Divine Atemporality and Tensed Facts
March 14, 2006 — 15:39

Author: Jeremy Pierce  Category: Concept of God Divine Foreknowledge  Comments: 14

In the comments on Kevin’s post Divine Eternity and Libertarian Free Will, several commenters have raised the problem of how an atemporal God can know what time it is. There are several directions someone can go with this question, and I wanted to comment on it in a little more detail than seemed right in a comment thread. The main problem is as follows.
1. God is omniscient, i.e. God knows every true proposition.
2. God is atemporal, i.e. God does not experience events in temporal succession but instead experiences every temporal event timelessly.
3. The A-theory of time is correct, i.e. there is a fact beyond the facts about what happens before and after other events, namely the fact about which of those events is taking place now.
4. By 1 and 3, God knows what time it is.
5. If God knows what time it is, then God is in time.
6. By 4 and 5, God is in time.
7. By 2 and 6, we have a contradiction. So something above must be denied.


I don’t know of anyone who rejects omniscience to avoid this argument. Most people attracted to divine atemporality are doing so because they think it helps explain how God can know what is future to us. Most people who want to deny omniscience have probably already rejected 2.
It’s easy to avoid the argument by thinking of God as temporal. If God experiences events in temporal succession the way we do, then God can know what time it is because God experiences time as it passes. Since this argument is generally presented as an argument against the timeless conception of God, and I want to explore what atemporalists can say about this problem, I’m not going to pursue that option.
What seems to me to be the best view is just to deny that there’s anything special about tense. If you listed all the facts about what takes place in the world across all time, and you placed them in order indicating which are earlier and which are later, the A-theory of time says that there’s something more to be said. We still don’t know which time is now. There’s something special about now. You could know the whole temporal order from start to finish and which things take place when without knowing what time along the whole timeline is the present.
The B-theory says that such a claim is nonsense. There is no such further fact. Once you list all the facts about what takes place and what order they take place in, you’ve listed all the facts. If you ask what time it is now, then all you’re asking is what time it is at the time you’re asking. You’re using the word ‘now’ in the same way we use ‘here’. When we ask if we’re in the U.S. or Canada, you get one answer if you’re south of the line and a different if you’re just north of the line. Similarly, you get different answers if you ask what time it is, depending on when you ask it. It’s not as if there’s some further fact about when now is beyond the fact about when it is along the timeline that you ask the question, just as there’s no further fact about where here is beyond where you are spatially when you ask the question.
I happen to think the B-theory is correct, and I think it’s the best way to avoid the contradiction, but I don’t want to rule out a third way of avoiding the argument without examining it more closely. Several philosophers have tried to argue that you can maintain an A-theory of time while retaining God’s omniscience and atemporality without leading to the contradiction. The two I know the most about are from William Lane Craig and Brian Lefow.
Craig’s view is easier to grasp in terms of how it solves this problem. He doesn’t think of God as just atemporal. He says God exists in a parallel time dimension that has no duration, a point-dimension. It’s an eternal present. But he also thinks God is in time, so God isn’t really timeless. God is both eternal in experiencing the eternal present and temporal in experiencing succession the way we do. When speaking uncarefully, Craig says God was atemporal and then suddenly he created time and became temporal because he knew he’d have to sacrifice omniscience if he didn’t. He remains in the point-dimension called eternity while also experiencing succession in the timeline he created. This view is logically incoherent as stated, however. When he speaks more carefully, Craig doesn’t use the temporal language of God entering time suddently at the beginning of time, as if the eternal point-dimension is before time. I’ve critiqued Craig’s view in some detail in my review of God and Time: Four Views, ed. Greg Ganssle, published in Faith and Philosophy a year or two ago, but most of my problems with the view don’t relate to this issue. I do think Craig retains an A-theory and omniscience while thinking of God as in some sense timeless, but he doesn’t really think of God as timeless the way most people who use that term to descrinbe their views do. They mean not experiencing succession. He doesn’t think God is like that. So I think Craig has just denied premise 2.
The other way to avoid this argument while retaining omniscience, divine atemporality, and an A-theory of time is Brian Leftow’s. His conception of eternity includes all temporal events in eternity with God rather than having God in time with temporal events, so in that sense his view is the inverse of Craig’s. This helps him explain how God knows the order of events despite all the events’ being simultaneous with God’s eternal present, because there’s an ordering of the events in some other-than-temporal way in the timeline as it occurs in eternity. But that doesn’t help him answer the question at hand, because we’re worried about how God knows what time it is, an A-fact that is only knowable by someone in time who experiences duration and succession. There is no fact in eternity about what time it is now. Everything about time is in eternity with God except the A-facts. God just has access to the B-facts about which things are earlier or later than others (or simultaneous).
Leftow says several things about this problem, and it’s not entirely clear to me which one is his preferred view, but the general thrust of his response is to deny that the kind of omniscience traditional theism requires does not require God knowing this sort of thing. In particular, there are other kinds of knowledge that we can have that God can’t have (e.g. what it is like for a bat to be a bat, which requires being a bat), and traditional views of omniscience don’t consider those to be failures of omniscience. This is parallel to God’s inability to eat, walk, breathe, or sin, none of which threatens omnipotence on standard conceptions of omnipotence. If Leftow can get away with this (and I’m not sure I see a reason why he can’t), then I think he may have succeeded in defendng the omniscience, divine atemporality, and the A-theory of time from inconsistency. So while I myself hold to a B-theory, I don’t think it’s as clear as some do that you need to take that way out to avoid the contradiction claimed by the argument above.

