On Choice and Judgment (Column 175)

With God's help

At the beginning of the week I was asked on the site about judging and condemning wicked people (Hitler). The questioner presented himself as a determinist and asked about the possibility of judgment in the libertarian picture. I wrote back that, in my opinion, there is no point in discussing the matter with him, because on his view his conclusions are forced upon him, and therefore they have no significance. Even if he now thinks something, or is persuaded in the course of discussion of something else, that is only because he is compelled to those positions; but he himself cannot infer from this anything about their correctness.

In effect, I shifted the discussion here from choice on the moral-value plane (the choice of what is proper) to judgment on the intellectual-scientific plane (the choice of what is correct/true). For a discussion of the relation between these planes, see also the thread here. In Column 35 I pointed out that libertarianism (the view that we have free will) has significance in the factual context as well, and not only in the value context. I argued there that facts too are not forced upon us, and that forming a position regarding facts (as in science) is not merely the result of a mechanical calculation. My motivation for saying this was that only within such a worldview can one speak of intelligence. Intelligence is not the capacity for mechanical calculation (after all, inanimate things have that too, such as water, which performs a "calculation" and chooses an extremely complex course of motion and behavior), but the ability to exercise judgment that determines which calculation we should perform in order to solve the problem and which argument is correct in the circumstances under discussion.

The questioner objected that there is no difference between me, as a libertarian, and him; for even on my view a person acts on the basis of intuitions, and therefore on my view too the truth is forced upon me. The intuitions implanted within me are what cause my decisions, and I did not decide them. They are somehow embedded in me. In effect, his claim is that even in the libertarian worldview, the criteria by which I deliberate are forced upon me. [By the way, I would note that this argument does not undermine my actual distinction that conclusions forced upon us should not be taken into account; it merely extends it into the libertarian's territory as well. Even if he is right, it would still be true that we should not take our conclusions into account, so what is the point of discussing at all?!]

I answered him that if so, the discussion is entirely worthless. He asked me according to my position as a libertarian, yet within his own words he objects to libertarianism itself. Therefore he must decide: either we are discussing whether libertarianism is correct, or we are discussing, within the libertarian framework, whether there is room for judgment and deliberation.

In any case, this exchange made clear to me that the analogy between choice in the value context and choice in the intellectual context, which in Column 35 I took for granted, requires further clarification and sharpening.

Choice in the value context

The discussion begins with the account I propose (in Column 173, and in greater detail in my book Mada'ei HaHerut (The Science of Freedom), among other places) of the libertarian conception. I argue that determinists sometimes attack a libertarian approach that is a straw man. Many such attacks assume that the libertarian who advocates free will is free of influences, that is, that he does not accept the very existence of various influences on a person (genetics, environment, education, and so forth). But as I explained there, a sober libertarianism certainly accepts the fact that a person is subject to various influences. It merely argues that, contrary to the deterministic conception, those influences do not determine the result but only affect it.

I suggested describing this picture by means of a model of terrain within which a person operates. A ball or a stream of water that found itself in the same situation would arrive at the point where the potential energy is lowest (rolling down into the valley rather than climbing the mountain); that is, the forces exerted on it by the terrain would determine its path unambiguously. By contrast, a person in the same state is indeed also affected by those forces (they act upon him as a material object), but in most cases (except for an irresistible impulse) he also has the option of deciding to act against them. In this model, the terrain reflects all the influences on a person: genetics, education, environment, and so on. All of these are integrated in the brain and create various impulses that try to cause the person to act in a certain way. But according to the libertarian, he has free will, and therefore he can also act otherwise (climb a mountain, rather than necessarily rolling down into the valley).

In one comprehensive term, I would call all the pressures that come from the terrain impulses. These include both the good impulses and the bad ones, since an impulse tries to make a person act out of urge rather than out of a value decision (and that gives rise to the problem of weakness of will, which I discussed in Columns 172–173). The task laid upon a person is to decide, and not to let the impulses (all of them: the good, the bad, and the ugly) dictate his decisions and actions.

