easily imagined errors

[Jon Aske]

> The point is that the child knows/learns structure.

But what, may I ask, is "structure"?   I'm mesmerized.  First you get rid of the *meaningful* relations which bind elements together into "structure".  And then when you notice that elements are bound together somehow, you wonder how that can be.  Don't you see the inconsistency here?  The child learns absolutely zero "structure" independent of the meanings associated with that structure.

> little knowledge of the complexities of structure-dependency.
> The "cognitive unit" business is just hand-waving.

And the empty categories and other hocus pocus categories and structure are not hand waving, right?

And about this business about the "complexities of structure-dependency".  This stuff only seems complex because you try to look at it without taking function into consideration.  Once you look at functions, everything starts to make sense, for the linguist and for the learner.  Sure you can't reduce form to meaning.  Nobody's trying to do that though, as you well know.

Jon