Divine Omniscience and Human Freedom

by Donald H. Wacome, Ph.D., Professor of Philosophy, Northwestern College, Orange City, Iowa 51041

This essay originally appeared in the Review of the Evangelical Philosophical Association

My aim in this essay is to explore some conceptual difficulties that attend the doctrine of divine omniscience. Specifically, I deal with problems that arise out of the proposition that God knows everything that will happen in the future, including free human actions.

It might at first seem obvious that God simply knows everything, in the sense that we can say: "God knows that _______" and fill in the blank with anything. But a moment's reflection reveals that this is not the case; try filling in the blank with "1 + 1 = 3". God does not know that; it's impossible to know this because it is false. God's knowledge, although it is infinite, has a particular structure, constrained by logical possibility. Realizing that the doctrine of divine omniscience does not imply that God knows absolutely everything, but that God's omniscience is a matter of his knowing what it is possible for him to know, prompts us to ask whether the kinds of knowledge traditionally attributed to God are kinds of knowledge it is possible for him to have. It is widely assumed among theists that God knows everything that will happen in the future, but since the future includes free human actions, this view is problematic. Is it possible for anyone, even God, to know what free agents will do before they make their choices? In this essay I argue that we should doubt that God foreknows the actions of his free creatures.

An ancient conundrum purports to show that divine foreknowledge is incompatible with those actions being really free. Suppose today (Monday) God knows that on Friday Marvin will go to the zoo. It follows that on Monday it is true that Marvin will go to the zoo on Friday. When Friday arrives, there will be nothing Marvin can do to change what was already true on Monday. It is impossible to change the past, so there is nothing Marvin does on Friday can change the fact that on Monday God knew he would go to the zoo, and it is not, therefore, in Marvin's power to bring it about that God did not know he was going to the zoo on Friday. Thus Marvin is fated for Friday at the zoo. His choice is irrelevant to this action; with respect to it he is not a free agent. What is true for this particular action of Marvin's is true for every human action. If there is an all-knowing God, there is no human freedom. Or so the argument goes. On first hearing the argument that divine foreknowledge precludes human freedom, many people sense that something is wrong, correctly objecting that God's knowing something will happen doesn't make it happen. In general, knowing that something will occur is not equivalent to causing it to occur. Although I know the sun will rise tomorrow I do not bring it about that the sun will rise tomorrow. But the point of the argument for theological fatalism is not that God by foreknowing makes things happens, but that his foreknowing "guarantees that they will happen. If I really know that the sun will rise tomorrow, this guarantees that the sun will rise tomorrow. If someone knows something, then it is true. But there may be a connection between God's knowing what someone will do and that person being caused to do it. If we ask how God knows what the people he has created will do in the future, we may discover that the only possible ways for God to know are ways that rule out human freedom, even though foreknowledge in itself does not rule out freedom.[1]

To begin thinking about how God might know the future let's consider how we know the future. Knowledge of the future is problematic in comparison with knowledge of the past because our knowledge of what is or has been going on in the world depends on how the world causally effect us. But we cannot know about future events in this way because things that have not yet happened can have no causal influence on us. The knowledge we have about the future typically depends on our present reasoning, not on what has not yet happened. Our knowledge of the future depends on inferences we make now. Suppose, for example, I have just dropped my pen; before it hits the floor I already know it will. How do I acquire this bit of knowledge about the future? The standard answer is that we know the future in virtue of knowing the laws of nature. I know that if an object is released in a gravitational field it will fall. That is a law of nature, albeit imprecisely stated. Conjoined with my knowledge of current conditions, viz. that I recently lost hold of the pen, that it is now in a gravitational field, that there is nothing to catch it, I make an inference to the conclusion that my pen will make its way to the floor. Of course, our knowledge of the laws of nature is partial and imprecise; sometimes our beliefs about what they are mistaken. Our beliefs about relevant current conditions is also incomplete, imprecise, and subject to error. Yet if all the things that happen in the world are, like the fall of my pen, governed by causal laws and thus causally determined, then anyone with complete knowledge of the laws of nature and of the current condition of the world would be in a position to acquire perfect knowledge of the future, simply by making the requisite inferences.[2] If what happens in the world is caused and there are laws that connect causes with their effects, then God, as the creator of this law-governed world, would know the laws and the state of the world at its beginning. On this basis, he would have complete inferential knowledge of all future events. Our concern here is whether God can know future free actions of his creatures. If human actions are causally determined there is no problem with God knowing them. He would simply need to know the relevant laws of psychology or neurophysiology or whatever governs human behavior, as well as the current conditions of human beings. On this basis he would have complete knowledge arrived at inferentially, of all future human actions.

