On 'The Limits of Contingency' & Metaphysical Possibility

An article by Gideon Rosen entitled 'The Limits of Contingency' recently peaked my interest. Rosen makes an interesting distinction between two conceptions of metaphysical possibility - what he calls Standard and Non-Standard. According to the standard conception of metaphysical possibility, a property P is possible for an object x iff P can be possessed by x in a world whose 'form' is the same as the actual world. By form, Rosen means something like the fundamental laws of how objects are composed or how causation goes about - something broadly like 'ontological rules'. According to the non-standard conception of metaphysical possibility, a property P is possible for an object x iff P is logically coherent with the 'what it is to be' of x. That is, only if P does not logically conflict with x's 'individual nature'.

The first thing to notice is that the standard conception requires a certain checking against the conditions of the actual world - a position which I think represents a respectful speculative sobriety. The non-standard version of metaphysical possibility only requires that a property be compatible with the 'nature' of an object and presumably, since an object's nature might very well be intact while it does not exist in the actual world (think Plantingian Essences or Platonic Divine Idea Essences), metaphysical possibility is not constrained by the conditions of the actual world.

I think that the two positions - the standard and non-standard - are, as Rosen claims, very common notions of metaphysical possibility. Those who think that sortals are the de re governors of what is metaphysically possible for an object clearly fall into the non-standard camp, though this seems a rather standard position in modern philosophy. And those who think that metaphysical possibility has something to do with the laws of the actual world will, in most instances, fall into the standard camp; those who do often want to identify metaphysical possibility with physical possibility, broadly construed.

At the risk of being philosophically trite, my sentiments lie with the middle position. In general, I think that the non-standard version is correct - metaphysical possibility must be de re in a significant sense and this sense must have to do with the 'nature' of the object in question. On the other hand, however, the non-standard conception does not appropriately respect the privileged status of actuality - there are no merely possible worlds over which we might peer out and look down upon to inspect the natures of the objects within. Any considerations that we have of this kind - where we think we have real knowledge of the modality of non-existent objects - seem to me to be based solely on our powers of conceptualizing, not on the nature of these objects themselves.

Metaphysical modality, I think, must be about objects and so must also be about their 'definitional natures', but it cannot merely be a matter of logical non-contradiction without consideration of the actual world. On the other hand, it equally cannot concern the 'form' of the actual world, except insofar as 'form' is taken to mean the natures of the objects within it.