So we all agree Satoshi Kanazawa is a bad scientist; what next?

Yesterday morning, Psychology Today published a blog post by Satoshi Kanazawa titled Why are Black Women Less Physically Attractive than Other Women. The link has gone dead, but if you’re curious, somebody mirrored the article here. News and reactions here and here and here and here (and plenty elsewhere). It has been causing a bit of a kerfuffle online. Plenty of people are duly taking Kanazawa to task over what he wrote. I’d like to reflect on some related issues.

1. Psychology Today needs to answer for itself. Psychology Today apparently pulled the article, but so far they have offered no explanation. Marianne Kirby says that’s not enough, and I agree. With no explanation, Kanazawa or his supporters can position him as the noble truth-teller being censored in the name of political correctness. Psychology Today needs to head off that argument by directly refuting the substance of what Kanazawa wrote, not just disappearing his blog post. They need to show that their decision to spike the article was an evaluation of the science. And the statement needs to come from the editorial staff – it isn’t enough just to let this be a back-and-forth with other Psych Today bloggers.

2. Let’s stop bothering to read anything that Satoshi Kanazawa writes. A few years ago statistician Andrew Gelman spent some of his valuable time writing a critique of another of Kanazawa’s claims. Gelman took Kanazawa seriously and was evenhanded, but concluded that Kanazawa had commited some serious statistical errors. Kanazawa (or his partisans) placed a rebuttal on Wikipedia, and went on to write a popular book about the disputed research. The lesson I took from that incident was that hunting down all of Kanazawa’s errors is a thankless job. Based on his track record it is probably safer just to assume that he’s always wrong and move on.

3. Don’t blame Add Health. I’ve seen some bloggers attacking Add Health, the longitudinal study of adolescent and adult health from which Kanazawa got his data. That’s misguided. Kanazawa had nothing to do with planning or running Add Health. Add Health is a publicly funded study that makes some of its data available to the general public and other, more sensitive data available to researchers who enter into security arrangements. They provide a valuable resource and should not be held responsible for how their data gets used or misused.

4. You don’t need cultural determinism to refute him. Physical attractiveness is a reputational construct, meaning it is irreducibly defined by how people perceive one another. And plenty of studies have shown that judgments of physical beauty vary by culture, historical era, by who is the perceiver and their relationship to the target, etc. But there is a world of difference between saying “culture matters” and saying “culture is the only thing that matters.” The former is indisputably true, and that gives you everything you need to indict Kanazawa. By not considering cultural explanations or perceiver-side biases, Kanazawa committed an enormous error in reaching his conclusions. You don’t need to go the next step and claim that there are absolutely no universals in how humans make judgments of attractiveness. For one thing, that kind of blank-slate cultural determinism is much more difficult to defend; for another, no serious universalist theories say anything about race differences. Don’t do Kanazawa the favor of including him in a sophisticated scientific discussion about the bases of attractiveness judgments; his mistakes are far dumber than that.