A Systematic Perspective on Free Will

With God's help

From the book “To Know Your Way in the Land”

Introduction

In this article I would like to discuss the question of free will. The dispute surrounding it has raged with full force ever since Greek philosophy, and in our day it has intensified greatly, though from a different point of view. From the period of Greece until a few decades ago, the dispute over this question was perceived as a philosophical debate, and therefore the arguments for and against determinism were a priori conceptual arguments. By contrast, in the last generation, with the development of neuroscience, the question has begun to appear as a matter accessible to empirical observation, that is, one that can be examined with scientific tools. As of today, quite a few researchers in this field think the debate is already over, since determinism is for them not only a philosophical assumption but a scientific claim belonging to neuroscience. As we shall see here, this question does not seem to have significant direct practical implications, but there is no doubt that it has far-reaching implications for our consciousness regarding the image of the human being and the psychological and ethical attitude toward him. Almost every ethical consideration we make depends on the starting point: is the human being an inseparable part of his surroundings, or is he a creature containing something beyond the rest of creation? A deterministic and materialist conception (see below on the connection between them) sees the human being as just one ordinary creature among the rest of creation, and from such a perspective it is difficult to accept that he deserves a special moral status, or to demand of him special moral obligations.

Following my book, The Science of Freedom (Yediot Sefarim, 2013), which is devoted to free will and neuroscience, I was invited to lecture on this topic in various places, including academic neuroscience research centers. Again and again it became clear to me that many people, including leading and prominent researchers in this field, are unaware of the conceptual nuances and the most basic philosophical principles of the issue, and therefore draw hasty conclusions and formulate their position on free will along a path riddled with errors. The minority among them who advocate libertarianism feel defensive, as though this were some kind of religious belief confronting scientific findings. The impression one receives is that in many of these research centers determinism is not only a scientific conclusion but almost an article of faith. As I will try to show here, it is certainly not a scientific conclusion but at most a widespread and unfounded a priori assumption. As for articles of faith, the prophet Habakkuk already said (2:4): the righteous shall live by his faith.

It is important to note that at least since Kant and humanism, human freedom and autonomy have been a fundamental part of the modern person's conception of himself. In fact, they are the deepest and most widely accepted basis of the major ethical conceptions prevalent in our world. At the same time, however, within the scientific camp, the deterministic-materialist approach has been spreading in recent years, and modern people naturally feel committed to the findings of science. The situation that has arisen is that many people in our world, mainly in the more educated strata, live with both of these conceptions simultaneously. They speak with pathos about humanism, about human dignity and distinctiveness, and at the same time about the human being as a lump of physico-biological matter, whose actions are predictable and fixed, without any capacity to decide and choose. It seems to me that there is here a kind of repression, and it finds expression at times in absurd apologetic theses (compatibilism is an excellent example; see below), which try to offer a conceptual and philosophical platform that reconciles these two contradictory poles. These conceptions, which in essence contribute nothing to the discussion apart from word games, at least allow those who feel this justified distress to live comfortably with the absurdity at the foundation of their worldview.

In this article I would like to deal with this issue in a systematic and conceptual way, and therefore I will do so almost without sources and quotations, and without describing the full range of approaches to the topic of free will (most of which are nonsense). I will describe here the libertarian worldview (which upholds human freedom of will and choice), together with its meanings and implications, and I will confront it with materialist determinism. These are the two significant poles on the table, and they call for a decision. In the process I will show that even today this question cannot be decided by scientific tools. It was and remains a philosophical question, and that is most likely how it will remain, at least in the foreseeable future.

From the above one can see that this discussion proceeds on three planes: the philosophical, the scientific, and the Torah plane. My methodological assumption is that if there truly is a reliable scientific finding one way or the other, that ought to determine our position. If scientifically the question is open, then we may move to the philosophical plane and examine the conceptual and a priori arguments for and against the two conceptions. In that framework I will also try to give the reader several diagnostic tools that may help him clarify what his position really is. Finally, I will also conduct a brief Torah inquiry (what is the position of the Torah and the halakhic tradition on this issue). But first of all we must carry out a conceptual clarification: what exactly is free choice, and what is determinism? As we shall see below, once these concepts and the differences between them are clarified, most of the arguments on both sides lose their force.

The discussion here must be brief, even in places that would merit greater length. I will therefore occasionally refer the interested reader to my aforementioned book, The Science of Freedom, which elaborates on all these topics in greater detail.

A. The Conceptual Plane

Materialism and Determinism

Materialism is a conception that sees the world as a purely material totality, and nothing more. It does not recognize the existence of souls or spirits, and usually not the existence of God either. At the opposite extreme from materialism stands idealism, which sees the world primarily as spirit and specifically denies the existence of matter. Since that is an esoteric approach, I see no need to grapple with it. What really stands opposite materialism is dualism, which sees the world as composed of matter and spirit, and in particular sees the human being as such a composite. The intuitions of us all, as can also be seen in the way we speak about ourselves, are plainly dualistic, but various philosophical and scientific considerations lead quite a few people to abandon those intuitions and adhere to materialism.

