Divine Causality and Human Freedom: a Synthetic Thomistic Approach
Barry Miller and W. Matthews Grant offer unique contributions to the challenge of how to reconcile the following commitments common to Christian belief:
1) Humans are libertarian free.
2) God is the universal cause of everything really distinct from himself.
3) God is simple and immutable.
Tensions arise from these three propositions almost immediately, some superficial, others serious. For example, there is the initial concern that if something is caused then that something cannot be free. However, even if it can be maintained that God’s transcendent causality is compatible with libertarian freedom, it then beguiles us how libertarian free human action doesn’t change God, especially God’s knowledge. In this article, I seek to summarize the contributions of Miller and Grant and show how they complement one another to strengthen the case for the genuine compatibility of the three propositions above.
First, something of a crabby introduction. For it does seem to me that metaphysical reasoning drives strongly in the direction of classical theism and divine simplicity.[1] It also does seem to me – if only by direct acquaintance – that I am truly, libertarian free. Because it strikes me that two things are actual, they must be compatible.
The point? I often worry philosophers and theologians run too quickly from theoretical tensions, losing the possibility of insight had only by working patiently through a puzzle that must have a reasonable solution, even if that solution is not immediately clear. Perhaps this is why my friend Jim has a coffee mug which reads, “Open theists are quitters.” I admit to similar feelings about theological determinists and Molinists, though I repeat myself.
Polemical remarks aside, let’s begin by summarizing W. Matthew Grant’s position.
First, we must ask: when is an act libertarian free? Grant tells us: “An act is free in the libertarian sense if and only if it its agent performs the act voluntarily and intentionally, and either the act is not determined (i.e., there is no factor both prior to and logically sufficient for the act), or the act is determined and the agent’s responsibility for the act derives from the agent’s voluntary and intentional performance of some prior act that was not determined.”
Alternatively, determinism means there are logically sufficient conditions for some effect E (which are also prior to E) such that if those logically sufficient conditions obtain then E follows inevitably. In which case we could also infer if E does not obtain then those logically sufficient prior conditions did not obtain (or were different in some respect).
So, if determinism is true, then the effect cannot be different without the prior cause(s) being different. However, as we’ll see in Grant’s model, if divine simplicity is true, then prior cause(s) are not different, even if the effect is different. Therefore, determinism isn’t true, or at least not necessarily true, if Grant’s extrinsic model is correct.
Grant eliminates the threat of determinism with respect to God’s causing human actions by illustrating how the extrinsic model of divine agency does not introduce any factor that is both prior to and logically sufficient for any of God’s effects. For example,
In a world where God doesn’t create there is just:
1) God (with God’s reasons)
In a world where God does create, there is:
1) God (with God’s reasons)
2) The effect (E)
3) The causal-dependence relation between God and E
Since God is not determined by any of his reasons to create and because God’s action is 1) basic and 2) God’s action is in the patient[2], the conditions of determinism are avoided with the extrinsic model, with the added result of pushing all contingency extrinsic to God. For example, God (and God’s reasons) are prior to E but not logically sufficient for E, since God need not have acted on whatever reasons may have “impressed” him. E is logically sufficient for E but is not prior to E (and with the production of E the causal-dependence relation, so that is not prior either). Thus, while libertarian agency hasn’t been directly defended, it has been indirectly defended by showing the conditions of determinism do not hold.
As Grant also points out, the extrinsic model is simply the result of common scholastic commitments, including divine simplicity and Aristotelian action theory. Since God is immutable, God does not change in bringing about his effect.[3] Thus, God’s acting on any of God’s reasons just is the effect coming about, in which case we can say that God “choosing E” or God “intending E” is nothing over and above E coming to exist with its causal-dependence relation on God. Atop the prospective benefit of preserving libertarian freedom, this model of divine agency is slim and elegant.
Contrast the extrinsic model what Grant calls the “intrinsic model”, where if God creates, there is:
1) God
2) E
3) God’s choice, decree, or intention to bring about E, which is intrinsic to God, is that in virtue of which God causes E and which would not exist were God not causing E.[4]
4) The causal-dependence relation between God and E.
