It has been argued by some listmembers that, since God is omniscient, man cannot have freedom of choice. Some years ago, I wrote a reply to a lecture by a philosophy teacher, Dr B, in which he argued that since man has free choice, God cannot be omniscient. I offer a section from that reply as a contribution to the present discussion.
***** ***** ***** ***** *****
Dr B argues as follows:
Suppose that God knows -- and knows today -- the truth value of every proposition. Then either God knows today that Jones will steal on his thirtieth birthday or God knows that Jones will not steal on his thirtieth birthday. Either way, the question is already settled, and Jones is not free to choose. But we know that Jones is free. Hence the premise (God is omniscient) is false.
Now the difficulty presented by this argument does not arise from any assumptions about God. The same difficulty occurs in exactly the same way in the following parallel argument against one of the fundamental laws of logic, known since the time of Aristotle as the Law of the Excluded Middle -- A or not-A -- "A proposition, provided it has a clearly defined meaning, must be either true or false."
> Suppose that at this moment every proposition is either > true or false -- A or not-A. Then either it is already true > that Jones will steal on his thirtieth birthday or it is > already false that he will. Either way, the question is > already settled, and Jones is not free to choose. But we know > that Jones is free. Hence the premise (A or not-A) is false.
Throughout this section, let us call Dr B's argument against omniscience the First Argument and the parallel argument against A or not-A the Second Argument. I maintain that every way of dealing with the Second Argument suggests a similar and equally satisfactory way of dealing with the First Argument. How might Dr B reply to this Second Argument? I suggest three possibilities:
(1) He might say: I agree. A proposition about a future choice by a free agent is neither true nor false but undetermined. When the agent makes his choice, and not before, then the proposition about his choice acquires a determinate truth value -- either true or false.
A former student of Dr B's has expressed the idea thus:
> It should be noted that an assertion may be neither true > nor false. For example, the statement "Another world war will > begin in the year 2000" predicts an event. Perhaps the event > will occur; perhaps not. In any case the events of the year > 2000 lie in the future and are therefore quite indefinite at > this moment. Since the truth or falsity of a statement depends > on whether it agrees with or contradicts reality, and since the > history of the year 2000 is, as yet, indefinite and undecided, > it follows that the prediction itself is neither true nor false > at this time. A statement cannot be true or false unless the > alleged facts it refers to are actualized in reality. A > statement that is either true or false is called a > _proposition_.<*>
<*> Paul Lepanto, RETURN TO REASON: an Introduction to Objectivism (Exposition Press, Jericho, New York, 1971) p34
Now what grounds are there for conjecturing that Dr B might take such a position? Well, he is on historically firm ground if he does. He is not repudiating Aristotle's Law of the Excluded Middle -- he is affirming it in precisely the sense that Aristotle intended. Aristotle considers the Second Argument, concludes that A or not-A, the Law of the Excluded Middle, is indeed incompatible with human freedom, and that since freedom is a reality, the law must be inapplicable to future human action. He says:
> In the case of that which is or which has taken place, > propositions ... must be true or false. ... > When, however, the subject is individual, and that which > is predicated of it relates to the future, the case is altered. > For if all propositions ... are either true or false, then ... > there would be no need to deliberate or to take trouble, on the > supposition that if we should adopt a certain course, a certain > event would follow, while, if we did not, the result would not > follow. For a man may predict an event ten thousand years > beforehand, and another predict the reverse; that which was > truly predicted ... will of necessity take place in the fulness > of time. > Further, it makes no difference whether people have or > have not actually made the contradictory statements. ... For > events will not take place or fail to take place because it was > stated that they would or would not take place.... > Yet this view leads to an impossible conclusion; for we > see that both deliberation and action are causative with regard > to the future, and that ... there is a potentiality in either > direction. ... > Since propositions correspond with facts, it is evident > that when in future events there is a real alternative, and a > potentiality in contrary directions, the corresponding > affirmation and denial have the same character. ... > It is therefore plain that it is not necessary that of an > affirmation and a denial one should be true and the other > false. For in the case of that which exists potentially, but > not actually, the rule which applies to that which exists > actually does not hold good.<*>
<*> Aristotle, ON INTERPRETATION, 18-19 (ix); BASIC WORKS OF ARISTOTLE, Random House Edition pp. 45-48
If someone says that he believes in Aristotle's Laws of Thought, it is certainly reasonable to suppose that he means that he believes them in the same sense that Aristotle did.