Is this the way the world is? The question of freedom and determinism raises its hoary head here, and I will not try to directly address it. But let me point out that there are good reasons to doubt whether there are, or even could be, laws governing all human action,[3] and good reasons to think that if there were such laws human action would not be free.[4] In what follows I will assume that there are no such laws and thus that, if God knows future free human actions, his knowledge is grounded in something other than his knowledge of laws governing human action.

The fact that there are no laws of this kind does not imply that what people will do in the future is totally unknowable. For instance, I know that if I offer a certain student the choice between an essay examination and a multiple-choice exam she will choose the multiple choice exam. If my knowledge is not based on my knowing a law that governs her choices in these matters then how do I know her future free action? The fact that we often know how people will act in the future does not imply that there are laws which warrant our inferences in these cases, but simply that people tend to act in regular, characteristic ways. We can say that free human beings usually act "in character". Because of this pervasive feature of human life, people's behavior is to a large extent predictable and explicable. The better I know someone, the better idea I have of what he or she will do in future situations.

In some cases, my belief about what someone will do in the future is just a hunch, a more or less educated guess. In these cases my belief is merely a belief, not knowledge, even when it turns out to be true. But in other cases my belief about someone's future action is a belief for which I have adequate reasons. When such a belief is in fact true, my belief is (barring Gettier-type cases in which my belief is justified and true, but true for the wrong reasons) a matter of knowledge. Sometimes we know people well enough to know what they will do under certain conditions. Does this mean that we sometimes know laws that govern human action? That is, that acting in character is a matter of acting under causal laws? If it is, then people would never act out of character, but they do. Human beings sometimes act in ways contrary to our most well-founded expectations. This is not always due to our ignorance. This would be so even if we knew every natural law.

This implies that while knowledge of laws would be sufficient for knowing the future it is not necessary. That there are no laws governing human action does not preclude our sometimes knowing what people will do in the future. But it does seem to preclude God's knowing what people will do in the future, at least if his knowing is at all analogous to the inferences we make based on our knowledge of persons' characters; our knowledge of future human behavior is acquired at the cost of sometimes being wrong. The risk of error appears to be a necessary feature of human belief about future free actions.

When, on Monday, on the basis of my extensive knowledge of Marvin, I justifiably predict that he will go to the circus on Friday, my belief may turn out to be true, in which case I knew on Monday that on Friday he would go to the circus. But if Marvin decides to go to the zoo instead, because as a free agent he is acting out of character, not acting as the circus aficionado I know him to be, then my belief is false and I did not know what he was going to do.

Suppose I don't like the idea of my most well-founded beliefs about what people will do in the future turning out to be mistaken and decide to make my beliefs invulnerable to being defeated in this way by refraining from forming beliefs about what people will do, and (assuming this were somehow really in my control) restrict myself to beliefs about what people are likely to do in the future. Now, instead of believing that Marvin will go to the circus on Friday, I hold the weaker belief that it is highly probable that Marvin will go to the circus on Friday. As is well known, Marvin would not falsify this belief by failing to go to the circus on Friday. His going to the zoo instead is consistent with it being true that it was highly probable that he would go to the circus.

If God only had beliefs concerning future free actions that were indefeasible in virtue of being probabilistic, the doctrine of divine omniscience would not be threatened by the fact that people do sometimes act out of character. But if God's beliefs about future free actions were all probabilistic he could not know what those actions will be. To know that p requires that one believe that p. Believing that it is very probable that p isn't enough. E.g., believing that it is likely that it will snow tomorrow is not the same thing as believing it will snow tomorrow. Someone who believes it will probably snow tomorrow need not believe it will snow tomorrow. Since knowing it will snow tomorrow requires that one believe it will, merely believing that it will probably snow tomorrow will not suffice for knowing it will.