The question with which this article deals is the question of determinism. Determinism is the view that everything that occurs in the world had to occur, and in fact is uniquely determined by the state that prevailed at the preceding moment. As Laplace said: give me the present state (and enough computational power), and I will tell you the future until the end of days. Opposed to determinism stands libertarianism, the view that at least human beings have the ability to act in a way not wholly dictated by the laws of nature. This approach upholds freedom of will, and usually the discussion concerns human free will or free choice. Very few people speak about freedom of inanimate objects, plants, or animals. Once again, intuitively all of us feel that we have some freedom to choose our actions and values. But theoretically this is a very difficult position, because it accepts the known fact that the human being is a creature made of matter, and as such is subject to the laws of material nature, and yet claims that there is in him something (free will) that enables him to act in ways not dictated by the laws of nature.

There is an obvious tendency to link the question of determinism with the question of materialism. In the conception prevalent in science today, materialism and determinism are bound up with one another. If the whole world is matter, then its conduct is subject to the laws of physics, which are deterministic (with qualifications, see below). But it is important to say at the outset that this connection is not logically and conceptually necessary, although there is certainly a substantive connection between these ideas. There can be a materialist conception willing to accept degrees of freedom within the laws of physics, that is, to give up determinism (libertarian physicalism). Conversely, there can be a dualist-determinist position, which accepts the existence of a nonmaterial substance (spirit) and holds that its conduct too takes place within a rigid system of spiritual laws.

Moreover, between these two axes one can define quite a few approaches, and almost every work that deals with the subject presents a fairly detailed map of them.[1] For example, some speak of epiphenomenalism, that is, a model in which the spirit is acted upon by the body but does not act upon it. Other approaches speak of parallelism, which sees the two complexes as if they proceed in correspondence but without any interaction between them (this is the theory to which Leibniz devoted what is called the “Monadology”), and more besides.

For the purposes of the discussion here, I will address only two poles. On the one hand stands the materialist-determinist, and essentially physicalist, conception, which sees the whole world, and the human being as part of it, as a material complex that conducts itself according to the laws of nature. As we shall see, such a conception is incompatible with freedom of will, despite repeated attempts to reconcile them. Opposed to it stands the dualist-libertarian conception, which holds that the human being is composed of two substances, matter and spirit, which stand in a two-way interaction with each other (this is called “interactionist dualism”). The spirit is free from the shackles of matter but influences it, and therefore it enables the human being to act in ways not uniquely dictated by the laws of nature. These are in fact the two possibilities that contend with one another; everything else is either empty wordplay or strange speculation that adds nothing to the discussion, and I will not deal with it here.

Before proceeding, I will present two examples of conceptual considerations that only confuse matters and add nothing on the essential plane. This will also illustrate my claims about these theories, but there is also a need to clear them off our discussion table.

Influences from Within: Compatibilism

First, let us briefly examine the approach known as compatibilism.[2] In my estimation this is the most common view today among philosophers and brain researchers, and as noted many of them feel that it allows them to remain humanists and to use ethical humanist discourse together with a rigid deterministic-materialist scientific worldview.

Schematically speaking, compatibilism holds that there is no contradiction between free will and determinism. Compatibilists accept determinism, according to which everything that occurs at the next moment occurs necessarily and is uniquely fixed by the states that prevailed up to the present moment. They accept that the future is a necessary derivative of the past, and in principle even calculable. And yet they do not see this as contradicting the claim that the human being acts freely. A person who acts freely, on their view, is someone whose actions are not dictated by a factor external to him. Internal factors are merely a description of the way in which the person himself makes his decisions, and therefore there is here no departure from determinism, but also no departure from the autonomy and freedom usually seen as part of libertarianism.

Without entering into details, the fundamental problem with this view is that it is nothing but a semantic game. In essence it is wholly ordinary determinism, according to which the human being is nothing more than a log, a sophisticated machine, or at most a two-legged mouse (perhaps with consciousness and a more developed intellect), with the addition of a semantic shift according to which such a state too can be called “free.” There is of course no point arguing about semantics, but there is nothing here beyond that. If we look at every implication of the dispute between determinism and libertarianism, compatibilism will be found on the deterministic side. This is of course unsurprising, for it is simply determinism in slightly different dress. If libertarianism has ethical or other implications, they will never be relevant to the compatibilist model. The compatibilist “solution” is like someone proposing to solve the liar paradox (or any other paradox) by calling the paradoxical sentence a “true statement” and thereby ceasing to be troubled by it. Alternatively, one can prohibit the expression of the problem (that is, construct an artificial language in which it cannot be expressed).[3] As noted, this is a kind of semantic-psychological sedative, and nothing more.

