Friday, August 11, 2017

You can't say that! It might be true!

By now you have probably heard about James Damore, the Google engineer who was fired for writing a memo about... well, that's where the trouble begins, because the memo was about two things.  But the media kerfuffle focuses on only one of them.

One of the two things the memo was about was the hypothesis that women might be less suited to careers in technology at least in part for biological rather than social reasons.  That is, understandably, the part that the media has focused on, and the part that led to Damore being dismissed.  But the second thing the memo was about, what was in fact its main thesis, was that, at least at Google, you cannot even advance the hypothesis that biology might be a contributing factor to women's underrepresentation in tech without putting your career at risk.  Ironically, by firing Damore, Google proved that he was actually right about that.

Before I go on, because this situation is absolutely brimming over with opportunities for misunderstandings, I want to say up front that I do not agree with Damore's hypothesis.  The evidence for it seems thin to me, and the best data indicates that there are few discernible differences in mental capacity between men and women.  I am not defending Damore's thesis.  I am defending his right to advance it without putting his livelihood at risk.  [EDIT: I used the word "right" too glibly here.  Employees generally do not have free-speech rights on the job.  But Google claims to encourage free speech and dissent.]    And I am going to go one step further and advance a controversial thesis of my own, namely, that one of the reasons that this is such a hot-button issue that that Damore's thesis is plausible.  It could be true.  Biology clearly can have an impact on cognitive ability.  Down Syndrome, for example, is a biological trait (caused by having an extra chromosome) that causes "mild to moderate intellectual disability".

I can hear the whoops and hollers already: how dare you compare being female with having Down syndrome!  Well, if you read carefully, I am not comparing those two things.  I am citing Down syndrome as evidence that biology can have an impact on cognitive ability, and hence it is not impossible a priori that having a Y-chromosome deficiency might have a similar impact.  However (and this is very important) it is just as plausible a priori that this difference could hew in favor of women as against them.  A "Y-chromosome deficiency" might cause cognitive impairment, but so could "testosterone poisoning".  (Note that I'm deliberately choosing ironic anti-euphemisms here to highlight the point that people respond to these arguments emotionally rather than intellectually.)

I can totally understand the desire to shut down this discussion.  I'm a liberal.  I want a world with equality of opportunity.  If (strong emphasis on IF) it turned out that being female did indeed have a measurable impact on one's ability to do certain tasks it would make the battle that much more difficult.  There are obvious differences in physical abilities between men and women (sports are uncontroversially segregated by gender), which has made it that much harder for women to secure the right to, say, serve in combat roles in the military, even roles which are much more intellectual than physical, like being a fighter pilot.

But suppressing opposing views is a very dangerous game, and not just because it can blind you to the truth.  It's dangerous because it undermines the very goal that it seeks to advance, namely, social justice.  By firing Damore, Google reinforces the belief held by many conservatives that liberals value social justice more than they value the truth.  In fact, the narrative goes, liberals fear the truth and must suppress it because social justice is not part of the natural order of things.

This point of view is directly supported by Google's CEO Sundar Pichai's response to Damore:
... we strongly support the right of Googlers to express themselves, and much of what was in that memo is fair to debate, regardless of whether a vast majority of Googlers disagree with it. However, portions of the memo violate our Code of Conduct and cross the line by advancing harmful gender stereotypes in our workplace. ... To suggest a group of our colleagues have traits that make them less biologically suited to that work is offensive and not OK.
At the same time, there are co-workers who are questioning whether they can safely express their views in the workplace (especially those with a minority viewpoint). They too feel under threat, and that is also not OK. People must feel free to express dissent. So to be clear again, many points raised in the memo—such as the portions criticizing Google’s trainings, questioning the role of ideology in the workplace, and debating whether programs for women and underserved groups are sufficiently open to all—are important topics. The author had a right to express their views on those topics—we encourage an environment in which people can do this and it remains our policy to not take action against anyone for prompting these discussions.
It is important to note that Pichai makes no attempt to actually debunk Damore's claims about women's biology.  He cites no sources.  He doesn't even bother to explicitly state that he thinks Damore is wrong!  Instead, he just says that some things, like "advancing harmful gender stereotypes", are "not OK."  Even if they are true.  And notwithstanding Pichai's nominal support of free speech, the elephant in the room is that Damore no longer works at Google.

The net effect of this is the exact opposite of what we should be striving for.  Suppressing dissent does not make it go away, it merely drives it underground, where it festers and grows and eventually re-emerges, usually (but not always) taking liberals by surprise.