Grant argues the intrinsic model invites determinism, since it introduces a factor (namely, item 4) which would be both prior to and sufficient for the effect. For example, God forming some prior intention or decree (considered as something intrinsic to God) in virtue of which God then produces some effect would clearly be prior to (ontologically, if not temporally) and logically sufficient for that effect. Thus, things could not have been otherwise, including human action. So, for example: God’s Internal Decree A = Your Action E. But then if not your action E, then not God’s Internal Decree A (God is different). Determinism holds.
However, looking back at Grant’s extrinsic model, your action could have been different (or not at all) and God been absolutely the same. Because God does not change internally or essentially to create – the change is entirely extrinsic to God – this means God is not both prior to and logically sufficient for any created effect. This is a striking result: it shows, at least initially, that libertarian freedom and classical theism are not only not incompatible, but the latter may be the only plausible theistic model for securing the former. Whether other models of theism can offer the same resources I leave to further research.
“In short,” says, Grant, “None of the items on the scene when God causes A constitutes a factor both prior to and logically sufficient for A. But, then, on EM, God’s causing A does not render A determined. What goes for A goes for any creaturely act. Given EM, God can cause all creaturely acts without rendering any of them determined. Thus, on EM, the only way a creaturely act caused by God would be prevented from being free in the libertarian sense is if God’s causing such an act precluded its creaturely agent’s performing the act voluntarily and intentionally. But I know of no good reason to think that an act’s being caused by God rules out its creaturely agent’s performing the act willingly and on purpose. So, given EM, it seems that all human acts can be caused by God and still satisfy both the strict and broad conditions for being free in the libertarian sense.”[5]
Grant’s model has come in for criticism, of course the most pressing of which is how it maintains divine immutability in light of contingent creation (libertarian agency aside). A recent iteration being Ryan Mullin’s aloneness argument, which suggests God’s knowledge must change with and without creation.[6] Fortunately, Grant and Pawl provide a cogent response, showing the impressive flexibility of their model for handling these concerns.[7]We will return to this issue toward the end of this article.
Barry Miller’s Approach
Barry Miller – himself a strong defender of divine simplicity – begins to solve the puzzle of human freedom and divine universal causality by reminding us that God’s causality is to be understood analogously to creaturely causality. Minimally, when creatures act as cause, they act always on pre-existing material. Their causal acts are limited by something other than the causal act itself – namely, something already existent. But this is not the case with God, who creates ex-nihilo. This is significant. Why? Because it shows that God’s causal act never acts on something (since it would have to exist before God created it, which is absurd) in the way creaturely causal acts do. Rather, God’s causal act is why there is something that itself can act or be acted on (as a secondary cause) by anything else at all.
Miller gives the example of Socrates moving his leg, and says, “… in affirming that Socrates does act on something (his leg), my interest is in making one point and one point only, namely, that the condition of the leg plays a part in determining the outcome of Socrates’ causal activity.”[8]
Again, the point being that creature causality is limited by factors independent of the causing, which is not the case with God, whose causality is subject to no such limitation given creation ex-nihilo.
In explicating Miller’s work, Elmar Kremer urges the use of external causal operations to express divine causality, as in,
a. God causes that (the earth spins on its axis)
And
God causes that (the trees in the orchard bloom),
And not the use of the internal operators, as in,
A’. God causes the earth to spin
And
b. God causes the trees in the orchard to bloom.”
The latter two examples imply (falsely) that God acts on the earth or the tree. By contrast, the former two are consistent with the fact that God’s causing is ex nihilo throughout.
Here’s the upshot. If God causes everything that happens in the universe, then God not only causes that things change in various ways, God also causes that (creatures cause things to change in various ways). Thus, God not only causes that (the trees blossom in the spring), but he also causes that (sunlight makes the trees blossom in the Spring). If anything is caused to happen by a creature, the creaturely causing is among the effects of God’s act of causation.
Moreover, as Miller explains, God not only produces but sustains the universe in one single eternal act; hence, God does not act on anything at any point. Or as Miller puts it, “There can be no multiplicity of divine acts at all, whether sequential or not. Rather, what we have is one act, the effect of which encompasses the whole history of the Universe, and in the production of which God does not act on anything.”[9]
For Miller, it’s also important to understand that God’s “acting upon” a creature doesn’t mean creatures are passive to God in the sense of waiting to receive something from God by which they can then act. Because every creature depends entirely on God’s action for its very existence, no creature can receive their action from God in any sense that would imply the creature is already there. Every creature for its very existence owes to the same divine causality by which their actions exist as well.
