Ontological Pluralism: the doctrine that there are different fundamental ways of being.
To put it more specifically to illustrate the point (although this not statement that ontological pluralists are necessarily committed to): the world is made up in such a way that our best metaphysical language will include multiple quantifiers.
Common candidates for ontological pluralist views are objects which are categorically distinct from one another; actual/possible, substance/accident, mental/non-mental, etc.
So, on this view, if we commit ourselves to the existence of actual and possible objects, the distinction between these two kinds of objects is fundamentally a distinction in their modes of existence, NOT their natures or properties. So, on this view, our ideal metaphysical language will include two fundamental quantifiers: actual-exists, possible-exists, which can be written as ∃a and ∃p, respectively (and the same can be done with the universal quantifier).
These quantifiers carve reality at its joints. As such, they are more fundamental than the general quantifiers of standard first order logic (∃ and ∀). I'm not sure if all ontological pluralists are necessarily committed to this view, but for my purposes here it won't really matter. I'm concerned with the metaphysical reality of modes of being. Problems with what language is best used to describe reality can be put to the side. Hopefully, what I've said so far illuminates the view.
The radical distinction between an actual and merely possible coin is grounded in their distinct modes of being (a-existence to the actual coin, p-existence to the possible coin), NOT their natures or their properties.
Okay, time for the argument. Consider the distinction between substance and accidents.
D1. S is a (primary) substance (def =) S is metaphysically capable of independent existence.
D2. A is an accident (def =) A is not metaphysically capable of independent existence but exists, if it exists, in a substance as its modification.
'Metaphysical' = 'broadly logical'
According to D2., an accident cannot exist apart from the very substance of which it is an accident. Therefore, there is a necessary connection between a particular substance S and its particular accident A. The necessary connection cannot be grounded in the nature of A, since natures are multiply realizable - natures must be stated in general terms. It cannot be part of the nature of A that it exists as an accident for this particular substance. Therefore, nothing in the nature of A can specify it to a single substance S. Therefore, the fact that A cannot exist outside of S is grounded in their modes of being (substance-existence, accident-existence, for example), not their natures.
Here's another argument.
Phenomenal states such as pain and pleasure have (as Searle puts it) a "first-person ontology". The existence of pain is identically its being perceived. But the existence of a brain is not identically its being perceived. Pain exists in a first-person way, brains in a third-person way. What accounts for this distinction in a particular case of a pain P existing in a brain x? It can't be the nature of P - since natures are multiple realizable. It cannot be that P by its very nature exists in x. Therefore, the ontological dependence of a particular pain P on a particular physical state x is grounded in its mode of being (phenomenal-existence, for example).
What are some ways one might respond to this line of reasoning? Here is one way:
I'm not sure exactly how the introduction of modes of being is supposed to ground the connection between, say, a particular substance S and an accident A. I'm not sure how the introduction of the mode of being 'accident-existence', and claiming that A has this mode of being, is supposed to explain A's necessary dependence on S. How does A's mode of being tie it down to this particular substance, any better than its nature does? The same can be said about the pain-brain case. How does a particular pain P's mode of existence (call it phenomenal-existence) explain its necessary tie to a brain state x, any better than its nature does?
I'm also wondering if there are other accounts of the substance/accident relationship than the one given above (an account that does not necessarily tie a particular accident to a particular substance). Also, if there are any other accounts of the relationship between the physical brain and phenomenal states (an account that does not necessarily tie a particular phenomenal state to a particular brain state)?