Acutally, no, I meant exactly what I said: Lojban was formed to test
the Sapir-Whorf hypothesis, and while that's largely irrelevant to
me, it *is* important to me, and many other people, that Lojban is
not like any natural languages in a lot of respects. Makes it all
mind-stretching and stuff.
This is an interesting issue, I think, because I see the idea of "mind-bending stuff," while certainly very cool, to be diametrically opposed to any hope of having Lojban widely adopted as an international/auxiliary/talking-to-computers/whatever language--there are plenty of things that natural languages do which are plenty mind-bending to speakers of English, but they don't violate general principles of what languages do and what they don't do, while Lojban does in any number of respects.
If Lojban wants to be mind-bending, awesome. If it wants to be a language which is logical and consistent, and thus primed to make in-roads in being widely used, also awesome. But it can't do both of those things.