Theistic Metaphysics is Alive and Well
September 13, 2006 — 20:20

Author: Trent Dougherty  Category: Christian Theology  Tags: , ,   Comments: 3

I just read Brian Leftow‘s “God and the Problem of Universals” which is in the latest Oxford Studies in Metaphysics (vol. 2) (Table of contents available here). It is divided into four sections of three or four essays each and Part IV is “Metaphysics and Theism”. [I previously interviewed Leftow for Prosblogion (link).]
I’ve mentioned my desire to explore divine conceptualist alternatives to the regnant Plantingan Platonism previously here. [By the way, another part of the Plantingan Paradigm is a “relational” vs. a “constituent” ontology. Michael Loux has an essay on this in the same volume (and Wolterstorff discusses it in “Divine Simplicity” Phil Perspectives 1991.]
Leftow’s piece is meandering and mysterious at times, but he’s doing front-line work which means there are also some very exciting ideas as well. I’ll mention a few things I find especially mysterious and especially exciting below the fold.

I apologize in advance for the length of this post, but it is near the least amount possible to give a taste of this essay.
A. Conceptual Foundations
First, actually, some preliminaries. Leftow’s main theses is as follows:
1. If God exists then arguments for Platonism (indeed realism about attributes/universals generally) are weakened.
2. If 1, then we have an argument for God’s existence from theoretical economy.
The view, which will have to be defined in terms of the discussion, is called “Theistic Concept Nominalism.” The basic idea is to follow Augustine–and Aquinas after him (ST I.84.5)–in placing Platonic universals with concepts in the divine mind. Since God’s concepts–like all concepts presumably–are concrete, abstracta are eliminated (more precisely: to the extent that divine concepts can do the work of Platonic universals, to that same extent abstracta are eliminated).
And when we have multiple models which cover the data–in this case the existence of trans-linguistic and trans-personal meanings, the existence of natural kinds, regular activities of entities, etc.–we ought-rationally to believe the simplest one, we have here an argument for theism. Plantinga has somewhere–in conversation at least as well as, I think, in his “Two-Dozen or So Good Arguments for the Existence of God”–made this same kind of argument in less detail. Leftow compares it to David Lewis’s argument or concrete modal realism: if more tokens reduces types, then we’ve gained in economy. Such reasoning can be questioned, but I’ll not question it here.
He defines Platonism as the view that there exist unexemplified properties. I think this is not a good definition at all. The argument itself involves not what is exempliefied or unexemplified but rather whether there are abstract objects. He runs into problems here (which I’ll note below), and so I think it best to define Platonism as the view that there exist subsistent abstracta. If we define an abstract object as one which is such that, possibly, it is exemplified, then a subsistent abstract object will be one which is in no sense a part of another object. Our concepts are in some sense a part of or a constituent of us, likewise with God (friends of Divine Simplicity already have to find harmless senses of “part” so there’s nothing new here regarding that).
B. Some Mysterious Locutions
1. Cats
Here’s one example of the kind of thing that puzzles me:
“God stipulates that cats shall be mammals, forming His cat-concept so that this is part of it” (326). I’m not sure about the referents of the “it”s here. I suppose the first “it” is the property [being a mammal] and the second is the concept [cat]. Later (340) he says “As Creator, God simply stipulates: ‘I declare it true that Boots is a cat.’ As a result, Boots is a cat.” This sounds like a Divine Command Theory of Attributes. He has me wondering if God “already” (talk of “forming” worries me in God’s case) had the concept [mammal] “before” he formed the concept [cat]. If he did, then it seems like he couldn’t have done otherwise, so I’m not sure what the force of “stipulates” is (and the quoted sentences is followed by “There is nothing Platonic to say him nay.” The idea seems to be that God could have made non-mammal cats.
Whether this is so depends on what the truthmakers for logical “laws” are. These propositions aren’t just ordinary old propositions, they seem to govern what even God can and can’t do. Leftow doesn’t say anything that I noticed about how to nominalize Laws of Thought. However one of Leftow’s exciting ideas can plausibly be applied here in such a way that the problem goes away and he doesn’t need such stipulation talk (and on the next page it’s God’s *concept* which does the stipulating, surely just a slip (or perhaps a superfluous recursive notion).
2. Gruppies
A related issue comes eight pages later (p. 336). Leftow has been considering a realist argument based on the “natural unity of classes,” which I’ll just call natural kinds. His answer to the question “What makes grizzly bear a natural kind and not gruppy?” (I’m actually using Plantinga’s example here b/c I like it better) is that when God made cats he intended to make it the case there exist something that falls under his cat-concept.
He considers the objection whether gruppy would have been a natural kind if God had had different intentions. He really doesn’t say much helpful in reply which is unfortunate because I think this is the kind of thing that could make or break a theory. The closest he comes to replying is to say this:
“Perhaps God would guarantee that we perceived as natural whichever properties He’d made so. And perhaps God’s concept cat just is the kind of concept He would no matter what make a natural kind: perhaps that’s jus how His mind works” (337).
I think the first perhaps is just to say there really are no natural kinds. The second one is not much more helpful as it stands. Is this really the kind of thing that can be a brute fact? Surely God’s thoughts go that way because of something about the concepts themselves. This needn’t have a Platonic reading, for we could apply some thoughts that Leftow develops later in the paper.
3. Brute facts
There are an alarming number of these scattered throughout the paper. One just mentioned was God’s thinking in such a way that [cat] marks off a natural kind. Here are some others:
p. 329: “We could treat being contentful, and contentful in one way rather than another, as a primitive fact about mental states or events.”
p. 330: “we can say: here’s a mental state with certain roles in God’s thinking an willing, such that when that state guides God’s action, He produces mamal cats. Why does it have these roles? It just is that way: it just does.”
p. 331: “But why did the state grasp this attribute, not another? No answer: it just did. It just does have the causal role of grasping one attribute, not another.”
p. 351: “But that a power is one to bring it about that P does not entail that it involves an entity, a proposition P, in its inner constitution. It may just be a brute fact about that power that this is what it does.”
The first three are not on exactly the same topic. Explanation must stop somewhere, but I felt he was a bit too casual about it in this paper. I think it takes pretty good reasons to reasonably posit a brute fact: reasoning about what *kinds* of facts can have this status and reasons for why a particular case is of this kind.
5. Pre-Creation Intentionality
As a consequence of a flawed definition of Platonism (see above) he is worried about the fact that God will have some concepts which nothing falls under. This leads him to say some mysterious things.

