Scott Alexander

Agreed, and I've told deBoer the same. But I don't demand someone be right where everyone else is wrong before generally liking them.

its a disservice to readers like me that look to u for solid recommends in the information deluge just recommend Haier to NEone that asks

Who's Haier?

Also, consider broader implication of supporting principle "never trust any work by person who disagrees with you about race/IQ".

This is basically the equivalent of no-platforming Charles Murray and ignoring his working on eg rural white working class.

I think it's important to protect ability to recommend someone you disagree with about other things

And I think this issue is especially important if you have unpopular ideas and expect to mostly be on the receiving end of shunning.

What does that mean?

http://rationalwiki.org/wiki/Gish_Gallop … Which is definitely Vox's strategy

Ah gotcha. Have we published a lot of pieces about Charles Murray? I think that was the only one.

There's also the one Garrett wrote about at https://twitter.com/garettjones/status/846923004881731584 …

Plus various less focused ones, tweets, and stuff by other sources.

I, personally, have definitely tweeted quite a few times that I find it bizarre that Charles Murray is considered a reputable figure.

Your own institution's article basically admits he was right about four out of five things, although it does its best to downplay it.

How does this relate to Noah's rule?

I think you can cite an analogue of Noah's point: if you've investigated and dismissed 2 representative articles, assume it's all garbage.

But many excellent articles making IQ-partly-nonhereditary point. Neither obvious Murray's right, nor does Murray say 100% is genetic.

If someone gave me the two best articles arguing for nonhereditary IQ differences, I would expect them to be excellent.

That's not the hypothesis being discussed, though. To refute the claim of a hereditary IQ differences, you'd need to, well, refute it.

I mean, you can never prove a negative. I think "IQ differences are often nonhereditary, here's some reasons these might be" is enough.

You should know this! Bayesian evidence against the hypothesis is easy - if the hypothesis is specified clearly. Hereditary>0 isn't a claim