I've been reading Ted Sider's 'Ontological Realism' in preparation for the 'Being' conference this weekend. One of the interesting questions Ted raises is whether there could be a motivated position that held:
(1) That there are genuinely natural kinds - objective structure or 'joints' in the world, which our theories might aspire to capture or reflect.
(2) That there are some logical natural kinds (perhaps that the world has an objective structure which mirrors or underpins the structure of the network of logical relations holding between propositions, or that there are logical 'joints' in the world)
(3) That existence isn't a logical natural kind.
I wondered whether somebody might suggest something like the following. The best way of fleshing out the metaphor of 'carving nature at its joints' is something like: having our choice of theoretical primitives (in some privileged, perhaps idealized language) reflect the objective differences between entities in the world. So, for example, there's a 'joint' between electrons and positrons, because electrons objectively differ from positrons.
If you had this kind of 'difference-first' approach to natural kinds, you might think that existence isn't a joint. It's not like we have two classes of things in the world, the existing ones and the non-existing ones (pace Meinong), which we have to differentiate from one another theoretically by distinguishing the ones we quantify over with our unrestricted ontologically-serious quantifiers, from the ones we don't. Beings v Non-Beings isn't like electrons v positrons.
Of course, there's a stronger position, that urges us not only to capture objective differences, but objective similarities. I guess beings all have something pretty objective in common.
But we might have independent reasons for taking difference to be ontologically more basic than similarity. (I've heard Peter Simons defend this view in conversation on truthmaker grounds, for example. It may be easier to find something that genuinely seems to metaphysically explain why things are different, than something that explains what makes them the same.)
If difference is the ontological ground of similarity, then we might have a principled reason for wanting our best ontological theory just to have enough structure to capture all the objective differences. And maybe that would provide a halfway principled reason for refusing to count existence as a logical natural kind?