Comments:
  • Christian Lee

    “but the general thrust of his response is to deny that the kind of omniscience traditional theism requires does not require God knowing this sort of thing”.
    Jeremy, I don’t think this response is nearly as strong in this context, as it “might” be in others.
    Example: God is not omnipotent because he cannot make a rock to heavy for him to lift.
    Response: Duh. Omnipotence is restricted to possible abilities.
    Example: God is not omniscient because he doesn’t know what it is like to enjoy torturing an infant.
    Response: Duh. God’s perfection restricts the things he can enjoy and hence know.
    Example: God isn’t omniscient because he doesn’t know what it’s like to be bat.
    Response: Duh. It isn’t logically possible for anything, but a bat, to know what it’s like to be a bat. And no bat is a perfect being.
    Now, take the temporal case:
    Example: God isn’t omnsicient because he cannot possibly know what time it is now.
    Response: Duh? Well, what explanation follows? If God is atemporal, then it is not logically possible for God to know what time it is?
    Anyway, the rest of the “inabilities” can, in some intuitive sense, be justified by talking about perfection. But not so with knowing the time. Atemporality does not seem like a perfect making property at all, considered alone, and especially doesn’t if it were to function to restrict God’s knowledge.

    March 14, 2006 — 15:32
  • If God is essentially atemporal, then how is that different from God being essentially not a bat? The issue is what’s compossible with God’s nature. It’s not compossible with God’s nature that God sin, experience time, or be a bat. How are these not analogous cases?
    Many (most?) atemporalists do argue that atemporality is a perfection. That’s even the primary argument for atemporality for some people. I get the sense that it’s one of the main arguments for Leftow. The idea is that a perfect being wouldn’t be limited by time in the same way that a perfect being isn’t limited by space. Greg Ganssle discusses this argument in his introduction to the aforementioned book.

    March 14, 2006 — 16:14
  • Christian Lee

    A parody argument:
    First, “If God is essentially atemporal, then how is that different from God being essentially not a bat? The issue is what’s compossible with God’s nature. It’s not compossible with God’s nature that God sin, experience time, or be a bat. How are these not analogous cases?”
    Second, If God is essentially omniscient, then how is that different from God being essentially not a bat? The issue is what’s compossible with God’s nature. It’s not compossible with God’s nature that God sin, experience time, or be a bat. How are these not analogous cases?
    And again, many people who think that God is omniscient think that omniscience is a perfection. That may be a primary argument for the claim that God is omniscient…
    As you can tell I don’t think the response works. It is much more clear that a perfect being would be omniscient than atemporal, in fact, I don’t see how existing at all times limits a being anymore than not existing at any time.

