Thread by @BretDevereaux

So this critique (foreignpolicy.com/2021/06/20/his…) in @ForeignPolicy on failures of the sort of data-driven pseudo-history we've been seeing a lot of lately is pretty spot on.

It's focused specifically on Joseph Henrich's recent 'WEIRD' book, but the critique is more broadly useful. 1/9

In some ways, the article's focus on Henrich is actually unfortunate (though this is one of those 'this isn't the article I'd have written things, so grain of salt at the ready) because it leads it into a bit of a rabbit hole about 'the West' particular to that work... 2/9

...whereas to me the more direct issue here - if I may indulge in a strained analogy - is not the particular ugly face mounted on the top of the data-driven pseudo-history statue, but the clay feet at the bottom. 3/9

The cliometrics data-driven psueo-science methods can be used to produce any conclusions, from conclusions I like to conclusions I dislike. The conclusions aren't the problem, the epistemological bankruptcy of the method is the problem. 4/9

A lot of these studies fall to some really basic first-year-of-grad-school kind of errors! I am consistency befuddled as to how anyone gets very advanced in some of these fields without learning the garbage-in, garbage-out principle, and yet here we are. Again.

5/9

Woefully incomplete datasets, problems in framing assumptions, lack of comparative basis, and an apparent ignorance of the nuances of the evidence and what it can tell us versus what it can't. 6/9

Honestly, in my 100-level survey course, I have assignments built around moving students to these sorts of realizations, "oh, this source doesn't give me evidence for what people thought, only for what *this* person said" kind of stuff.

And yet, here we are. Again. 7/9

And I do mean *again* - Fafinski here, with a real economy of words, does a great job of setting out this tradition of 'scientific' pseudo-history - from Spengler and Huntington to Turchin, Henrich and Safra. He could have easily added Arthur de Gobineau (d 1882). 8/9

Anyway, it's a good article. The key thing is to remember that no amount of fancy math can raise the reliability of a conclusion any higher than the reliability of the evidence it is based on.

A statue with feet of clay is made no stronger by casting the torso in steel. end/9