“Since (surely?) not all attributes of which God has concepts have instances, there are then attributes that exist uninstanced [and thus Platonism looms]. This line of thought broaches deep issues I cannot discuss here. But I will make a suggestion. If there are unexemplified attributes, the divine conceivings whose contents they are either do or do not produce them. If a conceiving brings one to be, presumably something about it, logically before the attribute exists, determines that the conceiving will bring *that* attribute to be–directs it to produce *that* attribute, not another. Whatever that is, why can’t such a targeting, on its own, without producing an attribute, make the conceiving state one directed toward the attribute, not another?….But then we can just suggest that the state’s ‘aboutness’ consists in this targeting, deleting all reference to the targeting producing an attribute, and hold that due to this targeting, if God in the right way acts on what He in this state conceives, He’ll produce a particular extra-mental attribute, perhaps by producing an instance of it. Such a state would grasp (target) what it would be like if God did produce it” (333-334).

And on it goes like that encountering a new technical problem that I won’t even go into. I’ve read this three or four times and can’t make anything of it. My main concern is that the whole problem is a red herring given a proper understanding of what is objectionable about Platonism: *subsistent* abstracta, not *unexemplified* exemplifiables (concepts might still technically count as abstract, but not in the way that should be bothersome).
C. Some Really Exciting Ideas
1. Power-based metaphysics
For my money, the fun really begins when he starts when he starts defending Divine Concept-Nominalism from charges of re-located Platonism. In spite of what I have said above, some will still insist that timeless, immutable, unexemplified divine concepts are no better than Platonic Universals. Leftow at length:

“When I say that God has concepts, my ontological commitment is not what it might seem. having a concept makes us able inter alia to discriminate, recognize, … , classify, use terms, and grasp, infer, affirm, and deny claims: broadly to have a mental life of a certain sort. Having a concept is being in a mental state which is or gives us a set of powers….In saying that God has concepts, the most I commit myself to is that there is in god whatever underlying reality makes it apt to speak of concept-possession. This reality may be just God’s possession of certain powers” (349).

