you must. The people who take such great care are scientists, particularly evolutionary psychologists. David Buss’s introductory textbook is an excellent place to start learning about their work.

7.7 Seventh Objection

The seventh objection tendered by The Economist presents “some reasons to be doubtful about an evolutionary basis for the specific differences you cite.” Apparently, before the 1980s, when computers were exclusively marketed to men and boys, “a much bigger share of those studying computer science at university in America were girls than is the case now.” To be clear, this statistic is not that there were more girls than boys studying computers then, but that there was a higher proportion of girls (as compared with boys per capita studying computers) then than there is now.

That’s of course interesting on first blush, and could prove significant upon further investigation, but it could mean almost anything until it’s investigated or explained more. How different is the proportional change? How different is the absolute change? What were other changes in the marketplace and higher-ed (generally and for women) that would have affected these numbers? Such are the questions that a scientist in this field asks and investigates. At his Slate Star Codex blog, psychiatrist Scott Alexander already has some persuasive suggestions: “there were lots of women in CS [computer science] in 1980 for the same reason there were lots of Jews in banking in 1800: they were banned from doing anything else.” As he adds, with quotations from women’s magazines of the era to support his suggestion, “computer programming was originally considered sort of a natural outgrowth of being a secretary.”

7.8 Eighth Objection

For its eighth objection, pinpointing gender differences that Damore is supposed to have ignored, the editors write: “Even personality differences vary from time to time and place to place.” This seems to be the same point made under the third major flaw, invoking the same reply. Generally speaking, behavioural scientists are aware of this fact and incorporate it and all its attendant subtleties into their theories. They offer hypothesis, test them, revise them in light of the data, and so on. To which of these subtleties are the editors referring? To which theory are they requesting revisions? Nothing is mentioned. Instead, “Larry Page” says only: “That suggests that at least some of the gaps we see in America are because women are still relatively powerless.” This is a social-constructionist hypothesis, suggesting (but only suggesting) that the gender differences that appear natural are in fact due to power differences. Yet this hypothesis has been tested by the cross-cultural studies and it’s failed to predict the robust results. Indeed, the results turn out the opposite of what it would predict (namely, gender differences grow larger as power becomes more equal). That’s scientific fact rather than “suggestion.”

7.9 Ninth Objection

“Larry Page’s” ninth objection is that “those supposedly ‘female’ traits vanish in the rare arenas where the competition is entirely among women.” Yes, women tend to behave differently when they are competing with all women than when they are competing with men or with a mix of men and women. The same goes for men. This has been studied extensively by psychologists and there are good data-sets, evolutionary theories, and conclusions that withstand critical scrutiny. The author is presenting as a vague objection against evolutionary psychology something that has been studied by it, with sophistication, for a few decades. See, for example, the work of Anne Campbell.

7.10 Tenth Objection

The tenth and final objection, to the effect that Damore has ignored relevant gender differences, goes as follows. “Senior engineers must manage teams—and by your own reasoning that should mean that women, with their greater empathy and interest in people, should be over-represented at that level, compared with their numbers in more junior jobs. That they are not should have given you pause.” Indeed, it should have. But remember: Damore was not trying to explain everything about the Google workplace. This fact about the gender gap in senior positions has been noticed by scientists in many, many fields, not just tech—the academy, for instance. They have studied it, and have developed sophisticated explanations of it.

In some instances, of course, social factors (bias, discrimination, harassment) are primarily responsible (see Uber for an example). Perhaps they are often part of the story. But evolutionary psychologists have argued that natural sex differences are always part of that story too. Generally, men are more stratified as a population when there is competition for external goods, whereas women are less stratified when they are competing. There are more men than women in lucrative leadership positions, yes, but there are also more men than women in poorly-paid and dangerous jobs such as those Damore listed (“coal-mining, garbage-collection, and firefighting”), not to mention among the imprisoned and the homeless. There are straightforward explanations for this greater stratification from one of the central elements of evolutionary psychology, namely sexual selection theory. Men who took greater risks and succeeded passed on their genes more than men who either took greater risks and failed or avoided risks. The same was not true of women.

Sexual selection is a form of natural selection where one sex prefers a specific characteristic in an individual of the other sex.