If Dr B takes this line, then the argument against omniscience collapses. We ask God whether the statement that Jones will steal is true or false, and God replies, "Undetermined." This is not a confession of ignorance, but is the correct answer to the question, and the only correct answer. Many modern physicists believe that a particle does not have exact co-ordinates of position and velocity, but only approximate ones. Regardless of the merits of this position, it is clear that not even an omniscient being could be expected to know the exact co-ordinates of a particle that does not have exact co-ordinates. It would be like insisting that, since God knows everything, he must know the address of my stockbroker even though I don't have a stockbroker. But the thoughtful critic will recognize that omniscience does not mean knowing the answers to all questions. It means knowing the answers to all questions that have answers, and knowing of the other questions that they do not have answers.
(2) Again, Dr B might reply to the Second Argument as follows:
> Note the ambiguity in statements of the form, "If A, then > necessarily B." We say, for example, "If Jones is younger than > Smith, then necessarily Jones is younger than Smith's mother." > And it is easy to read this as meaning that if the first clause > is true then the second clause expresses a necessary truth, an > inescapable truth, a truth guaranteed by logic. But of course > in such sentences logic does not guarantee B, but only the > derivation of B from A. The adverb "necessarily" modifies, not > "B", but "If A, then B." B is not a necessary _truth_, but the > relation obetween A and B is a necessary _implication_. Instead > of saying, "Given A, B must hold," we ought to say, "Given A, B > must follow." > We have here an example of what is sometimes called the > Fatalist Fallacy: "That which is, must be." People say that if > a man is sitting he cannot be standing and if he is standing he > cannot be sitting, and so either way his posture is a matter of > necessity and not of choice. But the "cannot" here does not > express an impossibility of sitting, but rather an > impossibility of sitting-when-he-is-standing, which is not the > same thing at all, and leaves him perfectly free to sit or > stand as he pleases. When someone says, "If it is true that > Jones is sitting, then Jones must be sitting," he is offering a > special case of the statement, "If p is true, then p must be > true." But this must be read as "p necessarily-implies p," and > not as "p implies necessarily-p." And thus read, it has no > overtones of non-freedom. In the sentence, "If x is greater > than three, then x must be greater than two," the mathematical > necessity expressed by "must" attaches (to use the technical > jargon of the logicians) not to the _consequent_ (x is greater > than two) but to the _consequence_ -- to the "if...then," the > conditional relation between the two clauses.<*> It is not that > the second clause necessarily _holds_, but that the second > clause necessarily _follows_ from the first. > Now, as for this Second Argument: > This pretended refutation of "A or not-A" confuses > "necessity of the consequence" with "necessity of the > consequent". It is the Fatalist Fallacy, which we have just > analyzed. Granted, if it is true that Jones will steal, then > Jones must steal. But "must" here means, not that it is > impossible for Jones to remain honest, but only that it is > impossible that _both_ Jones remain honest _and_ the prediction > that he will steal be fulfilled. And this does not impair > Jones's freedom.
<*> John Hospers, INTRODUCTION TO PHILOSOPHICAL ANALYSIS, page 281 (I chose Hospers because he is a friend of Dr B.)
If this is Dr B's reply, then the argument against omniscience collapses. For we see that the sentence, "If God knows that Jones will steal, then Jones must steal," and the sentence, "If it is true that Jones will steal, then Jones must steal," use the word "must" in the same ambiguous manner, and an analysis of the structure of one applies straightforwardly to the other.
(3) Another line Dr B might take is to say:
> If a situation obtains today that is incompatible with Jones's > not stealing on his thirtieth birthday, than admittedly Jones > is not free. But a proposition is not a situation. A situation > is tied down to a particular time and place. A proposition is > not, although it may refer to something that is. "It is true > today that Jones will steal on his thirtieth birthday" and "It > was true last week that Jones will steal on his thirtieth > birthday" and "It will be true a thousand years hence that > Jones will (or did) steal on his thirtieth birthday" are > synonymous statements. To affirm one and deny another is to > contradict oneself. A proposition, though it may be about > something in time, is not itself in time at all. In the > sentence, "It is true that Charlemagne was crowned in 800," the > "was crowned" is past tense, but the "is true" is only > grammatically present tense. Logically, it is no tense at all. > Change it to "was true last week" or to "will be true next > week" or to "was true in 600" and you have not changed the > meaning. Charlemagne's coronation is an event in time; the > proposition affirming it is not. > > A shoe is a material object, subject to time and change and > disintegration. It may wear out and fall apart, or be eaten by > rats, or be attacked by rot or mildew. But the proposition that > seven is a prime number, and the proposition (assuming it to be > true now) that Charlemagne was crowned in 800, will always be > true. There is no way for them to grow old, or fall apart, or > be destroyed. They are not subject to change or decay. They are > timeless. > > Similarly, the proposition that Jones will steal is not a > presently existing cause of Jones's future action. It is a > timeless entity. > > Thus there is no event occurring or situation obtaining today > that is incompatible with Jones's not stealing on his thirtieth > birthday, and hence no restriction on Jones's freedom has been > shown.
