Donagan, Alan - Universals And Metaphysical Realism

10/29/2010

The Philosophical Papers of Alan Donagan Vol 1 Historical Understanding and the History of Philosophy, Ch 12. Malpas, Eds. University of Chicago Press, 1994

In this paper Author tries to revive Bertrand Russell's realist theory of universals. Author lays it out using Russell's own language from Problems of Philosophy: 'I am in my room' has a relation 'in' that connects me with my room. The 'in' relation itself has a reality to it. (pg211) The reality of relations like 'in' gives way to the reality of universals quite immediately: me being in my room and your being in your room is not two different cases of 'in'. They are the same case; they aren't two particulars. But they happen at two different places (perhaps at the same time). Thus the two cases of 'in' aren't one particular either. So, the 'in' is a universal. (pg212)

The possibility of the existence of a relational universal is met with the following objection: it seems possible to establish the reality of any particular relational entity by means of a negation proposition. A negation proposition is something like ~x(Relation)y. This is because we use relational entities in negative statements (which are also true or false) just as much as we do in positive ones. For instance: "x isn't in bed" is ~(x (in) bed), and seems to establish the reality of the relational universal 'in', but without, in this circumstance, the 'in' being exemplified (the proposition is true: x doesn't stand to the bed with the 'in' relation). This leads to the absurd conclusion that there are hundreds of existing relations that aren't exemplified. Author claims Russell tried to finesse this by having a Principal of Acquaintance: we have had immediate acquaintance with every understandable proposition; no acquaintance, no apprehension. Author pushes back that this won't do, since it is possible that there are relations we don't in fact understand but we could nevertheless put into a negation proposition. (pg 215-6) [What is author's way out? To assert that it is ok to have such real un-exemplified universals?]

The first objection and reply author discusses is the objection that the universal entity is caught between being divisible among all the places it is exemplified, or that it is indivisible but somehow present in multiple places at the same time. This is the original problem Plato encountered. Author denies that the second choice is problematic: it is only problematic for particulars, not for universals. Universals are naturally indivisible and exemplified multiply: that's what they do. (pg218) An issue along this same difficulty arises regarding the nature of exemplification. Exemplification seems to occur when there are true instances of a universal relating to particulars. But the trouble arises when one tries to explain the relation: x is related to y with the exemplification of the relational universal 'in'. In other words, the exemplification relation is exemplified in the case of x, 'in', and y. In trying to explain the relation of exemplification, an infinite regress arises. (pg219-220) The solution comes from denying that exemplification is a relation. The difference between "x is in bed" and "x isn't in bed" is that one is true and another is false, not that, in one case, 'in' is exemplified and in the other case, 'in' isn't exemplified. (pg221)

The second objection asks whether Russell's version of real universals is too bold. A more moderate Aristotelian version talks about essences and the sharing of them in particulars; yet not every relation has an essence. (pg222) The Aristotelian story has an essence unified as one thing in the human mind through a process of abstraction, yet it is distributed through the various particulars it is instantiated in. Author criticizes this as being even more obscure than Russell's theory of universals (pg223).

A completely different objection is raised by Quine (and, to a lesser extent, Goodman) in the nominalist program. For Quine, what exists is a function of what items are quantified in a set of true statements. Because you can construct a set of quantified statements no variables of which are universals or relations, then you don't need them in your ontology: they are dispensable. Against this objection, Donagan offers two replies: 1) Sellars has argued that quantifying over variables is not the way to determine ontological commitments. 2) if you admit that "two dogs are white", then isn't there a quality that the two dogs both have? (pg224-5)

The next objection might have been offered up inadvertently by Russell himself when he suggested that there could be an artificial language where normal predicates like '...is white' are replaced by a name for a 'discontinuous particular' like 'White'. So, to say that a wall is white, you might express 'wall white', the meeting of two particulars, 'white' and 'wall'. Author pushes back against this, claiming that it wouldn't be possible to establish the name 'White' without reference to the color white, which would have to be put together using some sort of predicate. But of course the language has no predicates, so there wouldn't be a way to establish that 'White' is white. (pg227)

Another objection is Pears': that realism seeks to provide an escape from the 'maze of words' into the real world by means of asserting that the words all refer (at least all the primitive ones do). While this seems informative, it doesn't do the work of actually getting out of the 'maze', and is therefore unhelpful. It is uninformative but appears to inform; it is just a circular system of words. (pg228) Author just says that realism is informative because it explains a fact about language '... is white' using a fact about the world: whiteness. (pg229-230)

The final section of the paper involves the author looking at Russell's motivations to hold his "Realist Principle": 1) not all relational predicates can be disposed of-- some are necessary for a language about the world (pg232-3); and 2) that statements with primitive predicates will be true or false based on how the world is comprised, not based on fantasy or idealism or nominalist acrobatics. (pg233) While it seems modest, this seems to be a realism worth having, since the price of not having it entails errors of idealism or nominalism. (pg234)