The attempt to construe divine foreknowledge as a matter of God knowing what people probably will do engenders a dilemma: if his beliefs are probabilistic they will never be false, but he will never have knowledge. If he has beliefs about what people will do (rather than just beliefs about what they are likely to do) then occasionally his beliefs will be wrong, since free people will sometimes act in improbable ways, i.e. when they act out of character. When they do, God would not have knowledge but a false belief about his creatures' future actions. I assume that this is an unacceptable result; whether or not God knows everything about the future, he does not hold any false beliefs.[5]

Rather than looking to God's knowledge of the world and its inhabitants, perhaps the defender of divine foreknowledge needs to look to God's innate knowledge of logical possibilities. Since God knows everything there is to know about possible states of affairs ('possible worlds') it would seem to follow that he knows what every possible creature does at every moment in every possible world in which it exists. From this it would appear to follow that God knows what every actual creature will do in the future of this, the actual, world.

Consider Adam. On this account, there are any number of possible worlds God could have made actual. Some of these are worlds in which Adam exists. In some of these worlds in which he exists Adam is offered the fruit and in some of these he accepts it. God presumably knew, prior to the Fall, which of the possible worlds he made actual, and he knew what Adam does at every moment it this world, so he should also know what Adam will do when, in the future, he is offered the fruit.

Could God really have the kind of knowledge this argument requires? I do not think he could. Clearly, God knows all the possible states of affairs. He knows, for instance, all about the possible worlds in which Bob Hope becomes president of the United States in 1980, and he knows all about the possible worlds in which all chemists are plaid. God knows that "1 + 1 = 2" is true in every possible world and that "1 + 1 = 3" is true in no possible world. I think no one would deny that the creator possesses this a priori knowledge of logical possibilities. But this knowledge does not suffice for foreknowledge of free human actions in the actual world.

To show that divine a priori knowledge does not suffice for foreknowledge of free human actions, I must first offer a criterion for a human being performing an action freely. Suppose that on Friday Marvin chooses to go to the zoo. If he does this freely, then there is a possible but nonactual world which is exactly similar to the actual world up to the moment on Friday when Marvin makes his decision. From that time on the heretofore parallel worlds diverge since one contains, and one does not contain, Marvin's trip to the zoo and its consequences. The crucial notion here is that there is nothing in the actual world prior to Marvin's choice that necessitates his choosing to go to the zoo. Things could have been just the same up to the time he decides, but with a different outcome. He is free in that he could have acted differently than the way he actually acted, since there is nothing existing prior to the decision to go to the zoo sufficient to bring it about.

God knows there is a possible world in which Marvin chooses to go to the zoo and one in which he chooses not to go to the zoo. Before Marvin makes his choice, nothing distinguishes the one world from the other. Which world are we in? That is, which world is this, the actual world? One in which the future includes Marvin's trip to the zoo, or one in which it does not? To know which world we are in requires knowing the future of the actual world. Knowing everything about the actual world up to now isn't enough to tell which of the two worlds is the actual one. Nor is knowing all the possibilities: even if God knows what every possible creature does in every possible world he still won't have a basis for foreknowing free actions in the actual world.

One way to try to defend divine foreknowledge grounded in God's knowledge of logical possibilities involves the claim that the actions that a free human being does in the actual world are those he does in every possible world in which he exists. On this view, God's knowledge of what every possible creature does in every possible world would suffice for knowledge of future actions in the actual world. There is no distinction to be drawn between possible worlds in which Marvin chooses to go to the zoo and worlds in which he chooses not to; he does one thing in every world. This account will not do. It guarantees divine foreknowledge of human actions at the cost of there being no free human actions. Suppose Marvin goes to the zoo in every possible world in which he exists: this implies that his choosing to go to the zoo is a necessary feature of him; there is no possible world in which he exists and fails to go to the zoo. If there were no possible world in which Marvin does not choose to go to the zoo then we should agree that his choosing not to go is simply not a possibility, in which case he is not free to choose not to go. But, in reality, he could choose this, and he is therefore free to do so. It is reasonable to think the actions of free persons are somehow an expression of their natures or characters, but the nature of a free person does not determine his free actions. Free human agents sometimes act out of character.

