Beyond metaphysical immutability

For years I was convinced that the extrinsic constitution model of divine knowledge, which theists who accept divine simplicity must accept, solves the problem of divine immutability in an A-theoretic world where truth changes. The idea was that God’s knowledge of contingent facts is constituted by God’s unchanging essential features (which given simplicity are God himself) together with the changing contingent realities that God knows, that God’s gaze extends to. (This idea is not original to me. Aquinas already had it and probably many contemporary people have independently found it.)

But I now think that this was too quick. For let’s take the idea seriously. The point of the idea is that an unchanging God can have changing knowledge. But now notice that God’s knowledge is conscious. The language of “God’s gaze” that I used above (and which Boethius also uses in his famous discussion of divine knowledge of free actions) itself suggests this—God sees the changing reality. At one time God sees Adam sinless. At another time God sees Adam sinful. This is a difference in conscious state. Granted, this difference in conscious state is entirely metaphysically constituted by the changing reality. But it still means that God’s conscious state changes. It changes in virtue of its extrinsic constituent, but it is still true that God at t₁ is conscious of one thing and God at t₂ is conscious of something else instead. And I submit that that is incompatible with divine immutability.

I think there are two responses the classical theist who believes in changing truths can give. The first is to deny that God is conscious of the changes. I think this is unacceptable. The more vivid and the more vision-like knowledge is, the more perfect it is. The idea that God has merely unconscious knowledge of contingents does not do justice to the perfection of omniscience.

The second response is to bite the bullet and say that God’s conscious state changes but this is compatible with immutability as long as this does not involve an intrinsic change in God. I think this is untenable. That God’s conscious state does not change is, I think, a central part of the content of immutability, regardless of whether this conscious state is intrinsically or extrinsically constituted. For a non-physical being, change of conscious mental state is a paradigmatically central kind of change—regardless of the metaphysics of how that change of conscious state comes about. When God says in Malachi 3:6 that he does not change, it seems very implausible to think that the listener is supposed to say: “Sure, but sometimes God has one conscious state and sometimes another, and because this change is grounded extrinsically, that’s OK.” Malachi isn’t doing heavy-duty scholastic/analytic metaphysics. Similarly, when the early Church Fathers say that God is unchanging I doubt they would tolerate the idea that God’s conscious state changes. The extrinsic constitution story is an explanation of what makes God’s conscious state change, and I expect the Church Fathers wouldn’t have cared what the explanation would be—they would just deny the change.

Jumping from the Church Fathers to the modern period, Calvin says that God “cannot be touched with repentance, and his heart cannot undergo changes. To imagine such a thing would be impiety.” But if God’s conscious states are extrinsically constituted and can change, there would be nothing to prevent the idea of God’s “heart” undergoing changes: when people behave well, God feels pleased; when people behave badly and deserve vengeance, God feels vengeful. The differences in God’s feeling would be, one could imagine, constituted by the differences in human behavior and divine response to it. But it would be implausible to think that Calvin would say “Well, as long as the change is extrinsically constituted, it’s OK.” We then wouldn’t need Calvin’s famous story—itself going back to the Church Fathers—of the accommodation of divine speech to our needs. When Calvin insists that God’s heart cannot undergo changes, he isn’t just concerned about divine metaphysics. He is rightly concerned about a picture of a God with a changing mental life. And here at least, Calvin is with the mainstream of the Christian tradition.

If I am right in the above, there is a disanalogy between how God’s mental state behaves across possible worlds and across times. We have to say that in different possible worlds God has different (extrinsically constituted according to divine simplicity) conscious states. But we cannot say that God has different conscious states at different times.

Some thinkers, especially open theists, want the doctrine of divine immutability not to be about metaphysics but about the constancy of God’s character, purposes and promises. I think they are wrong: the doctrine of immutability really does include what we might call metaphysical immutability, that God has no intrinsic change. But metaphysical immutability is not enough. A mental and especially conscious immutability is also central to how we understand divine immutability.

