I maintain that there are modes of being. To be precise, I maintain that it is intelligible that there be modes of being. This puts me at odds with those, like van Inwagen, who consider the idea unintelligible and rooted in an elementary mistake:
. . . the thick conception of being is founded on the mistake of transferring what properly belongs to the nature of a chair -- or of a human being or of a universal or of God -- to the being of the chair. (Ontology, Identity, and Modality, Cambridge 2001, p. 4)
To clarify the issue let's consider God and creatures. God exists. Socrates exists. God and Socrates differ in their natures. For example, Socrates is ignorant of many things, and he knows it; God is ignorant of nothing. God is unlimited in power; Socrates is not. And so on. So far van Inwagen will agree. But I take a further step: God and Socrates differ in the way they exist: they differ in their mode of being. So I make a three-fold distinction among the being (existence) of x, the nature (quiddity, whatness) of x, and the mode of being of x. At most, van Inwagen makes a two-fold distinction between the being of x and the nature of x. For me, God and Socrates differ quidditatively and existentially whereas for van Inwagen they differ only quidditatively (in respect of their natures).
One difference between God and Socrates is that God does not depend on anything for his existence while Socrates and indeed everything other than God depends on God for his/its existence, and indeed, at every time at which he/it exists. I claim that that this is a difference in mode of existence: God exists-independently while creatures exist-dependently. There would be an adequate rebuttal of my claim if thin translations could be provided of the two independent clauses of the initial sentence of this paragraph. By a thin translation of a sentence I mean a sentence that is logically equivalent to the target sentence but does not contain 'exist(s) or cognates or 'is' used existentially. Translations are easy to provide, but I will question whether they are adequate. Let 'D' be a predicate constant standing for the dyadic predicate ' --- depends for its existence on ___.' And let 'g' be an individual constant denoting God.
1. God does not depend on anything for his existence
1-t. (x)~Dgx.
2. Everything other than God depends on God for its existence
2-t. (x)[(~(x = g) --> Dxg].
I will now argue that these thin translations are not adequate.
I begin with the obvious point that the domain of the bound variable 'x' is a domain of existent objects, not of Meinongian nonexistent objects. It is also obvious that the thin translations presuppose that each of these existents exists in the same sense of 'exists' and that no one of them differs from any other of them in respect of mode of existence. Call this the three-fold presupposition.
Now consider the second translation, (2-t) above. It rests on the three-fold presupposition, and it states that each of these existents, except God, stands in the relation D to God. But this is incoherent since there cannot be a plurality of existents -- 'existent' applying univocally to all of them -- if each existent except God depends on God for its existence. It ought to be obvious that if Socrates depends on God for his very existence at every moment, then he cannot exist in the same way that God exists.
I don't deny that there is a sense of 'exists' that applies univocally to God and Socrates. This is the sense captured by the particular quantifier. Something is (identically) God, and something else is (identically) Socrates. 'Is identical to something' applies univocally to God and Socrates. My point, however, is that the x to which God is identical exists in a different way than the y to which Socrates is identical. That 'is identical to something' applies univocally to both God and Socrates is obviously consistent with God and Socrates existing in different ways.
Here is another way to see the point. To translate the target sentences into QuineSpeak one has to treat the presumably sui generis relation of existential dependence of creatures on God as if it were an ordinary external relation. But such ordinary relations presuppose for their obtaining the existence of their relata. But surely, if Socrates is dependent on God for his very existence, then his existence cannot be a presupposition of his standing in the sui generis relation to God of existential dependence. He cannot already (logically speaking) exist if his very existence derives from God.
The point could be put as follows. The Quinean logic presupposes ontological pluralism which consists of the following theses: everything exists; there is a plurality of existents; each existent exists in the same sense of 'exists.' Ontological pluralism, however, is incompatible with classical theism according to which each thing distinct from God derives its existence from God. On classical theism, everything other than God exists-derivatively and only God exists-underivatively.
On the Quinean scheme of ontological pluralism, the only way to connect existents is via relations that presuppose the existence of their relata. So the relation of existential dependence that is part and parcel of the notion of divine creation must be misconstrued by the Quinean ontological pluralist as a relation that presupposes the logically antecedent existence of both God and creatures.
The ontology presupposed by Quine's logic is incompatible with the theism van Inwagen espouses. One cannot make sense of classical theism without a doctrine of modes of being. One cannot be a classical theist and a thin theorist.
Bill,
You have introduced the QuineSpeak predicate letter 'D' to translate the English predicate '---depends for its existence on---'. So, if 'a' is the QuineSpeak name for Abraham, say, and 'g' for God, then the QuineSpeak wff 'Dag' translates 'Abraham depends for his existence on God'. The QuineSpeak is no more than a compressed version of the English. Then you go on to say
This is where you lose me. Why have we moved from talk of predicates into talk of relations?Posted by: David Brightly | Friday, August 03, 2012 at 03:55 PM
Bill,
I enjoy the blog. First-time commenting. I apologize if the following is asinine, it just occurred to me and it seemed interesting to me. Still not totally sure if it will pan out.