The root of the problem

Up to this point I have described the libertarian picture regarding the value choices we make. My claim is that judgment is choice on the intellectual plane, that is, a choice of the intellect (and not of the will). When we move to the intellectual plane and want to speak of judgment by analogy to value choice, we must examine all the details of the model I proposed and see whether it does in fact describe the intellectual process as well. In other words, is there a similar model that reflects our judgment?

I think the main thing missing in the intellectual-factual context is impulse. When I deliberate whether to steal, there is a value, and opposite it stands an impulse or an interest. Hence I am in a dilemma, and that is where my power of choice enters (and this is the problem of weakness of will that we discussed in Columns 172-3). By contrast, when I deal with a factual, metaphysical, or scientific question, the situation is different. Here there does not seem to be an impulse standing opposite the truth and putting me into a dilemma that requires choice. Therefore the questioner claimed that intellectual choice is forced upon me (that is, that there is no real choice here).

Admittedly, there are factual questions in which impulses may exert influence. For example, whether or not God exists. Here I can understand that the evil inclination would try to influence my position, since this factual question has value implications. But when I come to build a scientific theory, or solve a mathematical problem, there does not seem to be any impulse in the background. The question is what seems right to me and what does not. In the terms of the previous section, here it seems that there is nothing beyond the terrain. If my calculation shows that there is a law of gravitation and pushes me to believe in it, then I will presumably also decide that such a law exists. Why should I not accept what my thought tells me? If so, then apparently even in the libertarian picture our intellectual and scientific decisions are the result of calculation.

An alternative formulation

Perhaps this can be formulated differently: after all, on the intellectual plane there is right and wrong in reality. What, according to my understanding, is the reality—that is the reality. That is forced upon me, and there is no reason to activate a mechanism of choice here. Even in a situation in which reality is unclear, that is, where there are various arguments and considerations pointing in several directions, my task is still to decide which is most correct. Therefore here too, ultimately, we are dealing with a mechanical weighing (a calculation) regarding the factual plane.

But this formulation is mistaken, and this time precisely on the side of values. It tacitly assumes that with respect to values there is no right and wrong, and therefore there is nothing that compels me to decide one way or the other, and from that it follows that it is precisely there that one can speak of a free decision. But as I have written here more than once (see, for example, Column 128, in the discussion of Ari Elon's mistake, and throughout that entire series), such a conception describes a lottery, not a decision. A decision always means an attempt to hit upon what is proper (= morally right). If there is no objective proper course at all, then our choice has no meaning and is essentially a lottery. Therefore, someone who speaks of free will in the sense of choice, and not in the sense of indeterminism (= a lottery), cannot deny the existence of a moral-value 'right' (= the proper).

So there is a proper and an improper on the value plane as well, and nevertheless the libertarian holds that it is not forced upon me. There and also in the columns on weakness of will (172-3), I explained that we have impulses that try to divert us from doing the right thing (or to cause us not to choose). We have free will that determines what to do. But as I explained there, we do not have free will that determines what it is proper and improper to do. That is dictated and given to us in advance. I explained there that it is precisely this dictate that gives meaning to our choice (otherwise it would be a lottery).

But precisely this greatly sharpens the problem in the intellectual context. We saw that in most cases there are no impulses here, since the matter is one of factual right and wrong. Therefore the decision here parallels the decision of what is proper and what is not in the value context, and not the decision of what to do (a decision between the proper and the impulses). If so, it seems that in the intellectual context we really are forced to our conclusions, parallel to the proper and improper in the value context.

Two kinds of dilemmas

In my book Mada'ei HaHerut (in chapter 14, on Libet's experiments), I distinguished between two kinds of dilemmas: a value-value dilemma and an impulse-value dilemma. The dilemmas I have described so far (and in the discussion of weakness of will) belong to the second type. The first dilemma, by contrast, is a conflict between two values. I gave there Sartre's example, in which he tells of a student of his during the Holocaust in occupied Paris who hesitated whether to flee France and join the Free French army to fight the Nazis, or to remain in Paris and care for his elderly mother. Here we are not dealing with an impulse versus a value, but with hesitation between two values. Here we must decide what it is proper to do (which of the two is more morally correct), and that is all. There is no further stage of decision. In that sense, this dilemma resembles intellectual dilemmas between two possibilities, or two considerations, both of which seem persuasive. The same is true of dilemmas between communism and capitalism (although some see this as a value-impulse dilemma. I think they are gravely mistaken), and of other value and moral quandaries.