If God's innate knowledge of logical possibilities cannot serve as a foundation for his knowledge of future free actions, let us consider another possibility. Recently, there has been a revival of the late-Medieval idea that God has "middle knowledge" of his creatures' free actions, i.e., he knows what subjunctive statements are true of any possible creature he might choose to actualize. Most prominently, Alvin Plantinga's 'free-will defense' of God's goodness in the face of the evil in the world uses the notion of divine middle knowledge. On the middle knowledge account God knows what Adam would do in certain circumstances. If he knows what circumstances will obtain at a certain time, then he knows what Adam will do at that time. Thus if God knows that Adam would freely choose to eat the fruit if Eve offered it to him , and he knows that this offer will in fact be made, then God knows that Adam will freely choose to eat the fruit. The middle knowledge account is designed to leave room for genuine human freedom. Although God knew, before he created Adam, what Adam would do, he also knows of the many possible worlds in which Adam freely refrains from accepting Eve's offer. God foreknew what Adam would do and, from this, what Adam was going to do; not what he had to do. The central question we need to ask here concerns the nature of the knowledge of subjunctive facts this account ascribes to God. Jonathan Kvanvig, an advocate of this approach, provides a precise account of this kind of divine knowledge.[6] He asks us to think of God as knowing individual essences before he created anything. An essence is a set of properties that at most one individual can exemplify. If E is the essence of a free human being it includes properties of the form: being such that if E were instantiated as an individual x and if x were placed in circumstances C, x would do A. A statement expressing this fact, i.e., one of the form: If E were instantiated as an individual x and if x were placed in circumstances C, x would do A is a "subjunctive of freedom". Since the essence of a free human being includes properties of the type just cited, each essence of a free human being has certain subjunctives of freedom true of it.

God knows which possible creatures he has made actual. So he knows which sets of properties he has caused to be instantiated as human individuals. Knowing this, he knows which subjunctive properties he has caused to be exemplified. So God knows that Adam, someone he has created, has certain subjunctives of freedom true of him, e.g. "If Adam were offered a piece of fruit from this tree he would freely accept it." On the assumption that God knows that Adam will be offered the fruit, he knows that Adam will freely accept the fruit.

I agree with Kvanvig that each essence of a free human being has certain subjunctives of freedom true of it, and that anyone who knew which subjunctives of freedom are true would know what free individuals will do in the future. But I don't think anyone, even God, knows which subjunctives of freedom are true, solely on the basis of knowing an individual's essence. Kvanvig's position depends on the assumption that a subjunctive of freedom is true because the essence of a free human being includes properties of the form: doing A in C if instantiated.[7] The subjunctive of freedom is a necessary truth about the individual essence, expressing an essential feature of it. In virtue of his innate knowledge of all necessary truths, God knows which subjunctives of freedom are true.

Why we should accept Kvanvig's assumption that the subjunctive property in question is a constituent of a free person's agent, rather than just a property the essence has, one of its contingent properties? Not every property of an essence is a necessary property of it. Essences have contingent, as well as necessary, properties. For example, it is a contingent fact about the essence of Plato that the individual who instantiates it lived in Athens rather than Grand Rapids. That essence could have been instantiated in Michigan rather than Greece. The time and place of an essence's instantiation are contingent matters. Knowing the essence of Plato doesn't guarantee knowing such inessential features of him.

It is, I believe, at least quite natural to regard properties of the form: doing A in C if instantiated as contingent properties of the essences to which they belong. But if they are contingent properties God does not know whether or not the essence has them unless and until it is instantiated and the relevant conditions obtain. Then if the individual who instantiates the essence does A God knows the corresponding subjunctive of freedom is true; otherwise he knows it is false. It appears that the subjunctive property must be a necessary property of the essence if middle knowledge is to ground divine foreknowledge of free human actions.