If Dr B takes this position, then the argument against omniscience collapses. The argument for the timelessness of a proposition is parallel to the argument for the timelessness of the knowledge of an eternal and omniscient being.
More generally, we may say that any refutation at all of the Second Argument will turn out to be a refutation of the First. For the First Argument actually presupposes the Second. It says, in effect: "Suppose (a) God knows now that Jones will steal. Then (b) it is true now that Jones will steal. Then (c) it is already a fact that Jones will steal. Then (d) it is impossible that Jones should not steal. Then (e) Jones has no choice between stealing and not stealing." The First Argument is that (a) implies (e). The Second Argument is that (b) implies (e). But the road from (a) to (e) lies through (b), and if you can't get from (b) to (e), than you can't get from (a) to (e). Hence, if the Second Argument is invalid, so is the First.
***** ***** ***** ***** *****
At this point, perhaps some critic will complain that I have pulled the usual mystical gambit of evading the force of Dr B's argument about Divine knowledge by explaining (or rather, anti-explaining) that God's knowledge is "timeless" and therefore quite different from what we mean by knowledge. I meet this objection by showing that even ordinary human knowledge before the event is no obstacle to freedom.
Consider the following argument:
> Suppose we have a situation on Monday from which it necessarily > follows that Jones steals on Tuesday. Then Jones's act is > predetermined and hence not free. Now if Smith correctly > predicts on Mondey that Jones will steal on Tuesday, it > necessarily fillows that Jones steals on Tuesday. Therefore > Jones is not free. Right?
Wrong!
For if one man says, "On Monday the weatherman correctly predicted rain on Tuesday," and the other man says, "On Monday the weatherman incorrectly predicted rain on Tuesday," they are not disagreeing about the events of Monday, but about those of Tuesday. The correctness of the prediction is not a part of the situation on Monday. If two persons agree on what Smith said but disagree on whether it was a correct prediction, they are not disputing about what happened on Monday. What happened on Monday is that Smith made a prediction. What happened on Tuesday was that Jones fulfilled the prediction. No description of the events of Monday that confines itself to Monday implies anything at all about the events of Tuesday. But the statement that Smith correctly predicted on Monday what Jones would do on Tuesday is not just about the events of Monday. It is about the events of both days, and how they were related.
A similar analysis holds if we put it more strongly and say that Smith knew on Monday that Jones would steal on Tuesday. For what does it mean to say that Smith knows that X is true? It is usually taken to mean that (a) Smith believes that X is true, and (b) Smith has good evidence that X is true, and (c) X is true.<*>
<*> John Hospers, INTRODUCTION TO PHILOSOPHICAL ANALYSIS, 144ff
But note that good evidence does _not_ mean evidence such as we never have for false propositions. If Smith correctly believes that Bismark is the capital of North Dakota because he has looked it up in the atlas, then we would agree that he knows it, without meaning to assert that atlases never contain misprints. So, if Smith believed on Monday that Jones would steal on Tuesday, and if Smith had good reason for his belief (such as hearing Jones discussing the prospective theft with an accomplice) and if Jones did in fact steal on Tuesday, we would say that Smith knew he would. And if Jones changes his mind at the last minute, then we would say that Smith did not know after all, but that his reasons were still good reasons. Thus, we see that Jones's freedom is not incompatible with Smith's foreknowledge. I remind the reader that knowledge does not mean infallibility. "John knows that snow is white," means that (a) John believes that snow is white, and (b) John has good reason to believe that snow is white, and (c) snow is white. It does not mean (d) John has better reason to believe that snow is white than anyone could ever have for believing something that is in fact false. That would exclude us from ever saying that we know the sum of a column of figures, since our evidence must necessarily be that we (or others) have added them up, and people have sometimes added wrongly, and failed to find the error on adding again. It does not mean (e) John is more confident that snow is white than anyone ever is of something that is false. The world is full of people who believe all kinds of false things with unsurpassable confidence.