    March 14, 2006 — 16:25
  • jon kvanvig

    I think there is an easier route to the compatibility of eternality and knowing what time it is now, at least if you think of things having essences. Take whatever account you are going to accept for how God knows what I know when I know I am tired, and extend it to temporal indexicals. My preferred route is through a triadic theory of belief, according to which there is, in addition to a person and a proposition, a mode of presentation for the proposition. God doesn’t need to know that JK is tired under the same mode of presentation as I know it, for that would require that he be me. Just so, if the present moment has an essence, then an eternal God can know what we know when we know it is now Tuesday, even though he can’t know it under the same mode of presentation.

    March 14, 2006 — 16:56
  • Christian Lee

    Jon,
    I think “I am tired” expresses a proposition that God cannot know unless, per impossible, I am God.
    I don’t think pure temporal indexicals have MOPs. They are paradigms of directly referential expressions.
    Now, God may know “Christian is tired” but that is a different proposition from “I am tired”. We know this because we could report one without reporting the other. There is also a rather well developed account of how this is so in Kaplan’s “Demonstratives”. A quick appeal to authority.
    But even if that’s all wrong, how do MOPs help. If you’re right, then God has indexical knowledge after all. I worry that is sufficient for God to be in time, i.e premise 5.

    March 14, 2006 — 17:46
  • Christian, Leftow does think God is omniscient. He doesn’t think God’s omniscience need require being absolutely omniscient about every proposition. At least that’s one solution he offers. I think he also considers something like what Jon is offering, where the propositions themselves are something that can be known regardless of one’s perspective. Either way, he’s not thinking of omniscience as requiring knowing what time it is in the way we do.
    This seems much less revisionist to me as an understanding of what omniscience requires than even the claim that there are no truths about future contingents for God to know. It’s technically easier to fit such a view of omniscience into standard definitions, simply by denying alethic settledness, but the scope of God’s knowledge is much less revisionist on the sort of view Leftow is offering than on any open theist view, and thus I think his view is closer to the traditional concept of omniscience. If there’s a choice between having one perfection over the other, it seems clear to me that exhaustive knowledge of future contingents is a higher perfection than knowing A-facts, and if the most perfect being could have just one of them then it would be exhaustive knowledge of what to us are future contingents. (This all assumes there are A-facts, of course, which I deny to begin with, which makes the choice much easier.)
    As for the argument that existing at all times isn’t a limit, would you say the same about existing at all places? Are you denying the standard view of God as non-spatial? If so, then I understand what you’re up to, but I’m suspicious of this move if you think of God as non-spatial. Even if time is fundamentally unlike space, it seems to me that existing at all times and existing at all places should be equally more of a perfection or less of a perfection with respect to existing at no places and existing at no times.
    As for “I am tired”, you’re making Leftow’s point for him. If there’s nothing unproblematic with God not knowing propositons expressed by such indexicals, then why not with tensed indexicals? If those express propositions not expressed by he untensed ones, as you think, then your view on “I am tired” seems exactly parallel. If it doesn’t express propositions not expressed by untensed ones, as Jonathan suggests, then God knows it. Either way I think Leftow’s point stands. There’s nothing unproblematic going on here in terms of omniscience.
    Finally, if your worry just boils down to thinking that knowing these indexicals is sufficient for being in time, then atemporalists can just say that God is in time in that sense but not in time in the sense they originally meant, i.e. that God doesn’t experience succession but experiences every moment as simultaneous while knowing the full chronology that a B-series contains. If you want to play the semantic game of redefining what it means to be atemporal, then they can just choose a different word for what they originally meant, because whatever you’re meaning by it is something else.

    March 15, 2006 — 7:35
  • Tom

    Jeremy: [Craig] doesn’t think of God as just atemporal. He says God exists in a parallel time dimension that has no duration, a point-dimension. It’s an eternal present. But he also thinks God is in time, so God isn’t really timeless.
    Tom: This is inaccurate, Jeremy. Craig’s view is this–God is atemporal sans creation and temporal since creation. Craig believes that God’s temporal mode of existence is a contingent attribute of God. So sans creation God is absolutely timeless. But with the act of creation Craig believes God abandons this atemporal existence for a temporal mode of existence.
    Tom