Granted, there is mystery to the notion of God’s causality, but mystery is not incoherence, especially if we already have reason to affirm something is the case, which, in this instance we do: traditional metaphysics tells us God grants existence to everything apart from himself, he doesn’t just rearrange things in existence. If that’s true, then God is not a cause like we are causes. God’s causality is, as Barry Miller puts it, the limit case cause.
Perhaps the best we can do to understand this analogous causal relation is by deploying… well, an analogy.
It has been suggested that God’s causal relation to the world should not be seen as just another agent in the world. Rather, God is more like the author of a story, rather than an actor in it. Edward Feser deploys this analogy helpfully as follows,
“Consider once again the analogy with the author of a story. Suppose it is a crime novel and that one of the characters carefully plots the murder of another, for financial gain. We would naturally say that he commits the murder of his own free will, and is therefore justly punished after being caught at the end of the novel. It would be silly to say, ‘Well, he didn’t really commit the murder of his own free will. For he committed it only because the author of the story wrote it that way.’ The author’s writing of the story the way he did is not inconsistent with the character’s having freely committed the murder. It’s not comparable to (say) some further character in the story hypnotizing the murdered and thereby getting him to commit the crime – something which would be inconsistent with the murder having been committed freely. If we got to a point in the book where such hypnotism was revealed, we would say, ‘Ah, so it wasn’t an act of free will after all.’ But we don’t say that when we reflect on the fact that the story had an author. It is perfectly coherent to say that the author wrote a story in which someone freely chooses to commit a murder.”
What Feser is illustrating is like the points Miller (and to various extends) Grant have made. Given that God’s causality is transcendent, God causing a story where we freely commit wrongful acts does not entail we were determined to perform those wrongful acts. God caused that [some sinful act occurred] is consistent with that sinful act not occurring and God being intrinsically the same.
If Grant and Miller are correct, then that person who freely chose to commit a murder could have done otherwise, in which case God would then have simply caused a world (story) in which that person freely chose not to be a murderer. And while it is true that God’s knowledge of that world would be different, such changing knowledge is a mere Cambridge, rather than real, change, which is compatible with divine simplicity.
The latter point must be substantiated, which this brings us to the final consideration of how God’s knowledge can be different, including God’s knowledge of the actions of libertarian agents, while God remains simple and immutable. This is critical to evaluate, since the security of libertarian freedom on this account depends upon divine simplicity, which many consider threatened by God’s knowledge of contingent things, which libertarian human action would be.
In an obvious sense God’s knowledge must be different between a world where Socrates freely decides to plant an apple tree and a world where Socrates freely decides to plant an orange tree. The question then becomes how this position is consistent with a God that is intrinsically the same across all possible worlds, as classical theism holds. As indicated, the solution comes by way of differentiating between real and Cambridge change and continuing to push all contingency – including contingent knowledge – extrinsic to God.
Here is the executive summary: If only God exists then God knows himself perfectly, and in God’s self-knowledge is embedded knowledge of all the ways anything else could participate in existence. God in knowing himself also knows – in one single, simple, self-reflective insight – all the reasons it might be good to create things other than himself, though he is free to refrain from doing so.
Moreover, God knows, in knowing himself, that nothing else could exist unless God causes it. Thus, it is not that God knows “there are no unicorns.” if God has not created unicorns; rather, it is that in his eternal self-knowledge God comprehends the nature of anything else that could be instantiated and included in that knowledge is the fact that such natures are 1) contingent and 2) could never exist apart from God’s causal influx. Such knowledge will always be true, invariant across possible worlds, if you like. Then, should God decide to create, the only difference is that whatever God decides to create is created and exists with a causal dependence relation on God.
But notice this: the only difference on the ontological scene between the world where God doesn’t create and world where God does create is the created object with its causal dependence relation on God, in which case God remains really and intrinsically the same across all possible worlds. Mutatis mutandis with God’s knowledge, for on Grant’s model God knowing some created contingent reality just is that created contingent reality existing because of God’s causing it for a reason (which is enough to say God acts intentionally). In other words, God knows the contingent reality in the intentional causal act of bringing it about (which is distinct from God).