At this point a critic may say: I will grant your definition of knowledge, and grant that good evidence is not the same as conclusive evidence (_i.e._, evidence that is simply incompatible with the falsity of the conclusion). But I deny that we can ever have good evidence (in the sense presupposed by the definition) for what someone will freely choose to do, and so I deny that we can ever know what someone will freely choose to do.
If we insist that Smith can never know in advance what Jones will (freely) do, we must say that knowing implies having such good reasons for belief as are never available where human freedom is concerned. But this commits us, if not to a universal skepticism, at least to a very drastic one. For our reasons for supposing that someone will not lie are very like our reasons for supposing that he has not lied in some instance where we cannot check him, and so farewell to all knowledge that depends in any way on human testimony. Indeed, given that humans have considerable ability to suppress or distort memories of incidents that they do not choose to remember accurately, it would seem (by this line of argument) that I cannot know whether I have chosen to remember something inaccurately, and then chosen to forget that I have so chosen, and so farewell to all knowledge that depends on my own memory of my own past experience. That does not leave much. In short, if I can never know what someone will (freely) do, then there is precious little that I can know.
I very much doubt that Dr B would agree with that conclusion. In view of what he has said on a closely related subject,<*> I should expect him to comment somewhat as follows:
<*> [What follows is taken with minute changes from a story used by Dr B to illustrate another point.]
> Suppose that you attend a gathering with a friend. At this > gathering, a stranger suddenly confronts you and charges you > with being about to commit a murder. You indignantly deny it, > but the stranger insistently repeats his charge. > "What murder?" you demand. Your accuser does not answer. > "Whom will I kill?" you demand. Your accuser does not answer. > "Why do you suspect me?" you demand. > Your accuser smiles slyly and answers, "I believe that you > will commit a murder this very evening. Can you prove that you > won't?" > You turn away -- and see that the friend with whom you > came is looking at you tensely. You cry to him, "_You_ don't > believe that I will murder anyone, do you?" > Your friend answers nervously, "No, of course I don't. I > mean ... he hasn't given any _evidence_ that you will commit a > murder, he's just predicted it. And since you have free will, > he couldn't possibly know that you will commit a murder. On the > other hand, he couldn't possibly know that you won't commit > one. Neither could I, for that matter, nor could you. If > anyone could predict what you will do, then your actions would > be fated and predetermined. I guess I'd have to say that nobody > knows whether you will murder someone this evening or not." > Thereafter, your friend is very fair and conscientious; he > makes it very clear to everyone that he does not believe that > you _will_ murder anyone; he is, he explains, an agnostic in > the matter, and his decision to leave the gathering at once and > without you is a purely precautionary measure. > If you were the victim of such a nightmare, you would feel > that a monstrous injustice had been committed -- specifically, > an epistemological injustice. And you would be right.
Another objection a critic might voice: It is true that human knowledge is compatible with freedom. But that is only because human knowledge is contextual, because every human knowledge claim is (or ought to be) provisional, subject to revision in the light of the further evidence. But theists claim that Divine knowledge is something quite different -- that it is timeless, which human knowledge is not, that it is not acquired, which human knowledge is, and that it involves omniscience, which human knowledge does not. To suppose that you can prove Divine knowledge compatible with freedom by proving that human knowledge is, is a blatant _non sequitur_.
To this I reply: Granted. But please remember how the discussion has developed. First, Dr B undertook to show that Divine knowledge, as traditionally understood, is incompatible with freedom. I undertook to meet him on that ground by showing that the idea of a timeless, omniscient being does not contradict freedom. In case someone complains that I am avoiding difficulties by bringing the mysterious notion of timelessness into the argument, I then argue in terms of ordinary human knowledge. But if a critic began by complaining because I brought timelessness into the argument, let him not now turn around and complain that I am avoiding difficulties by leaving timelessness out of the argument!
***** ***** ***** ***** *****
The above was my reply to Dr B. To the Christian reader I will recommend the chapter "Time and Beyond Time" in the book MERE CHRISTIANITY, by C S Lewis. In addition, the reader, even if not majoring in the physical sciences, might want to read a good elementary introduction to relativity for the non-specialist, such as that by George Gamow, or Isaac Asimov, or Bertrand Russell. (Gamow's books are his MR TOMPKINS series and ONE, TWO, THREE, INFINITY.) The goal here is to understand that time is so closely connected with physical processes (and may even be measured differently by persons on different planets, with each being right from his own point of view) that it makes sense to say that the concept of time has no application to God, who is not part of the physical universe, and in fact made it. </PRE>