    March 15, 2006 — 10:14
  • Tom, I don’t see the difference between what I said and what you said except that you used a word he can’t use, the word ‘since’. He acknowledges in more than one place that he can’t use it, and yet he continues to use it as if it’s important. It makes me think he’s trying to have it both ways. The reason he can’t use that word is because it gives the impression that for a certain length of time God is not in time and then suddenly creation occurs and God is after that in time. That’s what ‘since’ implies, since it’s a temporal term, but he can’t have that, because there is nothing before creation and no duration for timeless eternity. He acknowledges this, but he doesn’t want to if his language is any indication.
    You are right that he says that God is timeless without respect to time but temporal with respect to time. But isn’t that trivially true of anything in time? I think I’m timeless without respect to time but temporal with respect to time. He clearly means more than that. Every time I read him, I think he just sounds confused. The only way to make his view consistent is to use some of the things he says to describe the view the way I did and then to ignore the temporal language like “since creation” and “all of sudden, when creation occurs”. Maybe that’s unfaithful to him, but I’d rather be charitable to him than faithful to his language. If you’d rather be uncharitable and attribute to him a logically incoherent view, then please do, but I’m not going to do that, especially given his own statements that such language isn’t strictly speaking accurate.

    March 15, 2006 — 10:57
  • Christian Lee

    Jeremy, I can’t help but think “huge” methodological questions must be addressed for you and I to sit at this table together. Suppose I argue: Omniscience is required for God to exist, and omniscience requires knowing every knowable proposition, including the present time. So, an atemporal God is impossible. You respond: Omnisciences doesn’t require that, just knowing every knowable proposition consistent with God’s nature, which includes his atemporality. So, God could be atemporal. I respond: But that is not omniscience. You respond: It is omniscience, just not what “you” mean by omniscience.
    Well, I have the feeling that if God cannot know what time it is, what it’s like to be a bat, what will happen tomorrow, what I will freely do, what it’s like to learn, etc…all these things, then God isn’t omniscient. Or, if you like, God might be schmoniscient, and it just isn’t clear to me whether (a) anybody in the universe ever meant that but a few philosophers, (b) whether a being like that is perfect or deserving of concern.
    You are right about the future contingents, that omniscience doesn’t require knowledge of them is even harder to swallow. So much for omniscience I say.
    And about space, I wouldn’t think God to exist everywhere. That’s weird. But I could see reasons to think that claim is very different from existing at all times, for example, worries about colocation or having corruptible and bad proper parts.
    And “I” wouldn’t say I’m making Leftows point for him. I would argue that not only does God not know what time it is, he doesn’t know that “I am tired” and that is just another reason to think he is not omniscient. Again, the point is just to make explicit all those kinds of propositions that the being couldn’t know, and then ask: Is that an omniscient being?
    And my worry about God being in time is that people say things like “God is in time in that sense but not in time in the sense they originally meant, i.e. that God doesn’t experience succession but experiences every moment as simultaneous while knowing the full chronology that a B-series contains.”
    I don’t understand that at all. I think to be in time is occupy time or have a part that occupies time, or something basic. Here, time is like space, to be in space is to have part that occupies space. Anyway, assuming God doesn’t have parts and that he is immaterial and that he has thoughts, the only way it would seem we could argue about whether God is really in time is to argue about what kinds of thoughts he can have. I suppose I just don’t see how a being could think “This is the present moment” but not be in time. Suppose I am wrong, God can think that, but not think This moment is before, this moment, is before this moment…” Well, does it follow that God is, in some sense, not in time from that kind of consideration? I doubt it because I think my cat is in time, but I doubt he can think that thought. So, if it isn’t necessary, I don’t see how thoughts about succession make a difference.
    …By the way Jeremy, I’m enjoying this discussion and learning alot.

    March 15, 2006 — 11:09
  • jon kvanvig

    Christian, you’re stuck with the conclusion that God can’t know what I know when I know I’m tired if you are wedded to a dyadic theory of belief. Such theories have a really hard time with the cognitive significance of sentences involving proper names, since they have to resort to strange semantic maneuvers to solve Frege’s puzzle about identities between names. In addition, there are other reasons to be suspicious of the idea that you can’t know what I know when I know I’m tired, reasons I explore in the second chapter of my book on omniscience. Here’s a quick preview of one discomfort for the view: my wife says, “I have a dentist appointment tomorrow,” and I, being tired of having been told this information for the umpteenth time, reply hostilely, “I know that!” I’m right; I already knew what she told me.

