Intelligence, logic, statistics, and one's own petard


Categories: Personal Politics

A while back, Eric Raymond blogged what he called an intelligence test. It’s based on the much-commented on “the talk: non-black edition” piece that was making the rounds a while back. Raymond continues with some questions, then says:

If you answered “Yes” to any of the above questions, you failed the test.

There are flaws here. First, this is a test of logic, not intelligence. There are smart people who have never learned precise logic, and really, this test is aimed at a particular kind of rationality. However, there is a much deeper flaw. Let’s look more closely at Eric’s questions #1 and #4:

1. Did you fail to notice that the key paragraph in it is this one: “Thus, while always attentive to the particular qualities of individuals, on the many occasions where you have nothing to guide you but knowledge of those mean differences, use statistical common sense:”


4. Did you finish the article believing that John Derbyshire (the author) is a racist?

There are multiple flaws. First, you could have some reason other than the article for believing Derbyshire to be a racist. But that’s not the big one. The big one is the paragraph quoted in question #1. As Derbyshire and Raymond both agree, when you do not have enough information about the particular qualities of individuals, you should use statistical common sense.

Well. It turns out that, while there are doubtless exceptions, there are a whole lot of people out there who will justify racist courses of action by appeal to various statistics. So many that, in fact, if you do not have significant information to the contrary, it is a statistically sound and reasonable assumption that any such person is probably a racist. And question 4’s phrasing says “believing”, not “thinking you have irrefutable proof”.

So it seems to me that the very argument Derbyshire advances for treating people based on statistical norms in the absence of more specific data fully justifies people treating him as a racist until they have more information available.

There is a more subtle question to be had:

3. Did you at any point refuse to believe a fact claim in the article because you think the world would be a worse or uglier place if the claim were true?

This oversimplifies a little; contrast with refusing to accept a fact claim because you think the world would be a worse or uglier place if you believed it, and the fact claim is about a thing which is true or false in part because of human beliefs. Interpersonal relationships are heavily influenced and shaped by peoples’ beliefs. If you go around expecting people to be liars and cheaters, more people will lie to you and cheat you. People pick up on things and tend to act according to expectations. This is why good managers are effective; they expect the best and people know it. So choosing to adjust your beliefs in order to make a better world is not necessarily stupid or irrational.

I personally don’t know whether Derbyshire is a racist, but I do know that people who say things like that often are, and that thinking that way about people creates the sorts of circumstances under which the brain’s overactive pattern-recognition tends to create racism through confirmation bias. In the long run, if everyone non-black acted on his advice, the world would be a much worse place, and his descriptions would become more true; if no one acted on this advice (even the people who already are even if they haven’t seen it), the world would be a much better place, and his descriptions would become less true.

I choose to, in general, disregard statistical claims about populations because I am aware that the human brain is simply not well-equipped to process them in a way that doesn’t introduce more errors than it corrects. (For more background on how utterly shoddy the human brain’s heuristics can be, see Kahneman’s Thinking, Fast and Slow.)

In short, while I do not know whether Derbyshire was a racist, I know that humans who internalize models like that are virtually guaranteed to become racists over time through the magic of confirmation bias. The disclaimer paragraph is not the piece’s salvation, because it is irrelevant to actual humans, who simply cannot manage that kind of intellectual rigor consistently.