Everything but the beetles cancels out
This is a continuation of my last post and also partly a reply to concerns raised by my tutor Brian Weatherson.
Imagine a small community consisting of three elm experts A, B, and C.
First case: Each of A, B, and C knows enough to determine the reference of 'elm', but their reference-fixing knowledge differs. However, they belief that their different notions of 'elm' necessarily corefer. This is the case Lewis discusses in 'Naming the Colours'.
Second case: None of A, B, and C knows enough about elms to determine a partcular kind of tree, but taken together, their knowledge suffices. Moreover, they believe that 'elm' denotes the kind of tree determined by their joined knowledge. Such a case also seems possible.
In the first case, A's concept 'elm' picks out natural kind E iff E satisfies A's elm-platitudes, and A's and B's and C's elm-platitudes are satisfied by the same kind. Equivalently, iff E satisfies all three sets of elm-platitudes. Likewise in the second case. Here too A's concept 'elm' picks out E iff E satisfies all three sets of elm-platitudes. And of course exactly the same holds for B's and C's concept 'elm'.
It seems that to analyse the concept 'elm' in a linguistic community, the metalinguistic or deferential components of the individual concepts have always to be resolved and can afterwards be ignored, because they don't make any difference.
(I'm still not sure if any of this helps two-dimensionalism. On the one hand, the meta-linguistic part of A-intensions helps to explain how people can understand each other even though their individual concept's A-intensions differ. On the other hand, this very explanation seems to show that A-intensions can't do all the wonderful work they are supposed to do. This should be thought through much more carefully.)