Parthood and Indeterminacy

Normally I'd post this comment to the Phil. Mind. message board, but the graduate students are swamping that thing out of existence, and what I have to say is veering away from Phil. Mind., so I thought I'd post it here.

When discussing the indeterminacy of the word 'gavagai', Quine gives 'rabbit fusion' (that is, the set of all rabbits), 'rabbit', and 'undetached rabbit part' as possible referents of the word. It strikes me is that all these concepts are united in parthood. A rabbit is part of the set of all rabbits; an undetached rabbit part is obviously a part of the rabbit.

In our Phil. Mind. class, Professor Dennett suggested that natural kinds nestle into hierarchies, and it's indeterminate which of those natural kinds a folk term like 'water' refers to. Does 'water' refer to certain isotopes of H2O, all isotopes, or a higher-level natural kind that includes D2O (heavy water)? We might have different intuitions, but it seems we need to lay down a stipulation. Just appealing to a 'natural kind' as Putnam does doesn't help, because there are many. Another example: if a marsh tribe has a word for a smelly gas that erupts from the bog outside their village, that word might mean methane, gaseous hydrocarbon, or simply gas. Note that parthood seems to be in play in these examples too! Methane is a part of the set of gaseous hydrocarbons; gaseous hydrocarbons are a part of the set of gases.

Here's my question: what's the relation between Quinean indeterminacy and parthood?