But there is something odd about construing doing A in C if instantiated a necessary property of the essence. If it is a necessary property of the essence, then derivatively, it is a necessary property of the individual who instantiates the essence that he freely does A in C. If it were a necessary property of this individual that he does A in C (not that he freely does it) it would be clear that he is not free with respect to this action. The addition of the qualifier "freely" perhaps is intended to preclude this conclusion, but the result is paradoxical: we are inclined to think that if doing A is one of the individual's necessary properties then he doesn't do A freely. The middle knowledge approach requires that we accept that the individual must do A, but that he does it freely. It is not clear that the subjunctive property can be coherently ascribed to the essence, at least without radically recasting the concept of freedom.

Irrespective of the legitimacy of necessary properties of the form: freely doing A in C if instantiated, from the point of view of this essay, the problem with Kvanvig's defense of the middle knowledge approach is that he gives no independent reason to believe that the subjunctive properties necessarily belong to the individual essences God instantiates, and thus provides no reason to believe God can know them before the free action takes place or fails to take place. We should not see this as a flaw in Kvanvig's account: his avowed aim is to show how divine foreknowledge of free actions is possible, not to prove that God actually has this knowledge. If the subjunctive properties we have been discussing are somehow necessary properties of the essences of free individuals then divine foreknowledge of those individuals' actions is possible. For those already convinced of divine foreknowledge of free human actions Kvanvig's ascription of the subjunctive properties to essences should be a helpful move in the direction of making the middle knowledge theory a reasonable account of that foreknowledge. But here, where the question is not what (if anything) makes foreknowledge possible, but whether God foreknows, what we need is some reason to believe that in knowing the essence of a free creature God knows what that creature would freely do under any given conditions. Until this is supplied the middle knowledge approach has not justified the view that God foreknows his free creatures' actions. William Lane Craig's middle knowledge defense of divine foreknowledge of free acts set out in The Only Wise God has the same limitation. Craig suggests that God simply knows us so completely that he knows which subjunctives of freedom are true of us (p. 145). But he does not explain why anyone who is not already convinced of divine foreknowledge should think God can know anyone that well.[8]

We have examined and rejected various ways in which God might have knowledge of future free actions. We have not, of course, proved that he does not know in some way we cannot imagine. No doubt there are things we ultimately confess about God which we have little or no hope of understanding (that he is triune, that he became a human being) Yet I propose that, in the absence of further serious possibilities to examine, we should explore the consequences of a negative answer to our question. What follows from the conclusion that God does not know what his free creatures will do in the future?

First, consider what else God would not know. He would not know whatever depends upon free human choices. On the assumption that which people come into existence depends on what genetic components come together at conception, God would not know who will exist in the future. For the mix of genetic components that occurs at conception depends upon the precise moment conception takes place, and this is highly influenced by the many particular choices prospective parents make.[9] God would not just not know what free human beings will do; he would not know which free human beings will come into existence. Once an individual human being exists, to whatever extent his future is shaped by his own or other persons' free choices, that future would not be foreknown by God. To the extent that the social, cultural and political forms that human life assumes result in part from the interplay of billions of free choices, God would not know what these forms will be. Thus he would not know what life will be like for future human beings in certain general ways.

We must qualify these general assertions about what God would not know about the future by taking account of the knowledge and power would still have, even if he did not foreknow free human actions. Many find the idea that God does not know what's going to happen disconcerting. When we do not know what will happen we feel insecure, since we are more vulnerable to threats of which we are not aware. We are forced to ignore almost all the future's possibilities, preparing ourselves to deal with only those possibilities that seem most likely. Since we must ration our finite energies and resources, we often are caught off guard and forced to try to cope, however inadequately, with unexpected eventualities. Probably these and similar features of human existence condition our idea of what it must be for God to lack complete knowledge of the future. We who rely on God's providential care must wonder how he can exercise sovereign authority if he doesn't know what tomorrow will bring.