You can of course claim that in such dilemmas there really is no objective truth, and therefore no room for decision. But if so, then it is a lottery between two possibilities, since both are equally proper. Yet that does not fit any of our intuitions. When we deliberate over such questions, we have a clear sense that we are seeking what is right; otherwise we should simply conduct a lottery and be done with it. This is moral judgment. But as I explained, hesitation always indicates that there is something we are trying to hit, some pure truth of some sort. Between two equally weighted truths there is no room for hesitation, only for an arbitrary lottery.

Admittedly, in the factual context it is clear that this is not so. There is one correct answer there. Either there is a law of gravitation or there is not. If I deliberate between considerations and arguments for and against the existence of a law of gravitation, it is clear that a lottery is not an option on the philosophical plane. One can hold a lottery out of despair, or merely as a practical expedient, but there is no reason to assume that it brings us closer to the factual truth. The truth is the truth regardless of the outcome of the lottery: either there is such a law or there is not. Therefore, in the scientific-intellectual context, the decision cannot be a lottery. It must be based on a mechanism that will take us as close as possible to the truth. And if there is no such mechanism, we must leave the matter at "requires further study," and not decide.

Scientific hesitation of this sort consists in weighing all the facts and arguments in every direction, and arriving at the scientific conclusion. How should we understand deliberation on a scientific question that ultimately yields a decision? I assume that the above questioner would say that this is simply a longer and more complicated calculation, but still a mechanical one. Therefore, on his view, the factual-scientific result is forced upon us. But I want to propose here a different account, by way of the analogy between choice in the value context and judgment in the factual-scientific context.

What is judgment?

If I want to speak of judgment as choice on the factual-intellectual plane, I must apply all the details of the model we saw above to the intellectual context as well. The terrain is my instinctive system of considerations. It indeed performs a completely mechanical calculation, given a collection of facts and considerations, and presents me with a result. The result can be a single conclusion, or a dilemma between two possibilities that both seem reasonable. Now judgment enters and is supposed to decide which is the correct possibility. Judgment exists in both cases: even when the conclusion of the first stage is one definite conclusion, judgment can still decide that there has been a mistake here and that I should not adopt it. And when there is a dilemma between two possibilities, judgment decides which of them to accept (or perhaps neither of them). This is exactly what is missing from the deterministic picture, in which the result of the calculation is forced upon me.

On what basis is this libertarian decision made? On no basis. It is a creation ex nihilo, just as in value decision. If there were a cause that generated this result, or if there were some basis on which a calculation was conducted that led to this conclusion, then there would be no decision here, but rather a mechanical calculation. What is troubling in this description is that this decision is supposed to correspond to factual reality. This is not a free decision in which either of the two possibilities is valid simply because I decided that way. After all, I am ultimately claiming that, in my view, this is the correct scientific theory. But as I argued above, that is not the case in the value context either. There too there is a correct decision, and therefore the existence of judgment does not contradict the fact that one is aiming at a specific conclusion, and that this conclusion is measured against some objective truth.

Judgment is an attempt to look at the system of considerations and facts (or the value arguments, in the value context) that the mechanical calculation produced, and to examine which of them is correct. This is not an observation of reality, because that is what the calculation does. It is reflection on the observation, and a decision whether that observation was correct or not, or which of several possible observations is the correct one. This is exactly what we saw with value-value dilemmas in the value context.

