Hmmmm, well, I guess Fishman's "double dichotomy" of the linguistic relativity principle (between lexicosemantic and syntactic levels and between effects on linguistic and non-linguistic cognitive processes) is fine as far as it goes, and I've come to the first split in my own work on the eighteenth century, between relativity as conceived by Herder and by Humboldt. But it's hardly the principle as a whole that's being systematized: what about the dichotomy between "influence" and "determination" conceptions of relativism? The one between "universal relativism," where thought is relative to language full stop, and "relative relativism," where that relativity expresses itself differently relative to which specific language is meant (the real point of the thing, as far as I'm concerned).

The idea of a Whorfian "hypothesis" is also one that Fishman would reject later on, as it became clear that Whorf intended no such thing. There is also a summary of experimental work on the "hypothesis" to date, but since it was 1960 that has been superseded by more recent treatments. That's not to say there aren't also some neat connections made by this expansive and brilliant scholar: with Malinowski's "phatic communion," basically grouped unity through typed, idiosyncratic linguistic practice (say a penchant for certain kinds of word games or ways of celebrating verbally) or Freud's "word magic" (basically the ineffable particularity of a language). But overall I'd say Fishman's thinking on this question became more sophisticated later on. (This does clarify why in his later work he needed Whorfianism as a celebration of linguistic particularity and profusion of variation to be Whorfianism of the "third kind" as opposed to Whorfianism full stop, which would have seemed to be more in keeping with his own concerns as well as Whorf's intent: he had already "systematized" it in a way orientated to experimental work.) Behavioral Science 5(4). ( )