We should not exaggerate the situation, as if God's lacking foreknowledge of free human actions left him epistemically incapacitated. Some of the differences between God and his creatures mitigate the consequences of there being logical limits on his knowledge of human actions. So far as past and present are concerned, God is omniscient, so he is aware of all current tendencies, powers, plans, and whatever else might grow into a future threat to us. Our ignorance of the future is exacerbated by our ignorance of much of what the past and present hold. Because God knows all there is to know about the past and future of the world, and about its laws and potentialities, he cannot be surprised by what happens in the sense of encountering an unexpected possibility. Since he has infinite power, he need not conserve his resources for dealing with just the most likely eventualities. No matter what his creatures choose to do, he is fully prepared. As Richard Rice puts it, this lack of foreknowledge does not "render God helpless before a dark and mysterious future."[10]

God is omnipotent. Whatever this means precisely, it implies that he has the power to intervene and control the processes of nature and the course of human affairs as he sees fit. If God wants certain things to happen at a future time he can bring it about that events move in that direction. If he wants a particular individual to come into existence, he can bring this about. If he wants a particular individual to perform a certain action, he can cause him to do it, presumably without that person being aware of any interference. Thus God can know about future human actions. This knowledge is grounded in God's knowledge of his own intentions and powers. Perhaps this is the kind of account we should attempt to give of biblical passages that imply divine foreknowledge of future human actions. Perhaps we should expect these foreknown actions to occupy crucial junctures in God's redemptive activity, e.g. Cyrus' decision to rehabilitate Israel (cf. 2 Chronicles 36:22, Ezra 1:1). The one thing these actions cannot be is free. It is not possible that God causes someone to freely perform a specific action.

Have we arrived at a conception of the shape of God's knowledge of the future that is compatible with his omniscience? Is it enough for divine omniscience that God knows everything it is now possible for him to know, and that while he does not know what free creatures will do before they do it, he knows what everyone has done and is doing? The ultimate question is not, of course, whether God fits a preconceived definition of omniscience. The question is whether the idea that God does not know what his free creatures are going to do is compatible with the God who reveals himself in the Scriptures. If the hypothesis under consideration here ultimately conflicts with the biblical witness, those of us who take that witness as the last word on what God is like will not be able to hold it. But penultimately there are some reasons to think the God revealed in Holy Scripture does not foreknow free human actions.

I suspect that we often implicitly conceive of God as having freely limited his activity but not as having similarly limited his knowing. We think of God as having brought about limitations on the scope of his activity when he created free agents. When he made free agents other than himself God chose a world in which not everything that is done is done by him. Although he is a qualitatively different actor than all created actors, there are other actors. He does not do everything that is done. This does not impugn his omnipotence; it is a result of his free creative action that not all actions are his.

A correlative result of God's creative activity is that the future is not fully knowable by him. In decreeing the existence of free creatures, God brought it about that the future is open, in the sense that even his exhaustive knowledge of the past and the present does not suffice for exhaustive knowledge of what is to come. It is proper for us to see this as an expression of God's power and wisdom. He made a world in which there are other beings with whom he can personally interact.

In conclusion, I will bring into focus one assumption I have made from the beginning, but which I have refrained from stating explicitly. I have assumed that God is temporal, i.e. that he exists in time in essentially the same way that his creatures do, except that he has an infinite past as well as an infinite future. This is the view that God is everlasting. It is opposed to the view, held by many in the classical theological tradition, that God has no temporal properties, and that he is atemporal or eternal.

The belief that God is eternal has for many seemed a way to preserve divine omniscience so far as human action is concerned without endangering human freedom. Divine atemporality was offered by Boethius in the sixth century as a solution to the problem of theological fatalism. There is no question of God literally foreknowing human actions because God's knowing cannot be located at any point in time prior to someone's acting. The metaphor we have here is of God as 'outside' time, nowhere on the line of temporality, but with exactly the same epistemic access to each moment of time. The eternal deity timelessly knows our past, present, and future. Thus he timelessly, eternally knows what we freely do.

If God is eternal in the sense of being atemporal, the question of how he knows what his free creatures will do need not trouble us. The incompatibility of causal determination with freedom, the nonexistence of laws governing human behavior, and the impossibility of distinguishing currently indistinguishable possible world are no longer relevant. On (an admittedly distant) analogy with perceptual knowledge, God would simply "see" what his creatures are freely doing at all the times at which they exist.