Affirmative Action

Despite his numerous fallacies, “Larry Page” does make one sound point. The gender gap in one design team at Google led it to design software that was not appealing to women. Google Plus required new users to identify their gender on a public profile. “Presumably it didn’t occur,” write the editors of The Economist, “to anyone involved in development—all of them men—that many women choose to conceal their sex online to cut down on harassment.” This is a good argument for why design teams should have more women on them than they do. They are trying to market products to a population, after all, that is roughly half female. They should have many women participating in the process of designing those products so that they don’t make such mistakes.

This is a good argument for affirmative action, at least in business, that someone such as Damore should endorse. Even if it might appear unfair to some men (let’s imagine they were rejected for the project despite being as good or better at coding than the women were who were accepted or recruited for it), tough! Google’s trying to make good products, and this affirmative action program would make sense within that operation. Ironically, having someone who knows natural gender differences (an evolutionary psychologist) would also help a lot within that operation. Nothing sells like something that appeals to our innate interests. Indeed, having many people conversant in the arguments of evolutionary psychology, people who were encouraged to offer their arguments rather than being fired for making them, would strengthen the whole company. That was Damore’s main point.

This leads to a more general argument for affirmative action. “Larry Page” thinks it’s a good idea to implement it to overcome the resistance some women would have applying to so heavily male an environment. “Unless we make special efforts, some women will be put off applying to the heavily male culture.” This is an empirical generalization. As such, it should be studied—if it hasn’t already—and science rather than this vague intuition should be adduced as argument. In the meantime, here’s an argument, made by Scott Alexander, to be doubtful of it.

In the Victorian period, in the Western world, all the professions were almost exclusively male. In the 20th century, as these countries worked towards gender equality in education, the law, and so on, women went into the workplace. Within a relatively quick period (the fifty years of 1960-2010, let’s say), they achieved gender parity in some fields, exceeded men in some fields, and fell short of male numbers in other fields. Why? Why, for example, did women flock to the male dominated fields of tax collecting, optometry, and veterinary medicine? Why did they achieve something close to parity in photography and bus driving? Why are they still so outnumbered in construction and engineering?

“Everyone says, ‘Aha! I bet it’s because of negative stereotypes!’” remarks Alexander, whose post is an extraordinary piece of careful reasoning and research on this whole controversy, necessary reading for anyone holding an opinion in this debate. “This makes no sense,” he nonetheless answers. Why not? “There were negative stereotypes about everything! Somebody has to explain why the equal and greater negative stereotypes against women in law and medicine were completely powerless, yet for some reason the negative stereotypes in engineering were the ones that took hold and prevented women from succeeding there.”

And what about Sweden? The Swedes have tried harder than any other nation to achieve “equal distribution between men and women in all domains of society,” Yet the gender proportions among engineering students there are the same as they are in the USA. Moreover, 97% of Swedish truck drivers are still men while 89% of child-care workers remain women. Like the “paradoxical” results of the massive cross-cultural study of gender differences, comparative labor statistics can be initially surprising. In poorer and less egalitarian countries, where women have less freedom to determine their careers, gender disparities sometimes shrink. “Whereas most people would expect to find many more female engineers in the U.S. and Sweden than in Columbia and Bulgaria,” writes Maria Charles, “new data suggest that precisely the opposite is true.

Science, Reason, Freedom

Finally, “Larry Page” concludes with a bizarre list of suggestions that do not merit more than a rudimentary critique. “You could have asked some of your female colleagues,” he writes, “about their experiences in the industry.” But nothing they could have said would have touched Damore’s scientific argument. This recommendation puts personal narratives and feelings above facts, which is one of the things Damore criticized about discussions at Google. “You could have looked for evidence,” adds The Economist, “that conflicted with your biases.” But that’s the job of the behavioural scientists upon whose peer-reviewed work Damore was relying. If you rely on the best climate scientists to present a case, in a ten-page memo, that anthropogenic climate change is happening, it is no valid criticism of your memo to say that you didn’t look for evidence that conflicted with your biases. You would be right to trust the climate scientists to have done that.

That the editors of The Economist cannot see this in the case of behavioural science suggests that they never really trusted the findings of this science in the first place. If so, this confirms my suspicion that condemnation of Damore is typically inversely proportional to familiarity with, not to mention respect for, the science upon which he based his arguments. “Larry Page” began his “e-mail” by conceding that there are dozens of differences between men and women, implicitly endorsing the science upon which such claims are based, but in “his” practice of argumentation this endorsement is tepid at best. “He” does not go so far as to reject the whole field, the way some popular critiques of Damore have, but flirts with the contempt for science—whenever it seems incompatible with political ideology—that characterizes many of the thinkers on this side of the debate.