And this is not compatible with the A-theory of time, given omniscience. Which is too bad. While I myself am a B-theorist, the reasoning in yesterday’s post was giving me the hope that we could detach the A- and B-theoretic debate from theism, so that the theist wouldn’t need to take a stand on it. But, alas, I think a stand needs to be taken.

God's timelessness, the A-theory of time, and two kinds of Cambridge change

Classical theism holds that God is timeless and knows all objective truths. According to A-theories of time, objective truths change (e.g., what exists simpliciter changes on presentism, and on other A-theories at least what time is objectively present changes). There is a prima facie conflict here, which leads some classical theists to reject the A-theory of time.

But there is also a widely accepted reply. Classical theism also holds that God is simple. One of the consequences of divine simplicity is that if God had created a different world, he wouldn’t have been any different intrinsically—and yet he would know something different, namely that he created that world rather than this one. Seemingly the only good solution to this problem is to suppose that God’s knowledge is in part extrinsically constituted—that facts about what God knows about contingent things are partly constituted by these contingent things.

But the same move seems to save timelessness and the A-theory. For if God’s knowledge is partly extrinsically constituted, then as the created world objectively changes, as the A-theory holds, God’s knowledge can change without any intrinsic change in God. Basically, the change of God’s knowledge is only a Cambridge change in God—a purely relational change.

I have always been pulled two ways here. Since I accepted divine simplicity, the response seemed right. But it also seemed right to think there is a tension between God’s timelessness and the A-theory of time, thereby yielding an argument against the A-theory.

I haven’t settled this entirely to my satisfaction, but I now think there may well be an argument from classical theism against the A-theory.

First, note that the extrinsic constitution move is aimed not specifically at a tension between God’s timelessness and the A-theory, but at a tension between God’s immutability and the A-theory. The move shows how an immutable being could have changing knowledge, because of extrinsic constitution. But while any timeless being is immutable, the other implication need not hold: timelessness is a stronger condition than immutability, and hence there could be a tension between divine timelessness and the A-theory even if there isn’t a tension between immutability and the A-theory.

Here is why I see a tension. The crucial concept here is of a merely relational change, a Cambridge change. The most common example of a Cambridge change is something like:

1. Bob became shorter than his daughter Alice.

Here, we’re not supposed to think that Bob changed intrinsically, but simply that Alice got taller!

But there is another kind of change that I used to lump in with (1):

2. Dinosaurs became beloved of children around the world.

Both are, I suppose, Cambridge changes. But they are crucially different. The difference comes from the fact that in (1), the change is between the slightly younger Bob being taller than Alice was then and the slightly older bob being sorter than Alice was then. While the change was due to Alice’s growth, rather than Bob’s shrinkage, nonetheless it is crucial to this kind of Cambridge change that we be comparing the subject at t₁, considered relationally, with the subject at t₂, again considered relationally. It is, say, the 2018 Bob who is taller than Alice, while it is the 2023 Bob who is shorter than Alice. I will call this kind of thing strong Cambridge change.

But when dinosaurs become beloved of children around the world, as they did over the course of the 20th century, this wasn’t a change between earlier and later dinosaurs. Indeed, the dinosaurs were no longer around when this Cambridge change happened. I will call this kind of thing weak Cambridge change.

Strong Cambridge change requires an object to at least persist through time: to be one way (relationally) at one time and another way (again, relationally) at another. Weak Cambridge change does not require even that. One can have weak Cambridge change of an object that exists only for an instant (think of an instantaneous event that becomes notorious).

A timeless being can “undergo” weak Cambridge change, but not strong Cambridge change. And I suspect that change in knowledge, even when the knowledge is extrinsically constituted, is strong Cambridge change.

