God, Predication, and Ontology

[I originally posted this 4 July 2004 but felt it was worth bringing "back to the future," not only, I hope, to help some readers (?) with the recent material I've written about necessity, possibility, and so on, but also to show how this topic has been abiding interest of mine for longer than I realized. I have bolded the sections that I think have value for further consideration.

I sent a copy of this essay to a former professor of mine UF, Gene Witmer, and he was nice enough to reply, though I can't find, much less, recall the substance of his comments right now. I know he is a rather staunch atheist and physicalist––he joked drolly once about "Saint Alvin"––so even though my writings did little to convert him, I must thank him for giving me my taste for metaphysics. He was a bit rough around the edges but a very good teacher.]

The classic repudiation, à la Kant (AD 1724-1804), of the ontological argument for the existence of God (OAG), à la Anselm of Canterbury (AD 1033-1109) in his famed Proslogium, is that, since existence is not a predicate, necessary existence cannot be predicated of God. Anselm argued that God is the being than which none greater can be conceived, meaning that God is infinitely greater than every other entity. Part of this “maximal greatness" (to use a term of Alvin Plantinga’s in his modification of OAG) includes existing, since any existent being is greater than any nonexistent being.

A common example is that a perfect, but nonexistent, island is inferior to an even slightly less magnificent, but existent, island by sheer virtue of the fact that the latter actually has its magnificence, while the former does not actually have any magnificence. In lay terms, ten real dollars is greater than a billion imaginary ones. An actual bird in hand is better than two potential ones on the roof.

In reference to God, this fairly straightforward observation about the superiority of actuality over potentiality is that, if God is the greatest possible being, then he must exist. The key is to abstractly “depersonalize” God (as we in the Judeo-Christian tradition are inclined to imagine Him) for the moment, for the sake of argument, and call Him instead simply “the greatest possible being.” If we imagine such a maximally great being (MGB) – a term I’ll discuss below – we should admit that part of MGB's greatness entails its existence, since any other being, even one less great, would be ontologically “superior” to MGB by sheer virtue of actually possessing its iota of greatness. No matter how great MGB might be, it is less great than any actually existent thing because there is no MGB to have that greatness. Hence, if we conceive of God as the greatest possible being – and not even the greatest possible mortally imaginable being – then we should admit such a greatest being would have to exist, or it would not eo ipso be the greatest being. A thing’s nonexistence automatically disqualifies it from being the greatest possible being (a.k.a. God).

This sounds reasonable enough, but where Kant finds fault with it – and, in many people’s opinion, irrefutable fault – is that existence is not a predicate, meaning nothing can “have” existence to any greater degree than anything else. Predication basically means description. For example, you can predicate redness to two apples. One apple might be much redder than the other. By some obscure chemical process, you might be able to remove all redness from one of the apples, thus, no longer being able to predicate redness to it. That’s nice for things – predicates – like color, size, temperature, beauty, wisdom, weight and the like; but what about existence? Does it make any sense to predicate being to a thing? How can one thing be "more existent" than another? Existence is a binary feature of ontology: you either have it or you don’t. Existence is also very egalitarian: each existing thing exists just as greatly as any other existent thing. Hence, speaking of comparative degrees of being seems totally fallacious: a thing either is or it isn’t.

The significance of this point for OAG, according to Kant, is that existence shouldn’t even be brought into the consideration of God’s alleged greatness. Since there is no such thing as maximally existing or minimally existing -- but only existence in se -- we should not bother saying an existent thing is greater than a non-existent thing because it exists. We should restrict our criteria of ontological “greatness” to legitimate, variable predicates.

In his well known example [forgive me if I have this confused], Kant proposed two triangles, one existent, the other non-existent. "Adding" existence to the latter in no way increases its greatness over the former. However, removing one of the former’s sides – thus eradicating its triangularity – would indeed place its ontological "rank" below the nonexistent, whole triangle. The point is that this change – removing a side from one of the triangle – alters the triangles' respective predicates. This change has nothing to do with their existence, since existence is not predicable to them, while triangularity is predicable.

