By David Misialowski (2006)
In Part I of my response to Robert P. Taylor's dissertation on the philosophical problem of theological fatalism, I introduced my noisy neighbor – let's call him Sam – who assailed me with 51 consecutive iterations of "Night Fever", prompting me to murder him. At trial, I was acquitted by reason of insanity, after my lawyer bombarded the jurors with the same musical assault, 51 consecutive times. Afterwards, as I recounted, the jurors lined up to shake my hand, and one of them even asked for my autograph.
The upshot of my little tale was that the jurors had decided that I had lacked the capacity for free choice when I committed the murder, and hence was not morally culpable. The bombardment by the Bee Gees had prompted an involuntary and, as it were, instinctive response on my part, one that I could not control.
However, what we are trying to find out, in the context of the foreknowledge/free will debate, is whether any of our acts are free in the presence of an essentially omniscient agent, one that foreknows, infallibly, the truth conditions of all events, including those that occur in the future. In this context, even if the jurors had found me guilty, the theological fatalist might protest that I was foredestined to commit the crime by God's infallible foreknowledge of it – indeed, that Sam was foredestined to play the "Night Fever" 51 times in a row on that fateful New Year's Eve night.
In Part 1 of my response, I tried to show that this sort of argument fails. Briefly, it constitutes a modal fallacy, assigning necessity to a contingent event. In modal logic, events are necessary if they occur at all possible worlds, and contingent if they occur at some worlds and not at others. They are possible if they occur in at least one world. Since the modal status of events cannot change – they are logically fixed – it follows that Sam's playing the Bee Gees, and my killing him in response, were timelessly contingent events. From this it follows that God's foreknowledge did not entail these occurrences; on the contrary, the occurrences entailed God's foreknowledge.
More, I argued against the two forms of the Transfer of Necessity Principle that Robert invoked, hoping to show that the supposed accidental necessity of the past, and the alleged transfer of necessity of God's foreknowledge of an event to that event, are mistakes. The past, like the present and the future, does not have any accidental necessity about it, but rather timeless modal truth conditions involving, as always, necessity, contingency and possibility. Under this argument, God’s past infallible foreknowledge of my killing Sam was, and remains, a contingent fact of history, contingent on what Sam did, and what I did. That is the bare outline of my argument. If my argument is right, there is no problem of theological fatalism. Further considerations would seem to be moot.
Of course my argument could be wrong, but one would have to try to show where it goes wrong by a detailed analysis of it, and one would then have to present a counterargument to defeat it. In his dissertation, Robert, although initially broaching the modal fallacy, appears to believe that the transfer of necessity principle and the accidental necessity of the past are real roadblocks to resolving the supposed dilemma of making infallible foreknowledge compatible with free will. While I don't share this belief, let's look at the other solutions he proposes.
The Atemporal Solution
In the second part of his essay, Robert considers the atemporal solution. He begins by writing:
I must emphasize that we are in complete disagreement here, since I have hoped to show that A (presumably, in this context, God's infallible foreknowledge of all truth-apt future propositions) in no way entails B – in fact, the case is just the opposite. But, having said that, how would the case if be God was atemporal?
The idea here, as Robert presents it, as that the accidental necessity of God's knowledge in the past can be circumvented by appeal to God as existing either at all times, or transcending time, or in some fashion, both ways. Robert provides a series of detailed definitions for how his would be, which I will sidestep. Instead, let's look at the crux of the solution: which, as Robert writes, "posits a God who, despite having full knowledge of any event at any given point in time, lacks foreknowledge due to His being timeless and thus not located at any temporal point. ... God knows eternally what I will do at t3, but it is not the case that God knows, prior to my act at t3, what I will do at t3, since this would require God to be temporal."
This seems fair enough, but we must be careful to distinguish whether this constitutes a rebuttal of the (supposed) foreknowledge/free will dilemma, or an attempt to define it out of existence, in the same way that the problem can be defined out of existence by saying that since the future has not yet occurred, God has no knowledge of it (his omniscience being restricted to propositions that are truth-apt. An open future would have no truth-apt propositions to know).
In any case, I do not see how positing an atemporal or omnitemporal God defeats the dilemma, if there is a dilemma. An immediate problem may be found in the observation by Aquinas that Robert quotes:
In short, God sees everything at once – all the roads, and those who travel them. This introduces the problem of how God has certain knowledge of the past, present and future, and, as Robert notes, "...it is not clear how God could know which routes people will take, unless their routes are fixed."