Here is a piece of evidence for this thesis. Knowledge for us is partly extrinsically constituted—if only because (I am grateful to Christopher Tomaszewski for this decisive point) what we know has to be true, and truths is typically extrinsic to us! But now suppose that I have a case where the only thing lacking to knowledge is truth—I have a belief that is justified in the right way, but it just happens not to be true. Now suppose that at noon the thing I believe comes to be true (here we are assuming the A-theory). If we set up the case right, I come to know the thing at noon, though the change is a strong Cambridge change. But suppose that at noon I also cease to exist. Then I don’t come to know the thing! To come to know something, I would have to persist from not knowing to knowing. Prior to noon I was such that if the thing were true, I’d know it, but the thing isn’t true. After noon, I don’t know the thing, even though it isn’t true, because I don’t exist after noon. Change in extrinsically constituted knowledge seems to be at least strong Cambridge change.

Further, think about this. When God knows p in one world and not-p in another, this transworld difference is a difference between how God is in the one world and how God is in the other world, even if it is a relational difference. Similarly, we would expect that if God changes from knowing p at t₁ to knowing not-p at t₂, God exists at t₁ and also at t₂. And this does not seem to fit with God’s timelessness. (But don’t classical theists say God is omnipresent, and shouldn’t that include omnitemporal presence? Yes, but omnitemporal presence is not omnitemporal existence.)

In other words, I think for God to change in knowledge in lockstep with the objective facts changing, God has to exist in lockstep with these objective facts. To change from knowing to not knowing some fact due to the change in these facts, one needs to be a contemporary of these changing facts. And a timeless being is not (except should there be an Incarnation) a contemporary of anything.

In summary: A timeless being can only undergo weak Cambridge change, while it is strong Cambridge change that would be needed to maintain knowledge through a change in objective truths, even if that change is extrinsically constituted. One can uphold the A-theory with a changeless God, but not, I think, a timeless God.

Or so I suspect, but I am far from sure, because the distinction between weak and strong Cambridge change is still a bit vague for me.

And even if my specific arguments about God aren't right, I think the weak/strong Cambridge change distinction is worth thinking about.

Anthropomorphism about God

Consider an anthropomorphic picture of God that some non-classical theists have:

1. God is not simple, and in particular God’s beliefs are proper parts of God.

2. God’s beliefs change as the reality they are about changes.

Putting these together, it follows that:

3. I can bring about the destruction of a part of God.

How? Easy. I am now sitting, and God knows that. So, a part of God is the belief that I am sitting. But I can destroy that belief of God’s by standing up! For as soon as I stand up, the belief that I am sitting will no longer exist. But on the view in question, God’s beliefs are parts of him. So by standing up, I would bring it about that a part of God doesn’t exist.

But (3) is as absurd as can be.

And of course by standing up, I bring it about that a new divine belief exists. So:

4. I can bring about the genesis of a part of God.

Which is really absurd, too.

Why I can't believe in a God other than of classical theism

I can’t get myself to believe in a God who is an old bearded guy in the sky. That would be just a fairy tale.

What’s wrong with such a concept of God? It’s the beard! Seriously, the problem is that a guy who has a beard has parts and changing. Whether the parts are material or immaterial does not seem of very deep metaphysical significance. But having parts or changing, either one of these is an absurd anthropomorphism.

And hence I can’t get myself to believe in a God who changes or has parts. That leaves classical theism and atheism as the options. And atheism leads to scepticism, I think.

A problem for some views of a temporal God

Among those who think that God is in time, there are two views:

1. God has existed for an infinite amount of time

2. God came into time a finite amount of time ago when he created the world.

The second view is held by William Lane Craig. On this view, God isn’t essentially temporal: he wouldn’t have been in time if he didn’t create time or temporal beings.

It’s occurred to me that those who accept the first view have the serious problem of getting out of the time-of-creation problems: Why did God create the world when he did, instead of earlier or later? And why did he wait an infinite amount of time before creating?

St Augustine’s answer that time starts with creation doesn’t work for those who accept (1).