As a consequence, God’s putative greatness can be maximalized only in terms of His predicates – goodness, wisdom, power, etc. – but not in terms of His existence. Hence, even if He were conceived of as some “greatest being” – the most perfectly predicated being – that still would not entail His actual existence. A perfect and benevolent God may be better than a wicked Deity in terms of their predicates, but since you couldn’t predicate existence to either, perhaps neither exists. God may be perfect, but He nevertheless may not be. Basically, Kant imprisoned God in the realm of pure abstract possibility, granting even His Maximal Greatness only the possibility of existence.

Now, I am far from an expert on these matters, but I do know there has been some very advanced analytic work on Anselm’s little argument in the past fifty or sixty years. His humble thought experiment, which was conceived in Anselm's mind more as a prayer than a real argument, has died so hard -– or, perhaps more accurately, not died at all -– for some good reason. I understand that Alvin Plantinga’s most important contribution to OAG has been to cast it in terms of possible worlds (PWs).

For the sake of technical clarity, I should mention that Plantinga defines PWs as “maximal states of affairs,” meaning they can possess every possible attribute except being our world, the actual world (AW). Our world, AW, is the only actual state of affairs; everything else is possible. AW exists actually; all PWs exist possibly. There are other states of affairs that could exist, and could be very different from AW. For example, there could be a PW, say, PW’, that is exactly like AW except that in PW’ I have brown eyes instead of blue. Or there could be another PW, say PW’’, that is totally unlike AW, except that PW’’ also has oxygen (and only oxygen) in it. A PW’s relative similarity to AW is not really important. What is important is that, no matter how similar or dissimilar a PW is to AW, none of them actually exists, whereas AW does actually exist. Notice that this stipulation is not a violation of the non-predicability of existence, of which Kant made so much. It’s simply a statement about whether a state of affairs is or is not actual. Both PW’ and PW’’ are merely possible worlds, not the actual (i.e., this) world.

Now, basically – very basically! – Plantinga argues that if there is a possible PW in which some being necessarily exists, then that being necessarily exists in all PWs, including our world, AW. Hence, this alleged necessary being (a.k.a. God), posited in PW1, necessarily exists in this, our actual world. I admit that when I first encountered this idea, it struck me as utterly absurd. So what if some great being exists in some other possible world? Its “possible” existence in some other “possible” world doesn’t make it the case that it really exists in this, the real world. I was baffled, but I also knew Plantinga – and Anselm, and Hartshorne, and many other philosophers sympathetic to OAG – was not a complete idiot. He had to have some good reason for defending the seemingly trivial idea of a possible necessary being.

I have come to realize that Plantinga's good reason is that the nature of a necessary being (NB) is crucial for a fair assessment of OAG. In my very dim understanding, when Plantinga speaks of NB, he means a being that can’t fail to be, a being the essence – or “non-negotiable” attribute – of which is to exist. In purely abstract terms, it simply can’t be the case that NB does not exist. Now, such a being is certainly conceivable. Take a minute and imagine such a being, using only the terms of the definition I just gave; don’t get hung up on how foreign such a being is to our everyday, contingent existence. If you can imagine such a being without immediately recognizing an absolute contradiction in terms – on the order of a round square or 2 + 2 = 5 – then such a being might very well exist in some other, possible state of affairs.

In addition to the necessity of God in OAG, Plantinga discusses the "greatness" of this alleged NB. Plantinga begins by imagining a being that is "maximally excellent," which is to say a being that is totally loving, kind, wise, powerful, just, good, etc. As Graham Oppy says, "an entity possesses 'maximal excellence' iff [if and only if -- EBB] it is omnipotent, omnscient [sic], and morally perfect." A maximally excellent being (MEB) is certainly possible.

Next, MEB can be said to possess “maximal greatness” iff it is maximally excellent in every PW, which is to say it exists necessarily and is necessarily maximally excellent. In other words, MEB is MGB iff it is NB and necessarily MEB. Such a being, too, is certainly possible. We’ve already noted the possibility of NB, the possibility of MEB, and, now the possibility of MGB, so we should now see the consequent necessity of MGB. Since it is possible that MGB exists in some PW, it is necessary that MGB exists in every PW, including AW.

