A tentative new feature on philosophy ktl: quick notes, trying to describe what I think about whatever I read today.
The thesis, more or less, is that the appearance of universals in language is due to the fact that actual relations have to be represented with a meager supply of spatial relations of words. That something is on top of another thing could be represented easily with pictures, but the pictorial relation is referred to rather than shown in words. Result, there appears to be a universal concept "being on top of." Objects relate to each other, not to the relations which relate them. Because it is by their power of so relating that they are objects at all, the concepts are a function of the being of the objects.
I'm not convinced that Wittgenstein "dissolved" the problem of universals in the Tractatus. Anscombe probably doesn't either, but it's hard to tell. At any rate, she doesn't make anything of the fact that importing universals into objects as "properties" doesn't eliminate them. I think the virtue of this procedure, if its pretense to erasure can be erased, is that it returns the universals to their starting point.
Not sure what I mean by that except that I get an exciting sense of recollection when I think of universals as an illusory residue and transpose them into the unprotected, unguided sheer manifold of being of things.