“I shouldn’t have had to write this,” concludes this particular critique; “I’m busy and a little effort on your part would have made it unnecessary.” I know how you feel, Larry. But don’t worry: it was not all in vain. We have considered arguments on both sides of an urgent question and learned more about the world as it really is, rather than as our prejudices would have it be. So, where should we—whether at Google, in the academy, or in our broader national culture—go from here? Damore points the way.

First of all, let us bypass, as he does, the obsolete dichotomy between science and the egalitarian goals of feminism. They are not incompatible. To begin with, one is an empirical enterprise that describes the world; the other is a political movement that seeks to change it. They are no more in conflict than are math and Marxism. Rather than seeing evolutionary psychology as an opponent, then, feminists should see it as a tool. For no political movement’s goals can be effectively achieved without a good understanding of the world it addresses. (Strategies for minimizing workplace harassment, for instance, should take into account the Error Management Theory of Haselton and Buss.) Feminists should follow the example of intelligent Christians, who resisted Darwin’s thought at first, but are gradually learning to incorporate it. Students of Darwin, needless to say, will also benefit as feminists help correct for their biases.

Secondly, and most importantly, let us return to the classical liberal values Damore championed. Let us not shame, let alone fire, someone who dissents from the orthodoxy of an ideological echo chamber. Instead we should tolerate such people, welcome them, even admire them. It is through debate with them that fresh truths emerge. Such debates must respect certain rules. Ad hominem attacks, for instance, are strictly forbidden: not because they’re mean, but because, from the standpoint of reason, they accomplish nothing. Each contributor should rather provide evidence, and every participant can help by evaluating its soundness.

Let us, in sum, recall that “the only way in which a human being can make some approach to knowing the whole of a subject, is by hearing what can be said about it by persons of every variety of opinion, and studying all modes in which it can be looked at by every character of mind.” For as the great defender of classical liberal values, J. S. Mill, concluded, “no wise man ever acquired his wisdom in any mode but this; nor is it in the nature of human intellect to become wise in any other manner.”

 

Filed under: Philosophy, Science / Tech

by

Patrick Lee Miller is a philosopher teaching at Duquesne University, in Pittsburgh, Pennsylvania. He is the author of Becoming God: Pure Reason in Early Greek Philosophy (Bloomsbury, 2012), and co-editor of Introductory Readings in Ancient Greek and Roman Philosophy (Hackett, 2015). His recent philosophical writing uses Platonism to address current problems such as gender, sexuality, child psychology, pedagogy, virtual-reality, spirituality, indecision, honesty, and liberal government. His faculty page can be found here: http://www.duq.edu/academics/faculty/patrick-lee-miller

28 Comments

  1. Olivebirdy says

    Can’t find your response to Zunger. Mind giving me a link?

  2. Richard D. says

    Re Objection 7.7 – the Economist (and many others) have said that the sharp change in the proportion of girls and boys studying computing is down to personal computers being marketed exclusively at boys in the 1980s.

    There are several problems with this explanation, some of which have been mentioned already, but perhaps the most glaring problem is that more girls than boys now own computers (at least in the US), and this has been true for some time, yet the gap in studying computing has remained as wide as ever. (Reference: http://www.pewinternet.org/files/old-media/Files/Reports/2013/PIP_TeensandTechnology2013.pdf, p.6. As of 2012, 82% of girls and 77% of boys own a computer)

    • Eduardo L. says

      I’m no expert, but perhaps such a small percentage (5%) is not very valuable? Also those are stats from 2012, half a decade ago, so maybe it is not so relevant now that the tech landscape has changed. But again, I am no expert and I didn’t read the paper, just voicing my thoughts.

      • Richard D. says

        Yes, there aren’t hugely many more girls owning computers than boys, but I think the point is that there aren’t hugely many more boys than girls owning computers, which is supposedly how it was in the 1980s: we are asked to believe that the decline in women taking up computer studies from the late 1980s was due to boys having the computers in the early 1980s, and girls not having the opportunity to program them.

        But now (as of 2012 or earlier) girls have that opportunity as they own (at least) as many computers as boys, so you would expect (under this ownership/opportunity hypothesis) that the proportion of women taking up computer studies to pick up and match the pre-1980s (i.e., more equal) level. Since this hasn’t happened at all, we should reject the hypothesis. Or at least treat it with extreme suspicion.