Of course, this problem of how God infallibly knows the future, or infallibly knows anything at all, might be a problem for my solution as well - my solution holding that the entire conception of an incompatibility between foreknowledge and free will rests on a fallacy of modal logic. The various analogies that Robert offers – time as the circumference of a circle, with God as a central point within it, seeing all points in time with no single point taking precedence over any other; or God as an all-seeing point at the center of a sphere, seeing, simultaneously, all events that actually take place, as well as all events that might have been – seem, to me, unsatisfactory, for this reason: These solutions imply that the universe, including all spatial and temporal locations within it, just is, and we would have to agree with Parmenides that time, change, and physical separation are illusions: that reality is a single, unchanging, indestructible whole.
The problem is illustrated by Robert's discussion of the A and B theories of time. It is unclear to me whether very many people are familiar with these conceptions, so let me try to explain them in my own terms, and in so doing, we shall find that the ancient, pre-Socratic Parmenides seems to have been remarkably prescient. In fact, the problem is such a general one that it could be the subject of a dissertation all its own, for it could be the case that a Parmenedian conception of space and time rules out free will (rules out a lot of things!) even in the absence of an omniscient agent.
Let's return to my confrontation with the odious Sam and his Bee Gees bombardment. Initially I had presented this scenario as happening in the future – next New Year's Eve, in fact, the day before Jan. 1, 2007. But I have been speaking of the killing in the past tense – as though it had already happened, and that my trial and acquittal by reason of insanity were events of the past. In fact, talk of tensed temporal relations only makes sense under the so-called A theory of time, which is also known as presentism. Presentism holds that only the present is real. The past used to be real, but it no longer is; and the future is open.
The B theory of time does away with tensed relations, and makes the present an indexical property of existence, rather like "here" is indexical. That is, if I am in New York, then New York is "here" for me; and if you are in London, then London is "here" for you. There is no objective fact of the matter about "here". The same would be true for time, under the B theory. Robert captures the essence of this nicely when he writes: "Consequently, just as I can say 'I am listening to my iPod' and Socrates can say 'I'm learning that I know nothing', it is also true, under the B-theory, that I can say 'I am listening to my iPod now' and Socrates can say 'I'm learning now that I know nothing', where 'now' depends upon the temporal location of the speaker."
He might have added that under this conception someone in the future, relative both to the iPod-listening Robert and the knowing-nothing Socrates, can say, with equal justification, that "I am now doing x" and he would be right, also. In contrast to presentism, this conception of time is often called eternalism, and under it tensed talk of temporal relations is merely conventional or indexical. Under presentism, people and events are said to endure through time, while under eternalism they are said to perdure within it. Under the B theory, it is said that people and other objects in the universe have temporal parts, in the same way that they have spatial parts. To see how this analogy works, we could say that a person's spatial parts, on the vertical axis, are delimited by the soles of his feet and the top of his head. In a like fashion, a person's temporal parts are defined by the boundary conditions of her birth and death.
The B theory of time would explain how God comes to know all the facts of history. He could, so to say, stand outside the spacetime continuum, and look down upon it and see every place, and every when, all at once. But is the world really like that? And if it is, can free will survive in such a place, whether God exists or not?
This is a very problematic situation. The B theory, in effect, spatializes time. No one doubts that Mars (and all other points in the universe) have an independent existence, an indexical "here". If time is like that as well, then the past, present and future are all ontologically on par, and if that is the case, then how can anything I do now change a future that already seems to be fixed by virtue of the fact that it is indexically actual? In point of fact there are many eerie parallels between time and space, and in Chapter Eight of his book, Metaphysical Theories and Philosophical Constraints, Professor Norman Swartz discusses them in detail.
Why should we think that the B theory of time is correct, or even coherent? After all, this is nothing like what we actually experience. In human experience, time flows, and only the subjective present is real. Human intuition accords with the presentist, or A theory, account of time. And if the presentist account is right, free will seems maintained, and it could be argued that since the future is open (no truth conditions obtain) then God has no foreknowledge of it.
Bearing in mind that human intuition is often a poor guide to many of the apparent facts of reality (or at least the theories that we construct to interpret those facts, theories that themselves might be instrumentally useful, though not necessarily depicting reality as it actually is but instead models of it), there are two arguments against the A theory of time. The first is that the notion of time flowing from past to future is incoherent, no matter how intuitively plausible it seems. If time flows, how fast does it flow? One second per second? That makes no sense. For flowing time to make sense, it seems that its rate of passage would have to be measured against some meta-time, and then that time would have to be measured against yet another meta-time, leading to infinite regress. And the second reason that the A theory of time seems false is because of the special theory of relativity, which shows that there is no universal Now. What we call Now is relativized to reference frames, and in general two observers, depending on their relative rate of motion, may wildly disagree on what is past, present and future.
