I've heard his linguistics work is genuinely remarkable, and as someone with a passing interest in the topic I'd be curious to check it out.
His linguistics work is an awful scam which set the field back decades. Advances in the study of poorly-recorded indigenous languages (particularly those in South America) and psychology have both shown his understanding of universal grammar to be horrifically wrong. By contrast, the work others were doing in linguistics at the same time -- particularly Joseph Greenberg's use of mass comparison -- has been hugely vindicated by modern archaeogenetics; more attention given to them would've given us a better understanding of prehistory at an earlier date.
Also, genocide apologist and anti-capitalist. HP.
I'm not a syntactician ftr, and I'm not particularly invested in Universal Grammar, but the evidence on Pirahã is quite contested and depends a lot on how you interpret Everett's own documentation (and how reliable you think Everett's work is.) The arguments in favor of the presence of recursion in Pirahã don't seem crazy to me. I've also never really come across a good argument against the Poverty of the Stimulus argument, especially one that would work cross-culturally in contexts where children receive less linguistic stimulus.
On the other hand, Greenberg's work is really incredibly unscientific. It's not that his work is necessarily getting at the wrong things -- like you said, a lot of it corresponds to archaeology and genetics -- but fundamentally it just isn't rigorous. Mass comparison is really not reliable -- there's no way of getting around scientific comparative reconstruction. Part of it frankly, is an understandable but inaccurate desire to fill in blanks in the historical record, especially since comparative reconstruction gets significantly less accurate the further back you go (there's a reason why we don't have a good reconstruction of Proto-Afroasiatic, even though it's pretty uncontroversial).
This gets to be pretty problematic though, because there's a lot of Greenberg's classifications just don't meet the threshold for actual language families, ranging from "seems pretty likely (Na-Dene)" to "maybe but it's hard to say (Papuan)" to "false (like for Nilo-Saharan or Khoisan) to "laughably non-empirical (Indo-Pacific, Eurasiatic)." It's very tempting when suggestive linguistic similarities seem to match up with genetics and archaeology, but you can't just reject basic historical linguistics to read those similarities into the discipline.
Poll