As Grant explains, “If God knows contingent objects in the act of intentionally bringing them about and if, as EM maintains, God’s intentionally bringing an entity about does not involve anything intrinsic to God that would not exist were God not bringing that entity about, then God can know contingent objects without there being any intrinsic accidents in God.”[11]
At this point it should be added that, at least according to Thomistic psychology, to know a thing is to possess the form of a thing objectively. But God understands one object – namely, the divine essence itself – and through that all the ways anything could participate in existence. Thus, if God had not created, it is not as if God knows, say, propositionally and individually, “there are no zebras, there are no elephants, there are no humans, etc.” Rather what God knows is that God is the fullness of existence as such and that nothing else could exist unless he causes it (whether the form of zebra, elephant, human, etc), because part and parcel of God’s understanding of zebra (or any reality distinct from God) is the fact that no such reality could be real in the first place unless God brings it into existence. Further, because there is no form to possess objectively relating to the proposition “there are no zebras, etc.” there is accordingly no contingent accident that could be placed in God had God not created; instead, there is only the form of zebra, which God knows analogically through knowing himself. Then, should God create a zebra, the only difference is on the side of the created object: that is, the only difference is there is now some zebra existing with its casual-dependence relation on God, since God’s intentional action is 1) basic, 2) immediate, and 3) in the patient. So, God remains intrinsically the same whether God creates this, that, or nothing at all.
[1] For a good summary of arguments see Edward Feser Five Proofs of the Existence of God.
[2] As Grant explains, “Presupposed by EM is the idea that the causal act of an agent – what the agent is doing when it is causing something – may consist solely in the agent’s causing or brining about some effect, and that an agent’s causing nor bringing about some effect may consist in nothing more than the causal-dependence relation to the agent. … This understanding of causal action is not limited to theology. It is the operative understanding of causal action by at least some contemporary agent-causal theorists, and it is the understanding of causal action defended by Aristotle within a purely natural context and adopted by the scholastic tradition influenced by him.” Free Will and God’s Universal Causality, Pg. 58.
[3] It would be a fallacy of accident to assert otherwise. What is essential to being a cause is producing some effect; it is only accidental that most of the causes of our experience change intrinsically while doing so.
[4] Of course, one wonders what advantage the intrinsic model holds upon analysis, anyway. For one could ask what causes God’s forming the intention He does, which is that in virtue of which God produces some effect. Presumably, the proponent of the intrinsic model will say it is just a basic action of God. But if so, why bother with the unnecessary middle step? Why not just make God’s creating or intending or choosing our world the basic action?
[5] Free Will and God’s Universal Causality, Matthews Grant, pg. 61.
[6] The Aloneness Argument https://philpapers.org/rec/SCHTAA-53
[7] The Aloneness Argument Fails https://www.cambridge.org/core/journals/religious-studies/article/abs/aloneness-argument-fails/D07FC36BF50F8308335192D738D11706
[8] A Most Unlikely God, pg. 130.
[9] A Most Unlikely God, pg. 130.
[10] See https://www.theschooloftheology.org/posts/essay/barry-miller-divine-simplicity, “Miller’s analysis turns on a novel way of thinking about analogy, based on the distinction between a limit case and a limit simpliciter. Consider, for example, a series of circles of ever-decreasing radius. If there is such a thing, the limit simpliciter of such a series would be a circle with a very small radius indeed, approaching 0. But the limit case—and there certainly is such a thing here—is a “circle” whose radius = 0, which is of course not a circle at all, but a point. Or consider another—more controversial, but also more apt—example, of a series of regular polygons of increasingly many sides. The limit simplicter of such a series would be a polygon of very many sides, even an infinite number of sides apeirogon. But what if we were to go beyond an apeirogon: Miller contends that we would get a circle, the limit case of the series of n-sided regular polygons. In other words, the limit case of a series is that to which the series points, and not a member of the series itself. However, it may be sensible to speak analogically of limit cases in terms of the series that point to them.”
[11] Free will and God’s Universal Causality