I gave several examples of this. An atheist hears devastating arguments in favor of the existence of God. He has no good answer, but he feels that something there is problematic. So although the arguments and the result of the calculation are clear, he decides not to adopt it. This is the action of the control system that decides whether to adopt the calculation or whether it is misleading. The same is true of a fata morgana in the desert, where a person becomes convinced that there is an oasis before him, but rejects that conclusion through his intellectual judgment. The same is true of John Nash (see about him in Column 22, and more generally there on neutralizing emotions in decision-making), who experienced very vivid hallucinations and forced himself to ignore them because he concluded that they were imaginary. These too are the workings of our control system, which examines the results of calculations and decides, through judgment, whether to adopt them or not.

The motivation

To sharpen the point and its significance, it is important here to return to the fundamental motivation. My basic claim is that if the factual-scientific conclusion is forced upon us, that is, if it is merely a mechanical calculation, then we have no way of accepting it. A conclusion that is forced upon us can of course be correct, but for us, as beings located within the system, there is no way to know whether it is correct or not.

Take, for example, a computer that is fed a program that performs some calculation. Assuming we have no information at all about the programmer and the program, we have no way to know whether the result is correct. Suppose you have a computer before you. You enter into it two pieces of data, 1 and 7, and tell it to add them, and it outputs 8 (or 1/3). Will you accept this? Only if you know from an independent source that the result is correct, or that the computer is in fact equipped with a program for arithmetic addition. Let me sharpen the point and say that I do not mean to claim that the result is necessarily incorrect. It may well be correct, but we have no way of knowing that. This is, of course, on the assumption that all the information available to us is the result of the computer's calculation, that is, that we have no independent way to test that result. In such a case, our attitude toward it should be skeptical at best.

Now I will add another stage: for any given calculation, there is almost no program in the world that would perform it correctly. Take again the sum 1+7. To perform it, one must build the program specifically and tailor it to the task, so that it will carry out that particular operation (arithmetic addition) correctly. Almost no program in the world that was not built for that purpose will do so correctly. It follows that if we randomly choose a program from among all possible programs, there is no doubt that the result we receive for the addition will be mistaken.

So when we see a device that performs a calculation of arithmetic addition, and we have no information about it at all, nor about the result of the calculation, the assumption is that this is really no better than a lottery. Therefore it is reasonable and rational to reject the result of the calculation and not believe it. The same applies to the system that compels us. We have no information about it beyond the fact that it exists (and even that is doubtful. See the appendix). We have here a machine that receives inputs and produces some output (a conclusion). In such a situation, our attitude toward the result of the calculation should be more negative than mere skepticism. In fact, we have seen that it is reasonable to assume that the conclusion is incorrect. Notice: not merely doubtful, but incorrect.

Fine—but do we have information about this machine? After all, it was produced by evolution to yield correct (or useful) results. To that I say two things: 1. Evolution creates machines of maximal survivability, not of maximal reliability (and those are very far from being the same thing). 2. Even the information we think we possess about the compelling system (from evolution, brain research, and biology in general) is itself based entirely on using that very same system. Therefore it cannot be used to establish its reliability. This is a logical loop.

The conclusion is that within a deterministic framework that recognizes no judgment beyond mechanical calculation, the fact that we reached conclusion X should play no role whatsoever. It has no significance for the question whether X is correct or not, because we do not know the nature of the compelling system. In fact, as we have seen, it is much more reasonable to assume that X is mistaken (there is no reason to assume that by chance we happened to land on the correct conclusion). This is because, according to the determinist, there is nothing in us beyond that same compelled system, operating within the external system that forces us to arrive at our conclusions. We have no external check on it, for we are nothing but the programmed computer itself.[1]

This is the basic motivation for assuming the existence of free judgment, that is, libertarianism. The meaning of such judgment is that there is something in us beyond the system of mechanical calculation. That "something" reviews the result of the calculation and can examine whether it is correct or not. It is itself not a mechanical calculation, and therefore is not carried out through that very system. Without that assumption, there is no reason in the world to believe our own conclusions.

A skeptical query

An obvious question regarding the picture I have proposed here is: who says that this control system is reliable? Are the rules according to which it itself operates not exposed to the same query? Are they not forced upon us? This is essentially what the questioner asked regarding our choice of intuitions.