        And 2012 is a good time to predict the present uptake in computer studies now (under the childhood ownership hypothesis): boys and girls who were teenagers in 2012 will be about college intake age now. However, the proportion of men/women doing computer studies has remained roughly constant over the last 5 years, and has not returned to the pre-1980s levels.

  3. Randy says

    “This is a good argument for why design teams should have more women on them than they do”

    I don’t think so. Design teams should design to the likely users, not to themselves. I don’t think it would be very hard for a design team to locate a few likely users of a product and seek their input. In this case, to discover you should hide the “sex” field, it’s not like women are somehow foreign or unreachable. Just let them walk through your design, and they’ll hit this. You’d have a better case here if we were talking about a very different culture vs the USA, like Japan maybe. But even then, if you’re developing for Japan, does half or maybe all of your design team really need to be Japanese? Or do they just have to know how to translate requirements into user-appropriate design? What will they do when the product suddenly takes off in India, and we want to refine it so it works even better in the Indian context? Do we replace the Japanese?

    In any case, even as you describe it, this is more about making specific cultural knowledge part of the job description, rather than hiring someone just based on protected characteristics, which should be illegal no matter the purpose.

  4. Andrew POWYS says

    Thank you Patrick Lee Miller – I will be using this article as a resource for my students this semester – duly cited, of course. I hope to find a copy of ‘On Nothing’ by Gorgias – can you recommend a good translation? Andrew

  5. Just a few quick points to make at the moment.

    Regarding the reason for the wider disparity in higher-freedom regions, Artir’s criticism misses the mark in at least one important way: Even if Sadadin’s idea about men being more free to be domineering in those societies turns out incorrect, the more important point she’s making (and the more important point in general) is that it’s a wide leap to jump to any kind of certainty that the disparity reflects natural preferences.

    Regarding looking for contrary evidence, I don’t agree that it’s a scientist’s job in a peer reviewed piece of research to provide contrary evidence sufficient for what “Larry Page” is talking about here. Each research paper briefly presents some data and some conclusions, and a knowledgeable and interested consumer of “what the science says” will look at a variety of those papers for contrasting views if they exist. Every science research paper has an either explicit or understood “well, this is what our data suggests to us, but we’re publishing it as part of the larger project, beyond this paper, of finding out what all the data suggests objectively. We could well be wrong.” It’s not necessary, in the paper itself, for all such possible contrary evidence to be taken into account in detail. In particular, _of course_ a paper can’t respond to contrary evidence that comes later in the history of scientific research!

    As to asking women about their experiences, though that can do some work to informing one’s understanding of what the data suggest, much more importantly, it should inform our wisdom as to manner, place, timing and context when we decide to write out arguments about natural interest differentials etc. Since what’s at issue in “Page’s email” is not _only_ the validity of the scientific conclusions Damore argues for, but _also_ the appropriateness of his writing the memo in the first place, this (asking women about their experiences) is a fair piece of advice.

    See this essay for more on that point: https://www.vox.com/the-big-idea/2017/8/11/16130452/google-memo-women-tech-biology-sexism

    • You are a bit incorrect. A scientist’s job in a peer reviewed piece of research is not to bury contradictory findings in the body of knowledge. They are to discuss and present them all and what real scientists do is discuss the limits of the findings and how their findings address those limits with/without differing conclusion/results. Simply presenting a literature review validating your position and intentionally omitting contradictory findings meets the criteria of misconduct (something a doctoral dissertation committee would reject, for example). The problem is that peer review becomes friend review in the politicized fields of research particularly now that some mythical percentage of consensus is the goal versus empirical evidence describing the real world.

  6. > On to the sixth major flaw, therefore, which is that “even if you are right that more men than women are well-suited to the job of software engineer at Google, you are wrong that taking steps to recruit more women is inherently unfair to men.”

    Damore’s memo doesn’t say anything about it being wrong to recruit more women — the authors of the Economist editorial are arguing with their own imaginations here.

  7. anon says

    Using the word “memo” is biased against Damore because it suggests justification for his being fired by google. Legally, a memo is an official company document, not one’s personal thoughts on a topic. He didn’t write this as part of his job and submit it to googles database of memos. He didn’t email it out to management, HR, or cc to all employees or send it to the press to attack his employer.