I think this is just the right kind of move to make. It fits in quite nicely with the Scholastic idea that God’s esse is the end-all-be-all of explanation and it links up nicely with Alex Pruss’s recent paper in Blackwell Guide to Metaphysics which builds possible worlds out of God’s powers. Also, powers of agency are the kinds of things we know intimately, about as well as we know anything (a la Reid). Powers are capacity for causal effectiveness and we’ve already got reason to think causation is conceptually prime.
2. Divine Quine-events
One more move I just love is encountered in replying to an objection to the above move. The exercise of a power is an event and a popular account of events–Kim-events–has them as essentially property-instances. Leftow notes that Quine had a sloppier but perhaps better notion of an event. A physical event over some duration and throughout some extent was just the total content of that region of spacetime. Glossing this, I say that what occupies spacetime are substances. Yes, substances exist in certain ways (see also Heil 2003). These substances and their modes of being are all concrete individuals.
So how does this apply to the Divine Concept Nominalism? “[A] divine event is simply a part of the divine life…” (350). This then solves the aforementioned problem of the content of God’s concepts which led to postulating brute facts above. God’s concept [cat] is about cats because of what it’s a concept of Him causing to come into being. His concept [centaur] has what content it does because of what there would be if God caused there to be one. You’ve really got to read Leftow on this, the remaining two pages are poetic compared to the dense stuff in section B.
The final objection concerns the notion that substances have their causal powers in virtue of their properties. I suppose Leftow reverses the order of explanation. God just does have certain causal powers. His having the property of having these powers he does not think is a problem for it is not an unexemplified property. I would probably try to press an identity theory of this kind of property. It is just identical with the fact that He has such-and-such a power. This makes certain modal facts basic facts, but I think we were stuck with that anyway (unless you like concrete modal realism).
It is a truly fascinating and rewarding paper, and I highly recommend it. I hope a bunch of really smart people continue to whack away at this issue so epistemologists like me can reap the benefits.

  • Hmm. Ockham’s Razor in service of an argument for the existence of God. You still don’t have the economy of nominalism, of course, but I suppose if the Platonist arguments are supposed to win out over nominalism then this might just win out over Platonism. Leftow needs to say that the arguments for Platonism are still good arguments against nominalism but that their conclusion is not Platonism but some disjunction with Platonism as one disjuncct along with a theistic conceptualist disjunct as another disjunct, one that parsimony takes as the most likely disjunct.
    It’s the last step that bothers me a little bit. What kind of simplicity or parsimony do we care about? Berkeley faces this issue. With the external world and without God, we have a whole lot more stuff but no mega-being. With Berkeley, we have the mega-being with a lot less stuff. Berkeley insists that one mega-being of a kind we already admit to (minds) is more parsimonious than lots and lots of things whose existence is unnecessary. One very powerful being is more simple than lots of not very powerful beings. But the opponent thinks God’s being so powerful makes God a very different kind of being, and that makes for less parsimony in a different way. Both views seem right to me, but that’s because they’re relying on difference senses of parsimony. Which is the kind we mean when we think Ockham’s Razor is a good principle? I’m not sure if either is more in mind.
    Isn’t the same debate going on between the person who thinks God is simpler than Platonism? The Platonist has lots of these Platonic properties, and the theistic conceptualist just has God, one mega-being as opposed to lots of non-mega-beings. Which is more parsimonious, the one with lots of extra things or the one with a kind of thing that is way beyond what we might otherwise be sure of (at least if you start where you need to start if this is to be an argument for God).

    September 16, 2006 — 2:52
  • Daniel D. Novotny

    Thank you Trent for this post. I am convinced that Leftow is going in the right direction. Platonism vs. nominalism is a false dilemma. In my view, the middle course Aristotelian-Scolastic constitutionalism is the truth (whether in its more conceptualist Thomistic or more realist Scotistic version).
    Ad Ockham’s Razor: As one of the 17th century Scotists John Punch observed (and Jeremy as well), it it to be used prudently: just because we may explain many motoric phenomena by three fingers does not mean that we do not have all five.
    (On why Ockham’s Razor should be called Scotus’s Razor 🙂 )

    September 23, 2006 — 5:53
  • David Alexander

    Thanks for the post. I just got the paper and am excited to dive into it.
    Michael Loux has a paper in Midwest Studies in Phil in 1986 that seems to be pointing in the same direction. Also, Greg Welty (a phil prof at Southwestern Baptist Theological Seminary) has a very clear and concise paper on related points in the most recent Southwestern Journal of Philosophy

    September 29, 2006 — 16:31