Immanent Realsim: Another Objection and Another Response

                  One objection to Immanent Realism -- i.e., that universals are not abstract objects but spatio-temporal entities that are in some sense located "in" the various instances that instantiate them -- is that this implies that the very same object, say, the universal 'being yellow,' is located in more than one location at one and the same time. For if (1) universals are the principles of being in virtue of which different predicates that are predicated of or said of different objects are the very predicates that they are and not some other predicates, and also (2) that these principles are located "in" the various instances of which they are predicated, then it seems to follow that (3) these principles will themselves be variously located, depending on the spatial locations of their instances for their own spatial location. But this, it is said, is absurd. For this will imply that the very same thing, i.e., the property 'being yellow' can be wholly located in one location, say, here, on my shirt and also over there, say, "in" my neighbor's car.
                 There are a number of ways I think that the Immanent Realist might reply to this objection. But one that I'll look at just a little more closely here is simply to deny that the entity that is located in each instance of the universal is really identical across the various instances of that universal. The view itself is hardly an original one on my part. It's the view that was originally defended by St. Thomas Aquinas and arguably by Aristotle as well. (It also might have been defended by Aquinas' near contemporary John Duns Scotus. I say "might" here simply because while I don't know whether this is actually true, it is a view that would seem to fit well with Scotus' overall view of universals.)
                 And according to this view, it is only when one abstracts the individual natures (i.e., the multiply instantiated natures or, to use the more contemporary terminology, sortals) of, say, Socrates and Plato from the two men does one get something that is in some sense the same in both Socrates and Plato. For, in reality, according to this view, it is not possible to abstract a thing's nature from its individuating features, i.e., those features it has that determines its being the very individual that it is and its having the very species or kind that it has, without causing that thing to cease to exist. Thus, as these individual natures actually exist -- in the individuals that have them -- the natures themselves are not identical to each other (with identity here taken in the sense of strict identity, as codified in the form of Leibniz's Law), but are merely similar to each other. (By itself, the two-place predicate <the same (x) as> is neutral between qualitative identity and qualitative similarity -- in the present, I mean it in the latter sense. But, if used to refer to numerical identity, then clearly it does not have this leeway, and means only identity in the strict sense.) They are identical only if one abstracts or takes away from them all their individuating features, by means of which one arrives at something, i.e., some entity, that is not particularized or "individualized."
               Because it exists merely as an abstraction, however, such an entity can exist only as a mental entity; there is never a non-individualized human nature that one encounters "out in the world." 'Humanity,' then, exists only as the abstracted concept common to the individual human natures that Socrates and Plato share. It is abstracted because the human nature of Socrates and the human nature of Plato are such that they cannot exist without, respectively, either Plato or Socrates. And thus the universal 'humanity' is a concept because while such an entity exists, it exists only in the mind.
              Thus, contrary to the original assumption, it is false (on the present view at least) that there is something in Socrates that is identical to something in Plato, namely, humanity or human nature. But, then, this will seem to run into difficulty insofar as it commits one to denying the truth of the necessity of identity. Very roughly, the necessity of identity states that two entities x and y are identical only if x and y are necessarily identical. (Adopting them for at least the purpose of convenience, one might say that x and y are necessarily identical only if x and y are identical in all possible worlds in which x and y both exist -- i.e., that they meet the criteria sufficient for 'transworld identity' between one and the other.) But, on the present view, x and y are different entities just in virtue of these specific individuating features. And in this case, it seems true also that if it were the case that y possessed x's individuating features, y would have been x. (And likewise if x possessed y's individuating features, x would have been y.)
             But I think there is reason to question the counterfactual conditional that <if it were the case that y possessed all of x's individuating features, then y would have been x.>. For if the individuating features that are sufficient for x's being the individual associated with those features -- namely, x -- are absent, it is simply false to say that this individual is x. Moreover, if no individual has these features, it is false that there is an individual such as x. And if no such individual exists without its having these features, it is also false that some other individual besides x would have been x if this other individual had had those features instead of x's having them. For, then, this other individual would simply not have existed, and (of the two of them) it would be x alone that exists. (Say that this other individual is named y).
              Now, if y has both (i) all the features that are sufficient for being x and (ii) all the features sufficient for being y, then it follows that (iii) all the features sufficient for being x are the same as -- i.e., identical to -- all the features sufficient for being y. For assume that x and y are not identical. Then it follows that x and y are two distinct individuals and also that x and y are identical. But, by definition, no two individuals that are distinct are identical. So we have a contradiction. And therefore we must reject the assumption that x and y are not identical. And this in turn gives good reason for rejecting the earlier assumption that if x were to possess y's individuating features, x would be identical to y. The upshot of this last point, moreover, is that it gives good grounds for rejecting the earlier assumption that the current proposal -- that universals are abstracted concepts -- entails the falsity of the necessity of identity.