Supposing creation itself to be omnitemporally eternal only solves the problem with (1) if one additionally accepts a relationalist B-theory of time. For otherwise there is still the question why God’s omnitemporally eternal creation process isn’t all shifted temporally by a year forward or backward in time.

If classical theism rules out open theism, then classical theism rules out presentism

If presentism and most, if not all, other versions of the A-theory are true, then propositions change in truth value. For instance, on presentism, in the time of the dinosaurs it was not true that horses exist, but now it is true; on growing block, ten years ago the year 2019 wasn't at the leading edge of reality, but now it is. The following argument seems to show that such views are incompatible with classical theism.

1. God never comes to know anything.

2. If at t₁, x doesn’t know a proposition p but at t₂ > t₁, x knows p, then x comes to know p.

3. If propositions change in truth value, then there are times t₁ < t₂ and a proposition p such that p is not true at t₁ and p is true at t₂.

4. It is always the case that God knows every true proposition.

5. It is never the case that anyone knows any proposition that isn’t true.

6. So, if propositions change in truth value, then there are times t₁ < t₂ and a proposition p such that God doesn’t know p at t₁ but God does know p at t₂. (by 3-5)

7. So, if propositions change in truth value, God comes to know something. (by 2 and 6)

8. So, propositions do not change in truth value. (by 1 and 7)

I think the only controversial proposition is (1). Of course, some non-classical theists—say, open theists—will deny (1). But non-classical theists aren’t the target of the argument.

However, there is a way for classical theists to try to get out of (1) as well. They could say that the content of God’s knowledge changes, even though God and God’s act of knowing are unchanging. The move would be like this. We classical theists accept divine simplicity, and hence hold that God would not have been intrinsically any different had he created otherwise than he did. But had God created otherwise than he did, the content of his knowledge would have been different (since God knows what he creates). So the content of God’s knowledge needs to be partially constituted by created reality. (This could be a radical semantic externalism, say.) Thus, had God created otherwise than he did, God (and his act of knowledge which is identical to God) would have been merely extrinsically different.

But exactly the same move allows one to reconcile the denial of (1) with immutability. The content of God’s knowledge is partially constituted by created reality, and hence as created reality changes, the content of God’s knowledge changes, but the change in God is merely extrinsic, like a mother’s change from being taller than her daughter to being shorter than her daughter solely due to her daughter’s growth.

I agree that denying (1) is compatible with God’s being intrinsically unchanging. For a long time I thought that this observation destroyed the argument (1)-(8). But I now think not. For I am now thinking that even if (1) is compatible with immutability, (1) is a part of classical theism. For it is a part of classical theism that God doesn’t learn in any way, and coming to know is a kind of learning.

Here is one way to see that (1) is a part of classical theism. Classical theists want to reject any open theist views. But here is one open theist view, probably the best one. The future is open and propositions reporting what people will freely do tomorrow are now either false or neither-true-nor-false, but tomorrow they come to be true. An omniscient being knows all true propositions, but it is no shortcoming of omniscience to fail to know propositions that aren’t true. Then, our open theist says, God learns these propositions as soon as they become true. This is all that omniscience calls for.

Now, classical theists will want to reject this open theist view on the grounds of its violating immutability. But they cannot do so if they themselves reject (1). For the presentist (say) classical theist can reject (1) without violating immutability, so can our open theist. Indeed, our open theists can say exactly the same thing I suggested earlier: God changes extrinsically as time progresses, and the content of God’s knowledge changes, but God remains intrinsically the same.

So, what do I think the classical theist should say to our open theist? I think this: that God doesn’t come to know is not just a consequence of the doctrine of immutability, but is itself a part of the doctrine of immutability. A God who learns is mutable in an objectionable way even if this learning is not an intrinsic change in God. But if we say this, then of course we are committed to (1), and we cannot be presentists or accept any other of the theories of time on which propositions change in truth value.

