I don't think Sapir-Whorf (in a weak form) is just a case of inverted causality. Sure, matters of perception shape the language we speak; this happens all the t

I don't think Sapir-Whorf (in a weak form) is just a case of inverted causality. Sure, matters of perception shape the language we speak; this happens all the time. But the key point is that the reverse direction happens too. Language is a perfect medium to transport preconceived notions and because we don't think so much about it (indeed, we think lots of things are natural, when they're actually particular features of a single language) it's particularly subtle. Examples of this include gender/classifier systems, colour perception, evidentiality markers, or one of my favourites, spacial recognition: one famous experiment shows that members of a particular Australian aboriginal tribe (Guugu Yimitthir) think in terms of absolute cardinal direction even in everyday situations instead of left and right; in fact they don't even have words for left and right. Of course, the reason for this development is originally due to the landscape around them, but now the language which lacks such terms serves to reinforce this worldview. My only qualm with Sapir-Whorf is that it only credits language with this cultural transmission. But there are a lot of other things that can do this (arts, rituals, customs), even if language is a big part of it. Well that and that the original version ("the Hopi don't know the concept of time") is a bit ridiculous.