    • Actually, he did.

      He, along with others, were specifically asked for feedback on Googles diversity policies.

      And Google provided a specific forum for people to post their feedback.

      He received no feedback on his feedback, so he reposted it to an internal skeptics forum.

      Eventually one or more people who didn’t like it sent it viral internally.

      Then one or more people who didn’t like it leaked it to the “liberal” press.

      And then, when it went viral externally, Google management finally responded, saying the MEMO had harmed the company, but they couldn’t discuss individual employees issues in public.

      But then proceeded to discuss individual employees issues in public at length and in depth.

      Also saying that people were free to discuss contentious issues.

      And then fired him for doing so (but not the people who had leaked the memo, or discussed Damore’s personnel issues with the whole world!)!

      However, what’s REALLY biased was the way so many people and publications called it a “Manifesto”, as though he were the Unabomber, or Anders Breivik!?!?!?

  8. anon says

    The economist says “You could have asked some of your female colleagues…”.
    He posted it to a google “skeptics” forum inviting feedback. Does anyone believe he only wanted men to respond? Obviously some just wanted him fired and had no interest in a rational discussion.

    • anon says

      Also, anecdotes from female coworkers is of less value than scientific research.

    • Ironically enough, according to the Dave Rubin interview, getting feedback from female coworkers on his team was the first thing he did.

  9. Great article. This memo should not be lost and forgotten under the cover of identity politics.

  10. George Millo says

    This is the kind of article that I will send to people when I want them to understand why Quillette is such an excellent website. Thanks!

  11. Whatever about the Science, what about the simple business and HR aspects of this case? Here was a guy who is supposed to be writing code who has the time to be sending an incendiary ‘memo’ around the office based on his interpretation of an area in which he has no academic or scientific expertise.

    Google is a private company, they make money mainly by selling advertising. They are not a university or seat of learning. The memo was damaging and they had every right to let him go, simply for not doing his job and messing with the companies reputation.

    To make matters worse, right after being fired he went straight onto right-wing youtube channels for interviews. If he had no agenda and was simply interested in good science he would have spoken to real scientists in the field about how their work was being mis-represented or ignored. Instead he joined the culture wars.

    • I hope you didn’t write that on a company laptop on company time!

      Firstly, as I understand it, Google employees are supposed to spend a proportion of their time on personal projects anyway.

      And he wasn’t sending an incendiary, or any, ‘memo’ around the office: that was done by trouble makers who didn’t agree with it.

      As for it being based on his interpretation of an area in which he has no academic or scientific expertise, he has a degree in Biology, and a Harvard Masters degree in Systems Biology, which looks into, among other things, exactly this area, and he had been on a (was headhunted from? – it also involves a lot of bespoke high level programming) a Harvard Systems Biology Doctoral Programme.

      However, Google had specifically asked employees for feedback on its diversity policies.

      And Google provided a diversity policy feedback forum for people to post their feedback on.

      Which is where he initially posted his “Memo” AS REQUESTED/DIRECTED BY HIS EMPLOYERS!!!

      As he received no feedback on his feedback he reposted it to an internal skeptics forum for discussion.

      Eventually one or more people who didn’t like it sent it viral internally.

      Then one or more people who didn’t like it leaked it to the “liberal” press.

      And then, when it went viral externally, Google management finally responded, saying the MEMO (not the people who had leaked it to the press) had harmed the company, but they couldn’t discuss individual employees issues in public.

      But then proceeded to discuss this individual employee’s issues in public at length and in depth.

      Also saying that people were free to discuss contentious issues.

      And then fired him for doing so (but not the people who had leaked the memo, or discussed Damore’s personnel issues with the whole world!)!

      Oh, and Google don’t just make money by selling advertising, they make money by selling information on their users, on things like their tastes, preferences, interests, and the like, including on a group, or gender basis.

      Funny that, as they insist we are all the same and such differences are social constructs?!

      And why wouldn’t he go to the “right wing”?

      He’s initially sent his memo to a left wing company.

      It was leaked to the left wing press, who ripped him apart.

      And he did initially speak to a “liberal” news outlet, where the interviewer kept misrepresenting the memo, and attacking him.

      So where else would he (eventually) go?

      Oh, he also did a long interview with a psychology professor!

      And he’s spent years discussing this on his degree, masters, and doctoral programme.