But here too there is a mistake. My claim is that the control system is not a calculation according to criteria. It is judgment, not calculation. Judgment, by definition, is free, and precisely for that reason it can be trusted. I simply see that it is true, and that is all. There are no criteria here on the basis of which I perform a calculation. Returning to the example of sight: if someone were to cast doubt on what my eyes see, I would tell him that I see, and that is all. He would then ask: of course you see, but who says that the decision that what you see is true is reliable? On what basis do you accept that decision itself? My answer would be that this decision is not the result of a calculation but of a direct encounter with the truth. In the formulation I suggested above, in this decision (the judgment regarding sight) I am looking at the seeing, not at the scene. I look with my eyes at the scene, and then my control system looks at the mechanism of vision and decides that it is reliable.

Needless to say, I am not arguing here for absolute reliability. I do not mean to say that a person and his control system never err. Of course, all of us make mistakes from time to time. My claim is that error is an unfortunate case, not something essential to the system. The system is reliable, and therefore its conclusions should be treated positively, until it becomes clear that we were mistaken.

More skeptical questions

Now, of course, a claim may be raised on a more fundamental level: this is a pragmatist consideration. Who says that you really have such an external control system? Perhaps you are nothing but a system that performs mechanical calculations and nothing more. The fact that without the assumption that there is such a system one cannot believe any of our conclusions only shows why we would want there to be such a system. But what guarantees that it actually exists? I am simply positing the existence of external control in order to reach the result I desire. These are mere wishes, and therefore I am in the same position as the determinist. The claim is that this is pragmatism (forcing reality to conform to desire).[2]

But that claim is nothing but a skeptical query: perhaps there really is no such control system, and perhaps none of our conclusions should really be accepted. Fine, perhaps. But the same can be asked about what we see and think and experience. Perhaps we see or hear, but it is unreliable? If we are skeptics, then there truly is no point in talking about anything. This discussion, and in fact every discussion, proceeds under a non-skeptical assumption. Under that assumption, the conclusion is that determinism is a problematic and plainly unreasonable position. In other words, the only alternative to libertarianism is not determinism but skepticism. There is no possibility of consistent determinism.

Beyond that, I will mention here what I elaborated throughout the fourth booklet (especially in the first part, which deals with methodology)[3]. What I called there a "theological" consideration (as opposed to a "philosophical" one) is a consideration that assumes a conclusion and tries to derive from it the premises that lead to it. Thus, for example, I argued there in Part C that if a person accepts the validity of morality, an implicit belief in God is necessarily contained in that. In Part B I argued that if a person trusts his cognitive faculties, he too is tacitly assuming the existence of God. In these cases as well there is a tendency to see this as pragmatism, since we are creating God as a guarantor of morality or cognition (I explained there that this is exactly what Kant did; he was the one who put his finger on 'theological' thinking, without calling it that. Marx had already said that religion is the opium of the masses). As I explained there, this is a mistake. If the starting point were that we want morality (or reliable cognition), and from that we derived the existence of God, that would be pragmatism. But if the starting point is that there is morality (we really think morality is valid), that is, that this is a conviction rather than a desire, then it is a valid logical argument (in the 'theological' direction) and not pragmatist at all.

It is important to understand that my claim against the determinist is far stronger than the skeptical claim he raises against me. This is not a mere skeptical query, but a positive claim. I am not saying to him: who told you this is true? (That would seemingly be a 50-50 doubt.) I am saying to him that within the picture he holds, it is positively clear that it is not true (for, as I explained, the chance that the compelling system is reliable is negligible).

Appendix: Does this argument pull the ground out from under determinism?

Up to this point I have dealt with the validity of the determinist's conclusions. But there is a common claim (which I myself have raised more than once) that this also pulls the ground out from under the deterministic claim itself. In this appendix I will try to examine that in greater detail.

I argued here in general that if I reached conclusion X while holding a deterministic worldview, I should abandon that conclusion (that is, not trust it).[4] A few weeks ago I dealt with the question of what happens if, instead of X, we place determinism itself. When a person reaches a deterministic conclusion (that is, that we are a system of mechanical calculations, and that we have no free will and no judgment), does this pull the ground out from under that conclusion itself? Seemingly yes, for he holds that he himself is compelled to think that he is compelled, and as I explained here, there is no value to conclusions that are forced upon us. Ostensibly, that includes the deterministic conclusion itself.