Suppose that van Inwagen concedes that we cannot make sense of classical theism as you have construed it, but he can make sense of a closely related position; classical theism*.
He might insist that 1-t is false, for God bears the dependence relation to himself. He could either (a) say that 1-t is an incorrect translation of 1, on the grounds that the context tells us that what is meant is that God does not depend on anything besides himself for existence. Or he could also say (b) that 1 is false, but what we can affirm is 1*: God does not depend on anything besides himself for his existence.
Either way the translation that he wants to end up with is:
1-t*. (x) [~(x=g) --> ~Dgx]
we could then affirm the following:
3-t (∃x) Dgx
Also, because of the sameness of antecedent we can easily combine 1-t* with 2-t to sum up the classical theists dependency thesis:
4-t (x) [~(x=g) --> ~Dgx • Dxg]
Anyways, because classical theism* affirms 3-t, everything that exists bears the existential dependence relation to something, so we need not posit additional modes of being to makes sense of the view. Further we can still make sense of the derivative and the underivative notion, by saying that something exists derivatively if it bears the dependence relation to something besides itself, and underivately if it bears the relation only to itself. We could also take a page out of mereology and define 'proper dependency' as follows: x properly depends on y if and only if x bears the dependency relation to y and y bears does not bear the dependency relation to x.
I don't think that proper dependency could be used to generate the same kind of problem that you raised in your post, since it is just special type of the general dependency relation, the intuitive force of it requiring modes of being would be lessened. At least, it seems to me.
Perhaps you will insist that this is simply not the view you are interested in entertaining. But then, perhaps van Inwagen would insist that this view is close enough to do the trick, so to speak. The trick being making sense of the intuitions and motivations that generate the classical theists position.
Thoughts?
Posted by: CGibbs | Friday, August 03, 2012 at 06:22 PM
C Gibbs,
Excellent comments! Thanks.
I originally entertained the thought that God is not dependent on anything distinct from himself for his existence, but then I realized that a stronger claim is warranted: God is not dependent on anything (or at least anything concrete) for his existence. (And if God is simple then he cannot be dependent on his attributes, which are poresumably abstract objects.) For we certainly don't want to say that God causes his own existence. To cause his own existence, he would have to exist 'before' (logically speaking) he exists. God cannot be metaphysically prior to himself. If we think of divine creation as a relation, then it is an asymmetrical and irreflexive (not just nonreflexive) relation: if g creates x, then it is not the case that g creates g.
So I would maintain that (1-t) is correct.
More later.
Posted by: Bill Vallicella | Saturday, August 04, 2012 at 05:34 AM
C G,
A question: if God bears the existential dependence relation to himself, then wouldn't Socrates bear it to Socrates, and similarly for every creature? But if it makes a wee bit of sense to say that God bootstraps himself into existence, surely it makes no sense to say the same of Socrates.
Posted by: Bill Vallicella | Saturday, August 04, 2012 at 10:25 AM
David,
Predicates can be 1-place, 2-place, n-place. By deleting the name from 'Socrates is wise' we get the predicate '___ is wise' and by deleting the names (and replacing them with slots for names) from 'Socrates is taller than Crito' we get the 2-place predicate '___ is taller than ---.'
Properties are what is expressed by 1-place predicates and relations are what is expressed by n-place predicates where n > 1.
Posted by: Bill Vallicella | Saturday, August 04, 2012 at 10:55 AM
Thanks, Bill. A couple of further questions, if I may.
1. Don't the QuineSpeak predicate 'D' and the ordinary English predicate '---is dependent for its existence on---' express the same relation? If so, doesn't the 'presupposition' problem from which you say the QuineSpeak suffers also transfer to the English? Or is there a subtle difference between the ways that we are to interpret QuineSpeak and English?
2. I don't understand the presupposition problem. What goes wrong with the following presentation? Initially, just God exists, so the relation expressed by 'D' is empty, ie, there does not exist x, y, such that Dxy. Subsequently, God ('g' in QS) creates and sustains Abraham ('a' in QS). So now Dag obtains and obtains uniquely, ie, Dxy --> x=a & y=g. We have to say that the relation denoted by 'D' has changed, or at least, its extension has changed. It is as if we had two diagrams. In the first a blob labelled 'g' stands majestically alone. In the second a blob labelled 'a' is added together with an arrow from the 'a' blob to the 'g' blob labelled 'D'. I cannot see where the notion of 'presupposition' or 'logical alreadiness' comes in.
Posted by: David Brightly | Tuesday, August 07, 2012 at 03:02 AM