Maybe this seems like a lot of smoke and mirrors. To help clear the air, let’s consider the general nature of logical necessity and possibility. Consider the following proposition (q): 2 + 2 = 4. We immediately, intuitively and undeniably recognize that q is true, and necessarily true, in AW. But how about in some other world, say, non-AW? It should be obvious that q is also just as necessarily true in non-AW as in AW. q is a necessarily true proposition in every possible world. It can’t fail to be the case that 2 + 2 = 4; it’s a sure bet, in AW and in all PWs. By the very terms of the definition, it’s meaningless to rub your chin and say, “Well, perhaps 2 + 2 doesn’t equal 4 in some other world. Anything's possible, right?” Sorry, nope. Whatever is necessary in one state of affairs is the case in every PW; and whatever is the case in every PW is necessary in any state of affairs.

Now let’s consider a second proposition (qq): 2 + 2 = 1. It should be as plainly obvious – and necessarily obvious – that this is nonsense, both in AW and in all PWs. qq is necessarily false, which means it is – or "exists as" – false in every PW. It's as absurd to rub your chin and say, "Well, maybe 2 + 2 really does equal 1 in some other world. Anything's possible, right?" Well, no. You'll lose that bet every time, in every world.

With all this in mind, the important question is this: is God, construed as an NB, possible like q or impossible like qq? If God is like q, then He exists in PW1 just as necessarily as in AW. If, on the other hand, God is like qq, then He cannot exist in PW1 just as necessarily as He can't exist in any other world, including AW. But as we’ve already seen, God, as an NB, is not necessarily impossible; in fact, an NB’s possible existence ensures its existence in every world just as necessarily as q’s necessity ensures its validity in every world.

What is possible is possible in every world. What is possibly necessary in one PW is just as totally necessary as in every PW. More poetically, what is possibly necessary is necessarily possible; and what is necessarily possible is simply necessary. As with q, the very terms of the definition of an NB make it meaningless to rub your chin and say, “Well, perhaps a 'necessary being' could fail to exist in some other sate of affairs.” You might as well say 2 + 2 could equal 1, or that a square could be round. Since God, as an NB, is possible in at least one PW, then He necessarily exists in every PW, including AW. Hence, God necessarily exists in AW.

It should be clear that, in light of PWs, I don’t find Kant’s critique of OAG compelling. I agree that existence is not a predicate; but I disagree that OAG is simply about God’s predicates. OAG, more properly understood, is about various PWs’ predicates. As Kant argued, we shouldn't try to weigh one Being's greatness over another's based on their existence or non-existence. And now, with OAG formulated in terms of PWs, we’re not trying to establish one PW’s greatness over another PW based on either’s existence or non-existence. We’re simply discussing what might “populate” PWs.

Hence, I think Kant’s disregard for God’s predicates misses the point and actually backfires on him. He’s right that we should not judge God’s greatness based on His existence or non-existence. Yet, the predicate under question is not God’s existence, but his necessity or contingency. We’re not debating whether God’s be-ing is more truly existent than any other things’ be-ing. We’re simply discussing his mode, or way, of existing in comparison with other things’ modes of being. And, as I’ve discussed, God’s (conceptualized) status as an NB gives Him the same kind of omni-PW necessity as q or the triangularity of three conjoined sides.

The OAG, then, is not about God’s predicates, but about God’s predicability to various PWs. That difference, I think, dodges Kant’s predication objection to OAG. I could care less whether a possible God is greater than an actual one. What I’m concerned with – and what I think the OAG presses upon us – is whether it’s possible to predicate to a PW the attribute of “having an NB.” Insofar as such PW-predicability is possible, I think OAG provides a powerful defense of God’s existence, in PW and in AW.

AWmen.

P.S. I would appreciate suggestions, critiques, encouragement and any credit card information you might have to offer. Thankee.

P.P.S. I’m suspicious of calling location a predicate. Hence, insofar as I can see God’s possibility in PWs as a reference to His location, I think OAG has even less to do with God’s predicates.