Ignorance and prejudice cannot be fought with censorship.  It simply doesn't work.

15 comments:

O. said...

Solid points. The problem with advancing the theory because it might be true is that it immediately affects those it might be true about. Scientific discourse is slow and methodical. The layperson discussion and its "innuendo effect" in less rigorous circles moves more quickly, and the amount of effort required to dislodge beliefs that could have been true but have no empirical proof is large.

"African Americans are less intelligent/more violent than *."
"Native Americans are prone to alcoholism."
"Jews are ..."
etc.

These are examples of beliefs or patterns that may have at one point fit in the "it might be true" category, but whether or not they've been debunked, they are undead beliefs still held by people today.

There's a gap between allowing free and open discussion and disavowing distasteful ideas and the people who hold them that I don't know how to reconcile. But I also don't know how to tell women, black people, native Americans, etc that they have to pay whatever price these beliefs have while society discusses.

Ron said...

> "Native Americans are prone to alcoholism."

Well, that's a fact.

https://en.wikipedia.org/wiki/Alcohol_and_Native_Americans

Your other two examples are less clear cut. You didn't finish your "Jews are..." so I'll do it for you:

"Jews are disproportionately likely to win Nobel prizes."

That, too, is a fact (https://en.wikipedia.org/wiki/List_of_Jewish_Nobel_laureates).

Your third example is particularly problematic because intelligence is really hard to define, let alone measure. But it's a fact that blacks score lower on standardized tests than whites.

https://www.brookings.edu/research/race-gaps-in-sat-scores-highlight-inequality-and-hinder-upward-mobility/

It is also a fact that in the U.S. blacks are disproportionately likely to be imprisoned for violent offenses:

https://en.wikipedia.org/wiki/Race_and_crime_in_the_United_States

"There is general agreement in the literature that blacks are more likely to commit violent crimes than are whites in the United States."

The problem is not the statement of these facts. The problem is that some people take these facts, draw unjustified conclusions about causality, and then try to use those to justify discrimination. For example, some people reason that because blacks are more likely to commit violent crimes than whites, that justifies discrimination against blacks. No. It is entirely possible that the discrimination that results from this faulty reason is in fact the *cause* of the problem. In fact, I think it's pretty likely. I don't see any possible biological mechanism that would cause skin color to be correlated with a predisposition to violence. On the other hand, I see a lot of plausible mechanisms by which being treated as a second-class citizen all your life for no good reason could lead someone to be more violent.

*That* is the kind of discussion we should be having IMHO. But we can't because everyone is walking on eggshells. And that makes the problem worse because driving the faulty reasoning underground doesn't make it go away.

O. said...

"The problem is that some people take these facts, draw unjustified conclusions about causality, and then try to use those to justify discrimination."

This is the price I referred to. It's sort of a corollary to the bullshit asymmetry principle. Until we resolve how to keep the discussion we should be having from resulting in those unjustified conclusions and the chain of actions that follows, should we just allow those unjustified conclusions, and let the consequences fall as they may? These conversations are not always started in the spirit of discussion; sometimes they are started under the guise of "I'm just stating the facts" or "I'm just asking", but are meant to bolster the unjustified conclusions.

You point out statistical realities that I cannot possibly disagree with, and I do not.

But the difference between quoting the statistics and the unjustified conclusions (and the discrimination that can follow) is all too often lost on people, who say, for example "well, if black people are more likely to commit violent crimes, why on earth shouldn't we treat them differently? I can't address the causes myself, I can only keep myself and my family safe." A host of other actions can follow from what may feel like an eminently reasonable position.

I'm not suggesting that we shouldn't allow the statistics to be offered, and to the extent real discussions about these things exist and are useful, I'm all for them.

I'm asking, without asserting an answer (but perhaps rejecting the notion that all conversation is good conversation), what should we say or do when people offer those unjustified conclusions? Should we patiently and endlessly refute them, and relitigate them every time they come up?

This specific memo is just a recent well-known example of ostensible entreaties to honest discussion, and it's not really treading substantially new ground with its questions or conclusions. It seems people have been asking whether women are suited to STEM occupations since shortly after computers stopped being women.

I don't have the answer, but I do know that the work to refute such conclusions, and the consequences resulting from those conclusions land heaviest on the people who are the target of them. Women pay the price for the Google memo, regardless of whether any of its central conclusions are accurate or justified, or whether any of the conclusions are ultimately debunked.