      I’ll leave you to count up how many things you got completely, totally, and utterly wrong!

    • Mr. Mann put it nicely. Don shows the same suttering malice and devotion to dogma so nicely disposed of by the author.

  12. Pingback: The full memo: James Damore – Googles Ideological Echo Chamber | How blue

  13. Phillip says

    Really solid article about this.

    It’s rather troubling that in our society we even need to have people like you write this, but to me what’s even scarier is we in the west aren’t allowed to know men and women are different until We have scientists tell us. To see the absolute hysteria over this simple statement means things will get much much worse and everyone who can’t accept reality will be in for some serious pain when it’s done. It seems the smarter the person is their capacity for pure idiocy raises proportionally, or as Orwell said some things are so dumb it takes an intellectual to believe them.

  14. ccscientist says

    Damore’s point that if more men are in computer science in college there will be fewer women to recruit from is simple logic and it seems crazy to “deny” it.
    The point that the most extremes in ability exist for men (most very smart, most dumb) is beyond question but of course got Larry Summers fired as pres of Harvard. Google no doubt wants the extremes–the absolute smartest math people. This makes the recruiting pool even smaller for women and means that affirmative action will dumb down the hiring pool.
    Women in general do not find technical puzzles that interesting. I know many many guys that build their own stuff (computers, model railroads, real airplanes, houses) but NO women who do such things. In conversation, men will debate all the ins and outs of technical topics–gas mileage, how to do taxes, mac vs pc, antivirus brands, sump pumps, home remodeling, tails of statistical distributions, sports stats, anything within their group interest. Women almost never do this. I have known precisely 2 women engineers who could hold their own talking about technical stuff, and then largely only within their specialty. Men read technical books and the WSJ–women do not. Look around on an airplane at what people are reading. This is all about INTERESTS. The critics of the memo are either denying what women like in the face of data or deny that interests matter.
    Men are quite competitive inter-personally but don’t take it personally (not talking about back-stabbing at work). They can pretend to be about to push a guy off his stool at a party, lord it over the other guys that they just went camping or just got a promotion, and then the other guys will push back. But it is all good fun rough-housing, like kids wrestling and punching each other on the shoulder. The implication of lots of the claims about discrimination are that women can’t take it, they are too fragile, they need hand-holding, kindness, loving protection. That is, they are invoking sex differences!!! It may be in fact that women do need more kindness but that is not the same as saying men are evil. The world of business is pretty ruthless, not a love-fest.
    Why do women not end up in mgt positions? They may have other priorities like being able to be home with the kids and having friends. Is this bad? No, but it is not the fault of men. To say that women should be able to be a top manager and only work 35 hrs/wk when men in the same position work 60 hrs/wk is pretty blatant favoritism.

  15. “good time to reason through the best arguments on each side of the controversy. Who was right? What can we learn? How can we do better next time there appears to be a clash between the competing values of equality, science, and freedom of speech?”

    Q. Who was right? A. Nobody.
    Q. What can we learn? A. Social science is pretty much useless.
    Q. How can we do better next time there appears to be a clash between competing values of equality, science and freedom of speech? A. Admit the science is useless.

    Let’s take this out of the context of STEM and gender. Let’s take some other popular topics: Porn, video games and the death penalty.

    Take the death penalty. Does it deter crime?
    Take porn. Is porn addiction a common problem? Does porn addiction destroy marriages and cause men to lose jobs?
    Take violence in video games. Does video game violence lead to actual violence?

    The answer my friend is blowing in the wind. Call it the butterfly effect. Culture is the Earth’s atmosphere and biological variables such as the death penalty, porn and video games are the butterflies. Yes, we know the butterflies affect the weather. No, we don’t know any more than that. With respect to all the social science regarding gender, porn, video games and the death penalty it is nigh impossible to isolate the variables. Just like with the butterfly. What this means is that the science is moot. Everybody is just making after-the-fact observations of no future predicabilty and a random choice has the same chance as an informed one. Claim all you want the death penalty reduces violence. Good luck isolating it. Claim all you want that video games promote violence, good look isolating it.

    And this is the real evil in any of these op/eds. There is confidence of utility of model that doesn’t exist. Any time you see an op/ed like this or Damore’s for that matter ask them to put their money where there mouth is: start a business based on the confidence of the scientific models. If they don’t have that confidence to do that then they must admit the science has no modeling predictability capability.

Comments are closed.