I think the best response on the part of the classical theist who is an entrenched presentist would be to deny (1) and concede that classical theism does not rule out open theism. Instead, open theism is ruled out by divine revelation, and revelation here adds to classical theism. But it seems very strange to say that classical theism does not rule out open theism.

Simplicity and divine decisions

One of the most difficult problems for divine simplicity are how to square it with creation and divine knowledge of free actions. On its face, there are at least four distinct states of God:

1. God's essential nature
2. God's contingent decisions
3. God's knowledge of his contingent decisions
4. God's knowledge of creatures' free responses to his contingent decisions.

Calvinists can reduce (4) to (3) (say, by grounding (4) in (3), and holding that if state B is grounded in state A, that does not really multiply states in a way contrary to divine simplicity), thereby reducing the number of distinct states from four to three. Thomists, and presumably some Calvinists as well, can reduce (3) to (2): God's decision is identical with his knowledge of his decision. Even if we make both of these controversial moves, we still have the distinction between God's essential nature and his contingent decisions (which are then identical with his knowledge of the decisions and his knowledge of creatures' responses thereto).

My own preferred sketch of a solution to these problems is here. The solution proceeds by making the contingent aspects of (2)-(4) be extrinsic to God.

For those Christians who are unimpressed by the strength of the traditional commitments (in the pre-Reformation tradition, but also in people like Calvin and Turretin) to divine simplicity, and the arguments for divine simplicity, the natural solution will appear to be to deny divine simplicity, and then not worry about the problem.

They should still worry about the problem. For if one denies divine simplicity and holds that God has at least the two distinct constituents: his essential nature, N, and his contingent decisions, D, then one has to say something about the relationship between these two. Clearly, D is in some way explained by N: God acts as he does in part because of his essentially perfectly good character. The explanation is not a grounding-type explanation—to make it be a grounding-type explanation would be to hold on to a version of a divine simplicity explanation. In creatures, the corresponding explanation of decisions would be causal: the character causes (deterministically or not) the decision. So it seems that we have something very much like a causal relationship between N and D. And this in turn makes D be very much like a creature, indeed perhaps literally a creature. Since D is a constituent of God, it follows that a constituent of God is very much like a creature, perhaps literally a creature. But this surely contradicts transcendence!

Now perhaps one can insist that the relationship between N and D while being akin to causation is sufficiently different from it that D is sufficiently different from a creature that we have no violation of transcendence. Maybe, but I am still worried.

So if I am right, even if one denies divine simplicity, a version of the problem remains. And so the problem may not be a problem specifically for divine simplicity.

Immutability and split brains

The traditional Christian view that God is unchanging has been accused of being a fruit of Greek ideals of perfection (and what's wrong with that?). Here I want to motivate this view by thinking about our mental life.

But our conscious states are divided between times in much the way that the two centers of consciousness of a split-brain patient are divided from each other. My present state of consciousness only includes shadowy reminders of what I was aware of five minutes ago and vague premonitions of what I am about to be aware of. My temporality makes me like a patient split into untold numbers of centers of consciousness associated with different times (perhaps in a continuous way, with overlapping between close-by centers, since many of our mental states themselves persist over short amounts of time). We are deeply internally disunited--our "transcendental unity of apperception" is quite limited. Such deep internal division and disunion is surely not what the perfect being would experience (at least not in his proper nature—an Incarnation might make for such an experience, and the above reflection should make us grateful that he took up this deeply divided existence for our sake). This is not a matter of some "Greek ideal" of perfection. It is simply the intuition that mental division within oneself is an imperfection.

The above argument presupposes eternalism. But presentism only introduces even greater limitation in our mental life by making the future and past conscious states not be ours.

So we have good reason to think of God's mental life as all-encompassing, of God living an infinitely rich mental life all at once, as Boethius said. But God is a mind and surely all of his mental states are conscious. This gives us good reason to think God is unchanging.