But specifically regarding this particular X (the deterministic claim itself), the situation appears different, since even if we are indeed compelled to that conclusion, that itself would mean that we were right—namely, that we really are a deterministic calculating machine. Therefore, in this specific case, the assumption of compulsion actually supports the conclusion. In this specific case, it seems that either way the determinist is right: if he is right, then he is compelled; and if he suspects that he is not right because he is compelled, then again he is compelled. That is, in this specific case, the claim that he is compelled does not pull the ground out from under his conclusion.

To that I would say three things:

A. Even if this argument is correct with respect to the deterministic conclusion itself, that would apply only to it. But in the discussion here we are dealing with other conclusions, not with the question of determinism itself. With respect to them, I have shown that the deterministic assumption pulls the ground out from under their validity (as I explained above: they may be true, but for me, as one compelled to those conclusions, there is no way whatsoever to know that).

B. I explained in the aforementioned thread that even this discussion itself contains a mistake. We dealt here only with the question whether the fact that I am compelled pulls the ground out from under this conclusion itself. The conclusion was that it does not, because either way there is compulsion. But a person who reaches a deterministic conclusion must take into account two possibilities: 1. determinism is false and he erred. 2. determinism is true and he was correct. The argument above deals only with possibility 2 and shows that it is indeed consistent. But there is also possibility 1: that determinism is false and he erred. Moreover, even if he is right that he is compelled, the truth of that conclusion is only accidental. The conclusion that he is compelled does not follow from the fact that he thinks so (for if he is compelled, his conclusions have no significance); rather, he happens to hit the truth by chance. Therefore, either way, the fact that he infers determinism carries no weight in the discussion. We are left in exactly the same position we would have been in without his thought: the two possibilities, that he is mistaken or that he is correct, remain as they were and with the same degree of probability (as stated, the fact that he thinks so has no weight whatsoever). The bottom line is that the loop argument does not prove that he necessarily erred (for there is no internal contradiction here); it only shows that the fact that he thinks this has no weight and no significance in the discussion. Even if he reaches the deterministic conclusion, the doubt remains exactly as before.

C. Now I will add that even beyond this, in light of point A it follows that at least all his other conclusions (apart from determinism itself) are presumed false. But now it is important to note that his deterministic conclusion, too, is based on additional conclusions (such as the principle of causality, the laws of nature, induction, and so forth), and those are already presumed false. If so, point A pulls the ground out from under the deterministic conclusion itself as well, and brings us back to the situation in which the deterministic conclusion truly saws off the branch on which it is itself trying to sit. At most, the determinist can infer ad hoc that the compelling system misleads him in everything except the question of determinism itself. But that is an ad hoc hypothesis with no plausibility whatever (why should this question in particular be the exception?). Moreover, even if this is true, it is only a hypothetical possibility. He himself has no way whatsoever to know that (for he is compelled to his conclusions). And further, as I explained above, in such a situation he may be right, but only by chance (regardless of the fact that this is what he thinks), and therefore he himself has no way of knowing that he is right. He should ignore what he himself thinks (if, from the standpoint of the compelling force, he can manage to do so, of course :)).

[1] See the appendix at the end of the post for a discussion of the deterministic conception itself.

[2] Incidentally, this itself is a demonstration of the error of calculation in the deterministic picture. For the questioner here is basically assuming that I perform a calculation, and then he asks why I ignore the possibility that my conclusion is a product of my desires. Those can create the appearance of a conclusion. Here we have an admission against interest that conclusions arising from calculation are inadmissible. Only the existence of a free control system (which is not a mechanical calculation, but judgment) makes it possible to escape these queries.

[3] See also at the beginning of the first booklet, in the discussion of Anselm's opening prayer.

[4] And do not ask me what the determinist's attitude toward this consideration itself is supposed to be (namely, that one must abandon every conclusion we have reached).