Sketches on Divine Simplicity: II

Stump, Kretzmann and Ross in their brief debate shared a commitment to Aquinas' account of divine simplicity; they do not discuss or clarify but rather assume the doctrine as background for their discussions. Hasker, I think, is relatively uncommited to Aquinas' notion of simplicity, and as a critic of Stump and Kretzmann, lacks Ross' interest in explaining how divine freedom and simplicity might be compatible.

O'Connor (Faith and Philosophy, v. 16, n3, '99) was not a party to the debate I recounted in Pt. I, but he seems sympathetic--at least--to Aquinas' account of simplicity, and his article "Simplicity and Creation" takes up what roughly amounts to Stump and Kretzmann's problem, that "simplicity is incompatible with the possibility that God might have created another world (or not have created at all)" (O'Connor, 406). O'Connor's strategy is to give an account of divine agency that is not committed to, and does entirely without appeal to, intentions. The result is a rather unintuitive account of divine agency, one that does not look much like human agency at all. That may be a welcome result to one already expecting "agency" to be equivocal or merely analogous in some way between applications to God and creatures. Others expecting univocity are more likely to be bewildered.

I. The Problem
In contemporary theory, "the causal theory of action" is most often appealed to in explaining our acts. A bodily behavior, B, is an act of an agent, A, only if B is approriately a causal consequence of factors including A's reasons, R. We may say A controls his action, B, on account of the fact A's reasons for doing B are causally efficacious. If R does not cause A's action, B, then B is not intentional from A's point of view--it would be like a twitch or a heartbeat: not truly an action. The reason is here treated as part of a mechanism: B is triggered by A's having R, et alia (406-7).

Divine action is not to be seen as fitting the causal theory's model, as an account of divine agency is constrained by the contingency of God's creative action; the causal theory would render the creation necessary. Rather, God is pictured as freely bringing about an intention, X, to create a world that is not caused by any group of reasons, R. God has a purpose constituting reason(s) to create, R, and he recognizes that world W1 would--say--uniquely satisfy R, but this does not cause an intention to create W1.

But then we are faced with the problem: yes, we have a rough account of divine agency that respects divine freedom. It could be filled out a bit more by appeal to O'Connor's theory of agent causation, but that is another rather controversial matter; let's stick with the controversy we brought to the dance. The account O'Connor sketches here implies God has a contingent intention, X: "[i]f God's willing this particular order of things is contingent, then it might have been different. This contradicts the absolute simplicity thesis, as it implies that an aspect of God's intrinsic nature in this world--his state of intention that our world obtain--might have been different, though God himself would still have existed...." (408).

II. O'Connor's Solution
O'Connor begins to give a solution by recounting a rather popular medieval doctrine present in Aquinas' work: God cannot be really related to a creature, even in creation. "If God is His utterly simple nature, then, in Himself, He must be utterly unchanging across all possible circumstances" (408); for "the creation relation" is "merely notional for God" (408). It seems to me that Stump and Kretzmann in their article did not take this part of Aquinas' teaching seriously; their work would have been rather different if they did. In this respect, O'Connor seems closer to the undiluted Thomas, for whatever that is worth.

For as a consequence of Thomas' denial of real relations between God and creatures there cannot be "anything like a state of intention whose intrinsic nature would have varied in accordance with the state of affairs intended" (408), as having contingent intentions would imply a real relation running from God to creatures. Thus, O'Connor suggests we amend the model of divine agency above so as to omit X, God's intention. That is, we are not to see a contingent intention--or a necessary intention--or any kind of intention whatsoever come between God's reasons, R, and the created world. That is, "God doesn't form an intention to create our world and consequently do so, He creates the world directly. His activity entirely consists in a causal relation between himself, who is unchanging, and the dependent, contingent reality" (408).

III. An Objection and a Reply
Well hang on just a minute, a skeptic might say, "this move makes the cognitive aspect of God'a agency mysteriously alien, to the point where we've lost our intuitive grip on the model" of divine agency (408-9). Surely "at some point God has to decide, to intend, to will, or what have you that this particular creation be actualized" (409).

At first blush, yes, the amended model of divine agency looks rather mysterious with its picture of direct creation without any intervening intentions. But look closer, O'Connor suggests, and you will see the original model, over which we still had some intuitive grip, does the same type of thing the amended model does. It posits a direct action without an intention intervening between it and the agent's reasons. In fact, it is quite difficult to see how any theory of action could avoid positing such direct action somewhere, if only tacitly.

The original model posited an act, the production of an intention X, for which God had reasons R, but which itself had no separate intention Y. That is, the production of X was direct in exactly the contested way. Thus, "[w]hatever mystery resides in conceiving the dependency between God and creation in the intention-less, simpicity-based model is perhaps equally present between agent and intention in the original model" (409). But if direct action was no mystery in the original model, consistency requires it not become a mystery in the amended model. As long as we admit direct action, what need have we for intervening intentions? Indeed, "the locus of agency is not withing some special kind of intrinsic state" like an intention, "but rather with the exercise of causal power bringing about a state" (409). Thus, and here is the big payoff: "Had God created differently, the only difference would be in the contingent order, not in God" (410).

IV. Caveat Lector
O'Connor's picture takes a little getting used to, no? I am not sure that his picture works out, despite his protest that it does not imply God is unable to exercise will: "my model has indicated nothing about God's intrinsic, simple nature other than its embodying reasons for creating each of those contingent orders that are possible" (409). That is, for all O'Connor has said about divine agency, we could still be justified in speaking of divine willing. However, "it will not be or involve a distinctive intrinsic state of God" say like an intention (410). God "is not other than his simple, intrinsic state"; his willing cannot be like human willing--God's willing "will not involve such an internal change" wherein an intention is produced disitinct from other cognitive states (410). Instead, think "God's willing this world is his being its creator--which, intrinsically,is nothing other than his essence willing himself" (410).

Well, that last quote is unfortunate, I think: it makes him look as if he merely shifted the problem around, away from contingent intentions to something else. I do not think he meant to look that way, but now we may ask about God's being a creator, which we now hear is a matter intrinsic to God. That is, God creating is intrinsic to God, and cannot be "written off" as a one-way relation to God on the part of creation. But if the act of God's creating is contingent--and Aquinas surely holds that it is in some sense--then does it not follow that contingency is written into God's very essence? Although we no longer need worry about contingent intentions in God--very well--we now must worry about the contingency of creation as implying contingency in God's essence. Maybe O'Connor means to tacitly acknowledge this difficulty in his closing discussion of Kretzmann (410-11); there O'Connor seems to tentatively entertain the notion that God's act of creation is not contingent, that is, that "God who is perfect goodness will naturally create...." (410).

Another less pressing difficulty: you might take O'Connor's reasons for denying intentions to God as doing much more: implying there are no such things, even for humans. He is aware of the problem, I think, for he addresses it in giving reaons why humans must have intentions in willing: "The having of separate, content-bearing states of intention for each action is a way of harnessing such causal mechanisms within our bodies and environment, often triggering into action mechanistically-encoded, latent action routines...and always guiding the completion of behavior through complex feedback mechanisms" (409)--but of course, God does not have a body.

O'Connor does not see his amended model of divine agency as a magic bullet, slaying all dangers to a Thomist account of divine simplicity; he closes by noting a continuing difficulty. "Even if I have been successful, an important worry remains about the apparent complexity of reason states on the second model" (411)--i.e. do not take his amended model as final. It is better, I think, to take his work as a suggestion whose adoption would require modification and elaboration, but which might well at least initially seem intriguing for a thomist.