O. said...

Apologies for anything that sounds accusatory in my last comment (or the previous one). I sincerely did not mean to imply anything about how you feel about women or any group, or the consequences, and I don't believe you support/believe Damore's conclusions or any of the other less palatable stereotypes we've referred to so far ... not that you wouldn't be/aren't free to do so.

Ron said...

> Should we patiently and endlessly refute them, and relitigate them every time they come up?

Yes. That's the only way to win.

> if black people are more likely to commit violent crimes, why on earth shouldn't we treat them differently?

Because treating them differently becomes a self-fulfilling prophecy. Self-fulfilling prophecies are actually *true* if enough people believe them (that's what makes them self-fulfilling). So the only way to stop their deleterious effects is to stop believing them.

BTW, note that this reasoning applies *even if the underlying prejudice is actually true* (for example, because it is the result of a self-fulfilling prophecy) because the predictive power of even justified prejudice is low while its destructive capacity is high. Marginalizing a segment of the population on the basis of things over which they have no control is extremely harmful to society (to say nothing of the negative impact it has on those being marginalized). And note too that your beliefs are not something you have control over. You can't choose your beliefs, only how you act on them.

> Apologies for anything that sounds accusatory in my last comment

No apologies necessary. It thought everything you said was eminently reasonable.

TheAccomodator said...

I would say the price here was paid exclusively by Damore.

O. said...

"Because treating them differently becomes a self-fulfilling prophecy."

This doesn't present any incentive whatsoever for a person who believes any of what we might term less palatable/less rational beliefs to stop believing them. Put simply, if I already think 'screw them', the fulfilling of that prophecy makes no difference to me.

"So the only way to stop their deleterious effects is to stop believing them."

This makes responsibility for these concepts very diffuse (as in, it's on every individual person to reach enlightenment and stop believing on their own, and there's nothing a society can or should do to stop the reinforcement these beliefs before they have consequences).

"Marginalizing a segment of the population on the basis of things over which they have no control is extremely harmful to society (to say nothing of the negative impact it has on those being marginalized)."

But the people who do the marginalizing are likely to be unaware, or at the least unaffected, by the negative impact, so the incentive to stop marginalizing is minimal, if not zero, if any of said people are already inclined to marginalize them. This presupposes an inclination on the part of people to be sufficiently concerned about the effects of their beliefs on others who they may have no connection to.

I don't think you've addressed the notion of the consequences of these "unjustified conclusions" on the subjects of the conclusions, the "price" as I've called it. If you're not averse, I remain curious what you would say to or about those subjects about those consequences. As I've taken enough of your time, I'll take my answer if any "on the air", as they say in radio. I hope I managed not to come off as accusatory in this comment either. Thank you for engaging!

Will said...

Thank you for posting this. For all of the drama, I've actually been pleased by how level-headed many liberals/progressives have been over this whole ordeal.

But I think it's important to note that Sundar Pichai's interpretation of Damore's claim is arguably incorrect.

Damore did not say that his female colleagues were less than capable because of their gender. He did speak of population differences between men and women in terms of both interest and ability (that has an arguably biological basis[0]) but also qualified that individuals cannot be reduced to their population. If true we can infer that a) the women at Google might be exceptional for the gender[1], and b) that achieving a 50% representation between men and women among Google leaders may not be feasible. It does not mean that the women at Google are less capable, and Damore never said that it did.

Damore did claim that there is arguably a double standard for women (and other minorities) at Google. If true, that would call into question the abilities of Google's female employees. However, the question is not based on their sex/gender, but based on Google's hiring practices.

I've known a number of minority individuals (women, people of color, etc.) who avoid places that engage in affirmative action hiring precisely because it hampers their ability to signal competence. Certainly it makes an awkward situation for female employees that James made the accusation. Not all such accusations are made in good faith (though I believe his was). But if it's true, then the women (and other minorities) of Google should be mad at Google, not James. A double standard for hiring for any group only creates a "Lemon Market"[2] for that group, and drives down their wages.

Note that I believe it is possible to do more to recruit minority candidates while still having an equal standard [3]. I think both James and I would agree that using such tactics would be a good thing.

[0] http://slatestarcodex.com/2017/08/07/contra-grant-on-exaggerated-differences/
[1] Well, more exceptional. It's fair to say that most engineers and leaders at Google are exceptional regardless of background.
[2] https://en.wikipedia.org/wiki/The_Market_for_Lemons
[3] https://martinfowler.com/bliki/DiversityMediocrityIllusion.html

Richard said...