Einstein's relativity seems strongly to support a B theory conception of time, though some philosophers have disputed this.
I must admit I am not sure how Robert, in his dissertation, concludes that the atemporal solution is not compatible with the conventional understanding of the B theory of time. I happen to agree with Craig, who, as Robert notes, thinks that the atemporal solution is committed to B theory. And I think that the atemporal solution is incompatible with the A theory.
Be that as it may, if we accept the B theory of time we can go much further, philosophically, and posit a reality that is strange indeed. For example, Julian Barbour, a British theoretical physicist, believes that time does not exist, that all motion and change is an illusion, and indeed that our memories of the past are themselves illusions. We are all, he writes, timelessly existing "time capsules", and those entities that just happen to have coherently ordered illusions of an asymmetrical past (time's arrow flowing from past to future) will be those entities that have the illusion of sentience.
For our purposes, this is a digression, so let me finish up my analysis of the atemporal solution. Robert worries (I don't, for reasons already stated) that the accidental necessity of the past (i.e. God's foreknowledge) obliterates free will. He worries that:
The atemporal solution, allegedly, offers a way out of this conundrum by asserting that it is timelessly necessary that God knows I will do p. But he worries that this does not defeat the problem of accidental necessity, writing, "Instead, the scope is shifted such that the proposition that is the object of God's knowledge is accidentally necessary, rather than God's knowledge per se being accidentally necessary."
Since I don't agree with the formulation of the accidental necessity of the past or of the transfer of necessity principle, holding that both arguments arise from a failure to incorporate modal logic, I will pass over this. Next, he worries how an atemporal being could observe or interact with temporal events. This is a very justified worry, in my estimation, and one that is often at the core of atheistic objections to traditional conceptions of God. We must admit that the nation of an atemporal God also having temporal interactions could be self-contradictory and therefore incoherent, and if that is right, the atemporal solution fails for that reason alone. Robert writes: "Sorabji, for example, suggests that a timeless being cannot differentiate between 'x occurs' and 'x occurs at t' God will hence observe me typing this sentence, but my typing the following sentence will also be observable to Him."
This idea seems passing strange, and cannot, in my view, succeed. In fact, it seems self contradictory on the face of it. How could it be that an omniscient being can't even tell what time it is? If God cannot tell what time it is, then God is not omniscient.
Robert then discusses "eternal simultaneity" and "temporal simultaneity", or ET-simultaneity. This is an attempt, it would seem, to explain how an atemporal omniscient being can observe or interact with temporal events without the absurdity of suggesting that an all-knowing God can't tell what time it is. The idea here, as I understand it, is that events may be simultaneous with respect to the so-called Eternal Present (God's eye-view) but not with respect to each other. Robert writes:
I guess, by this analogy, God sees the whole reel, and we see the successive images, one following the other. But does analogizing existence to a movie reel make any sense if we wish to keep free will? Implicit in this conception is the notion that the "movie" has already been scripted.
Robert worries that under ET-simultaneity, a questions arises as to how God distinguishes what might happen from what does happen. He writes: "How does God, being atemporal, determine what will happen and what might happen, and if He is able to determine what will happen, as opposed to what might happen, then the problem raised by accidental and timeless necessity, viz. that what God knows must happen, is not defeated."
I'm not sure I understand the difficulty here, because again, by way of analogy we have just said that while we see the successive frames of the movie, one coming after the other, God sees the whole reel at a glance, as it were. Again, though, if this account is right, it implies (or so it seems to me) that we must be ontologically committed to a B theory of time, and the B theory of time, by itself, poses threats to free will that need to be addressed.
I will conclude by saying that the atemporal solution has real problems, but I believe these problems are outside the scope of the alleged foreknowledge/free will incompatibility. I personally do not see how an atemporal God can consistently interact with, or observe, temporal events. As has been noted, taking atemporality seriously can lead to bizarre conclusions, like saying that a God who knows everything that there is to know can't even tell what time it is. It seems to me that a more plausible picture (which might itself be just a restatement of atemporality, but whether it is or not, it is easier to grasp) is that God is omnitemporal, just as he is omnipresent: he is at all points in space and all points in time, simultaneously. This would explain his infallible true beliefs of all truth-apt propositions, but a problem immediately presents itself: the concept of omnitemporality seems, at least on the surface, to return us to a Newtonian conception of time and space in which time and space are fixed and absolute, a stage, as it were, on which objects exist and events play out. The radically different conception of space and time presented by Einstein does not seem to be compatible with an omnitemporal agent, but that is a subject for a different discussion.