It's ontological. My forthcoming Qui Parle essay is going to talk about this a bit. If you think that Graham really means that objects have a “molten core” that is spatial, material or temporal (that is dimensionally inside objects, or materially a substrate of them, or temporally revealed at some point), then:
(1) You are a reductionist (sorry but yes, you are). You think that objects have roughly two parts: an outer shell and a molten interior. This molten interior is more real than the shell. You have reduced the object to a smaller component.
(2) You are prejudiced against non-3D objects. Suppose a 2D object is real. It too withdraws. An object in a 5 dimensional phase space withdraws. A 1D object withdraws.
Neither is withdrawal in the object's future as opposed to its present; nor is withdrawal in the object's past. If it was, then at some point it would not be withdrawn. So withdrawal would not be a deep ontological fact.
Withdrawal means: even if you penetrate to the “core” of a 3D object you will not find it.
Withdrawal means: even if you monitor the object for trillions of years you will not find it.
If you think withdrawal is spatial, material or temporal, you are smuggling an ontic prejudice into an ontological realm. Ontic meaning somewhat clichéd everyday factoid (kind of). It would be better to let Graham's Tool Being explain this as he does it about 200 times (I got it the 6th, 35th, and 107th times...) and I have little space here.
Now if you don't like the metaphor, strictly, then let me stick up for withdrawal a little bit:
(A) All metaphors suck to some extent.
(B) Our culture privileges extraversion over introversion. We don't like things that are hidden. Everything must be visible (Twitter, phone cameras, etc.). Queer theory (and queer ecology) must redress this (frankly) violence. I respectfully suggest that the anxiety that everything must be seen fuels some of the anxiety about withdrawal.
(C) All other metaphors will have their unique faults, precisely because they are translations of a deep ontological fact.