Most of Damore's arguments were not about women's mental capacity, just that because of biology they may be predisposed to certain types of roles. He also stressed that he was not saying that these make women deficient in any way, but just that there may be ways to change the way things are structured to be more inviting to them.

Ron said...

@O.:

> if I already think 'screw them', the fulfilling of that prophecy makes no difference to me.

If it really makes *no difference* to you, then you are either completely unreasonable, or you have not taken on board what it means for a prophecy to be self-fulfilling.

In the antebellum South it was illegal to teach slaves to read. As a result, very few slaves knew how to read. Many white people then justified slavery by pointing out that slaves didn't know how to read and so were not suited to any role in society other than being slaves.

Do you really need my help to spot the flaw in that reasoning?

> Thank you for engaging!

My pleasure.

@Will:

> Damore did not say that his female colleagues were less than capable because of their gender.

That's not true. He wrote:

"[W]omen generally hav[e] a harder time ... leading"

"Women, on average, have more ... neuroticism (higher anxiety, lower stress tolerance) [than men]."

To which I would point out that women might well *appear* to have a lower stress tolerance than men because they have to deal with more stress from things like, say, men claiming that they have lower stress tolerance and using that as a justification to discriminate against them (see my points above about self-fulfilling prophecies).

This is exactly the problem: these statements are so toxic that people either focus on them as if they are the most important things that Damore said, or they ignore them and pretend he never said them. *No one* is actually taking this head on and asking: could these things be true? How can we know? And how does that impact the validity of the rest of his argument?

Will said...

> That's not true. He wrote:
>
>"[W]omen generally hav[e] a harder time ... leading"
>
>"Women, on average, have more ... neuroticism (higher anxiety, lower stress tolerance) [than men]."

Hm... perhaps I am missing something. He did indeed make claims about the population differences in terms of ability, and I acknowledged them in my comment. But let us consider Pichai's statement:

"To suggest a group of our colleagues have traits that make them less biologically suited to that work is offensive and not OK"

As I read this, it seems to me that Pichai is accusing Damore of making a gender reductionist claim about the women are already working at Google - that they specifically are inferior at doing the work they are already doing. As far as I can tell, Damore actually goes out of his way to make it clear that he is making no such claim. Talking about population differences is not the same as making claims about specific individuals, do you disagree?

Richard said...

I took Damore's main point to be... "Company X employs less women than the general population" does not mean that Company X is discriminating in any way. It could just be that women (we're talking statistically here, think trends, not identities) don't like those roles. Take child care roles as an example in the other direction. While I don't disagree that there is likely a huge number of males who would like to or already do work in the profession, I would be very surprised if the percentage of males in that field isn't extremely low. I don't think this is because the child care industry is actively preventing males from working in it (but again, there is some!) but rather that guys just don't want to do that work.

Ron said...

@Will:

I interpolated Damore's remarks as follows:

1. Women in the aggregate have trait X (what he actually said)

2. The women who work at Google are women (not what he actually said, but a not-unreasonable assumption), ergo...

3. The women who work at Google likely have trait X in the aggregate.

Luke said...

@Ron:

> 1. Women in the aggregate have trait X (what he actually said)
> 2. The women who work at Google are women (not what he actually said, but a not-unreasonable assumption), ergo...
> 3. The women who work at Google likely have trait X in the aggregate.

This seems rather suspect until the various selection biases are ironed out. Here's why:

1.′ Men in the aggregate are not good at programming.
2.′ The men who work at Google are men.
3.′ The men who work at Google are likely bad at programming in the aggregate.

The above is obviously erroneous, excepting those nitpickers who would point out that perhaps fewer than half of Google's male staff are programmers.

Publius said...

Testosterone

No, the Google manifesto isn’t sexist or anti-diversity. It’s science
As mentioned in the memo, gendered interests are predicted by exposure to prenatal testosterone – higher levels are associated with a preference for mechanically interesting things and occupations in adulthood. Lower levels are associated with a preference for people-oriented activities and occupations. This is why STEM (science, technology, engineering and mathematics) fields tend to be dominated by men.

We see evidence for this in girls with a genetic condition called congenital adrenal hyperplasia, who are exposed to unusually high levels of testosterone in the womb. When they are born, these girls prefer male-typical, wheeled toys, such as trucks, even if their parents offer more positive feedback when they play with female-typical toys, such as dolls. Similarly, men who are interested in female-typical activities were likely exposed to lower levels of testosterone.


After I started taking testosterone, I became interested in science...

Testosterone, Social Class, and Antisocial Behavior in a Sample of 4,462 Men