John Heil has been arguing for a while (cf. his book From an Ontological Point of View, OUP, 2003) that an important source of problems in the metaphysics of mind (e.g., about mental causation) is the tacit though widespread endorsement of the following principle [I am simplifying Heil’s formulations a bit]:
(P) Every non-empty predicate refers to a property.
Heil thinks that instead of (P), we should endorse the following weaker principle:
(P*) Every non-empty predicate applies to objects in virtue of the objects’ properties.
Heil’s point is that under (P*), we don’t need to suppose that there are higher levels of being (i.e., higher level properties). And without higher levels of being, we don’t run into the problems of mental causation that, e.g., Kim’s career has been built on.
I just read a recent paper in which Heil accuses Putnam of tacitly endorsing (P) in his influential formulation of functionalism, contributing to its acceptance among philosophers and with it, to the problems that it generates. Heil’s paper (“Functionalism, Realism and Levels of Being” in the volume on Putnam edited by Conant and Zeglenis) is followed by a baffling response by Putnam:
“I did not accept your (P) in my arguments, but I do want to defend it. One reason for accepting it is that all of its instances are theorems of what I think of as the only good formal theory of properties (i.e., of predicates in intension) that I know, namely Russell’s Principia Mathematica or ‘ramified type theory’ minus the Axiom of Reducibility” (Putnam 2002, p. 143).
I asked Heil if I was missing something, and he said I am not. He added that he presented the paper at a small conference on Putnam that took place in Poland. In person, Putnam seemed to like much of what Heil said. But in his written reply, Putnam is less clear. It is hard to see how the position as he now describes it connects with the current debate about mental properties and their causal relevance.
Heil added a philosophically interesting anecdote. At that meeting, there seemed to be two groups of philosophers. One thought (P) is obviously false, to the point that no one could possibly accept it. The other group thought (P) is obviously true, to the point that it would be crazy to doubt it. As to Putnam, he seemed to be in both camps.
Any takes on (P)?
Some thoughts, in no particular order:
1. I know you’re simplifying Heil here, but why restrict (P) to non-empty predicates? A predicate might express a property that’s contingently empty, but has a nonempty extension in other worlds. Maybe ‘is a unicorn’ is an example. For that matter, couldn’t there be necessarily uninstantiated properties? Certain logically complex predicates seem to express those: F & -F, for instance. (Assuming that there can be complex properties, of course.)
2. Modulo point 1, on a pleonastic conception of properties, (P) seems trivially true as well. See Schiffer’s ‘Language-created language-independent entities’ (Phil Topics 1996).
3. There are lots of purposes for positing properties. If they’re supposed to be entities grasped as part of linguistic understanding, then there should be as many of them as there are meaningful predicates. Ontological motivations for properties (explaining causation/realization relations, e.g.) might lead one to be more stingy with them. Compare, too, Lewis’s distinction between properties and natural properties (in ‘New work for a theory of universals’).
It’s been a long time since I read it, but I remember Mark Wilson’s ‘Predicate meets property’ as having some discussion of this point. Ditto, interestingly, for Putnam’s ‘On properties’.
E. J. Lowe offers the following proof that P is false: consider the predicate “is non-self-exemplifying”. Given a liberal theory of properties, many properties will satisfy this predicate, but there can’t be a property corresponding to it without paradox. Maybe Putnam invokes ramified type theory in order to try to avoid this problem? (Can he avoid it?)
Dan – Heil is definitely motivated by ontological considerations.
Here’s another reason to deny (P). Rene Descartes’s evil twin Skippy
argues that since his mind has the property of being indubitable by
him, and his body doesn’t have that property, his mind is distinct
from his body. The textbook refutation of this argument is that the
predicate ‘is indubitable by me’ doesn’t refer to a property.
As Daniel Weiskopf points out, what you think of (P) will depend a lot
on what you think properties are for. Putnam’s 1970 paper “On
Properties” (in Philosophical Papers v.1, originally in Rescher, ed.,
Essays in Honor of Carl Hempel, Reidel) is about the idea of “physical
properties”, which I think is not the same as the conception of
properties Putnam says (in this recent paper) we find in Russell.
So, given a suitable conception of properties, (P) may be true or
false. Heil is interested in ontology, and he is interested in
causation, and so he proposes (P*) instead of (P). This is one way out
of the mental causation problems that come from holding that the
predicates of psychology refer to properties that are not identical to
the physical properties.
One way. If you think you have another way out of the mental causation
swamp, then even if you agree that (P) is false in general, you could
still say that the predicates of psychology refer to properties. Ditto
for color, for example: if you can work out a theory on which there
are color properties that are not physical properties, the falsity of
(P) is not a reason for you to reject your theory.