    March 15, 2006 — 13:52
  • Christian Lee

    Jon,
    Perhaps I should check out your book! Cool!
    I don’t think the Fregean Puzzle’s make a difference here. The puzzles are raised with respect to names and definite descriptions. But, at first blush, those considerations are just not relevant to indexicals. So, I think one could accept MOPs for names, motivated by the puzzles, and then go to deny them for indexicals, because the puzzles do not apply to indexicals.
    I am also sympathetic to the appeal to linguistic intuition with the “I have a dentist appointment” example. I am not convinced, however. When youi respond “I know that!” I suggest you wrong, not right, if by ‘that’ you intend to refer to what is said by your wife. However, you are right, if by ‘that’ you intend to pick out the proposition ‘That “you” have a dentist appopintment tomorrow’. Of course, this is exactly what someone who defends the privacy of such sematic values should say.

    March 15, 2006 — 14:11
  • jon kvanvig

    Christian, yes, that is exactly what a privacy theorist should say. And, if they are honest, they’ll also admit that the response is strained, since they have to claim that I was wrong. It’s clear what the referent of ‘that’ is in “I know that.”
    On the first paragraph, I don’t think you’ll be able to handle the full range of versions of Frege’s puzzle as you suggest. In particular, the fundamental puzzle has nothing to do with descriptions, but rather with identity claims between names, and indexicals are one kind of name in such discussions.

    March 15, 2006 — 15:44
  • Tom

    Jeremy: Tom, I don’t see the difference between what I said and what you said except that you used a word he can’t use, the word ‘since’. He acknowledges in more than one place that he can’t use it, and yet he continues to use it as if it’s important.
    Tom: Just to clarify. What Craig finds properly inappropriate is talk of what was or wasn’t the case “before��? creation. It makes no sense to say God was this or that “before creation.��? To say this is to assume time before creation, something Craig denies. This is why he talks about God’s temporal status ‘sans’ creation (not ‘before’ creation). As for “since,��? there’s no problem at all in Craig’s view in speaking of this or that being the case “since creation.��? Since time comes into existence with creation, “since creation” is meaningful speech.
    Jeremy: The reason he can’t use that word is because it gives the impression that for a certain length of time God is not in time and then suddenly creation occurs and God is after that in time. That’s what ‘since’ implies, since it’s a temporal term…
    Tom: I hope you see now that this isn’t the case. To say “since creation��? is perfectly intelligible if time comes into existence with creation. So there’s no problem in speaking this way. What’s not really intelligible is talk of a “before creation��? on the same view of time. But this is why Craig prefers to speak about God ‘sans’ creation. This is a meaningful alternative to “before creation.”
    Jeremy: You are right that he says that God is timeless without respect to time but temporal with respect to time.
    Tom: That’s not Craig’s point, though. He says God is timeless ‘sans’ creation—in the absence of any creation—and temporal with creation (or since creation, which is equally meaningful). There’s no mixing of the two and no question that since creation God is temporal, period. God becomes irrevocably temporal when he creates. It’s not that he’s “timeless without respect to time��? but “temporal with respect to time.��? This doesn’t sound like anything Craig has ever or would say. Can you provide a reference where he expresses himself in precisely these terms?
    Craig is an A-theorist presentist (as am I). But his problem is that he adopts a B-theorist semantics. For example, he argues that the tenseless versions of tensed propositions are true at all times (and timelessly true ‘sans’ creation). He needs these tenseless propositions to be timelessly true because they are what inform God’s choice to create this world as opposed to any other of the logically possible worlds God could have created. What grounds those truths? Nothing, says Craig. He admits they are a metaphysical surd.
    You’re right in supposing Craig to be confused. He is confused, but not in his claim that atemporality and temporality are logically contingent and contradictory attributes. He might very well be right on this. I’m not sure. But he admits God can’t bear both properties. So God is only atemporal ‘sans’ creation and only temporal with creation.
    Tom

    March 15, 2006 — 15:55
  • Tom

    I was looking at my last sentence there and wanted to clarify. I said that for Craig God can’t bear both properties (temporality and atemporality). But of course he means God can’t bear both symultaneously.
    Tom

    March 16, 2006 — 7:39