Despite the attractiveness of eternality as a solution to the problem, I do not think we should accept it. For one thing, it is quite difficult to square the theory of divine eternality (as opposed to his everlastingness) with the biblical account, on which he seems very much a temporal being, acting within time. Advocates of the eternality theory have traditionally had to read many biblical descriptions of God's activity as anthropomorphic. I prefer to take God's temporality at face value and instead seek a reading of texts that imply divine foreknowledge that does not commit us to God foreknowing free human actions.

Finally, we should ask whether asserting that God is eternal ultimately preserves his omniscience as traditionally conceived. There is reason to think it cannot. Nicholas Wolterstorff, in an essay which trenchantly argues for God everlasting, claims that if God is timeless he cannot be omniscient in the received sense.[11] For if God has no temporal properties, as the eternality theory requires, he cannot have any knowledge the possession of which presupposes temporal properties. Wolterstorff argues that there are things one cannot know unless one undergoes change. For example, one cannot eternally know that it is raining, since this statement's truth value changes from one time to the next. The usual way to deal with cases like this is to assume they can be translated into tenseless, 'time-indexed' statements: "It is raining on March 18, 1989." The core of Wolterstorff's argument is that a present-tense statement such as "It is raining" implies "It is raining now." But the tenseless translation, "It is raining on March 18, 1989" does not imply "It is raining now." The tensed statement and the tenseless statement have different implications and therefore have different content. The hoped-for translation fails. If God knows every true statement then he cannot be timeless.

If Wolterstorff is right, then I see little hope for the classical interpretation of divine omniscience. He contends that the assumption that God is eternal rather than everlasting implies that there are things that God does not know. I have argued from the assumption that God is everlasting, not eternal, and from other strong but reasonable assumptions (that free human behavior is not causally determined, that backward causation is impossible) to the conclusion that there are things God does not know. Either God is everlasting or he is eternal. Both possibilities appear to imply that the classical conception of divine foreknowledge is flawed. The view I believe we should seriously explore is that we have a temporal God who has chosen a future he does not fully know, because he has put it, in part, into his creatures' hands.



Footnotes

[1] For a good account of why divine foreknowledge of itself would not rule out human freedom, see William Lane Craig, The Only Wise God, pp. 67-74.

[2] The case I am making here does not depend on the probably false supposition that all events in nature are causally determined. My point will be that even if they were, natural law would not provide a basis for divine foreknowledge of free human actions.

[3] The classic defense of this view is in Donald Davidson's essay "Mental Events," in Experience and Theory, edited by Lawrence Foster and J. W. Swanson. Amherst: University of Massachusetts Press, 1970.

[4] A good recent defense of the incompatibilist view is Peter van Inwagen's An Essay on Free Will, Oxford: Clarendon, 1983.

[5] This assumption, however plausible, needs to be squared with such texts as Jeremiah 3:7, which appears to assert the opposite.

[6] Jonathan Kvanvig, The Possibility of An All-Knowing God, New York: St. Martin's Press, 1986.

[7] Kvanvig, pp. 121-126.

[8] Craig, William Lane. The Only Wise God. Grand Rapids: Baker, 1987.

[9] Parfit, Derek, Reasons and Persons, pp. 351ff.

[10] Richard Rice, God's Foreknowledge and Man's Free Will, p. 58.

[11] Nicholas Wolterstorff, "God Everlasting," in Contemporary Philosophy of Religion, edited by David Shatz and Steven M. Cahn. New York: Oxford University Press, 1982.

Sources

Davidson, Donald. "Mental Events." In Experience and Theory, edited by Lawrence Foster and J. W. Swanson. Amherst: University of Massachusetts Press, 1970.

Kvanvig, Jonathan. The Possibility of An All-knowing God. New York: St. Martin's Press, 1986.

Parfit, Derek. Reasons and Persons. Oxford: Clarendon Press, 1984.

Rice, Richard. God's Foreknowledge and Man's Freedom. Minneapolis:Bethany House, 1985. (Originally published as The Openness Of God, 1980)

van Inwagen, Peter. An Essay on Free Will. Oxford: Clarendon, 1983. Wolterstorff, Nicholas. "God Everlasting." In Contemporary Philosophy of Religion, edited by David Shatz and Steven M. Cahn. New York: Oxford UniversityPress, 1982.