Influences from Without: Brain Plasticity

In the context of brain research one finds an attitude that is the exact opposite of the compatibilist one. There it is customary to speak of a deterministic picture of the human being only if he is operated by internal factors. By contrast, when one reaches the conclusion that the brain is plastic, that is, capable of changing over time, even if that change in fact occurs deterministically through external factors (education, environmental influence—human or nonhuman—and even illnesses and the like), then supposedly one has departed from determinism. In the terminology common there, this is a non-deterministic picture, and it is quite surprising to discover how much the discovery of brain plasticity serves as a basis for adopting libertarian conceptions that continue ethical humanist discourse without batting an eyelid, even though we are dealing with complete and perfectly ordinary determinism.

It is important to understand that in order to be a libertarian, it is not enough to claim that education or environment influence the person, that is, that the state of his brain when he lives and finally dies is not exactly the state his brain was in when he was born. The important question is the nature of that influence: if it is a mechanical influence (the environment and educators are in effect programming him), then we are still within the deterministic picture. By contrast, if the influence passes through decisions that the person makes freely (he may of course be influenced by the environment in a non-deterministic way), then we are dealing with libertarianism. Later we shall define this more precisely.

To sum up, as long as human conduct is not the product of a free decision that is liberated from the shackles of external or internal influences (see below for a more precise definition), innate or acquired, we are dealing with determinism. One should not confuse terminology used in specific contexts with the terminology used in the philosophical issue of free will. It is important to examine the contribution of any proposal such as compatibilism or brain plasticity in terms of its implications. In section B we shall see that in fact the main implication (and perhaps the only one) of this dispute is moral. Therefore, in the philosophical-scientific discussion we must primarily examine whether the proposal under discussion enables us to impose moral and legal responsibility on the human being or not. Everything else is wordplay and shifts in the meanings of concepts, and nothing more.

So what, after all, are we talking about here: what is libertarianism?

If libertarianism does not accept conduct that is the product of external influences, and also not of internal influences (the person himself), then what is happening here? What, then, affects or generates our conduct? Are we speaking of something random, causeless, released from every scientific or other description? The answer to some of these questions (not all of them) is to some extent yes, but this requires clarification.

The first remark is that the libertarian does not claim that the human being is unaffected by various factors, external or internal. His claim is that the person is not always affected by them, or that the effect is not complete. In other words, he is not determined by them but at most influenced by them. Second, the libertarian also agrees that sometimes a person acts under the influence of irresistible drives, and then he is not choosing. Third, he agrees that sometimes a person acts routinely, without making a choice and decision, and then again he is not choosing. Fourth, the libertarian fully agrees that the laws of physics act on the human body, and on the brain as well. Thus the libertarian does not deny environmental and other influences on human action. His sole claim is that the sum total of all these influences does not uniquely determine the conduct itself. He claims that human conduct is not a simple sum of all of these.

One may liken this to movement within some topographical configuration. When a small ball moves across a surface with a given topography, its motion is uniquely dictated by the layout within which it moves. It will tend toward minimum energy, and the forces acting upon it will dictate its path. When it encounters a valley it will roll into it, and it will never climb a mountain (unless it had prior momentum). But what of a human being moving within such a layout? Clearly he too will feel the various forces trying to bring him to a minimum-energy state. The difference between him and the little ball is that the human being weighs the totality of the forces acting on him, but in the end decides for himself whether to flow with them (toward minimum energy), or to move in another direction (to climb the mountain rather than slide down into the valley). As noted, the environment certainly does influence him, but it does not uniquely determine his conduct.

This topographical layout is, for our purposes, a metaphor. The totality of all the influences, external and internal, that act on a human being (genetics, the laws of nature, the structure and state of the brain, environmental influence, education, culture and values, and more) creates some topographical layout within which the person acts. This layout certainly influences his conduct, and tries to push him toward certain forms of behavior or to repel him from moving in other directions. But unlike the little ball, the human being can choose whether to comply with the dictates of this topography or to act otherwise. That final decision is not the result of any influence or factor, since all the influencing factors are situated within the surrounding topography. The decision is completely free, though it weighs and takes account of all those influences as well. Clearly, given a different topographical complex, the conduct may be different. The same person, with the same degree of love for scenic lookouts, when standing before a small hill will probably decide to climb it in order to view the landscape. By contrast, when he stands before a high and difficult mountain, he will tend to remain below and not climb it, even if it is the same landscape. Does that mean his decision is dictated by the environment? Not necessarily. It is influenced by the environment, but in the final analysis it is his decision.[4] This is a schematic description of the libertarian conception.

Three Models: Randomness, Determinism, and Free Choice

If we summarize what we have seen so far, we can speak of three mechanisms of conduct:

• Deterministic conduct, according to which every state is a necessary result of previous states. The circumstances that prevailed at the preceding moment (the topography and the person’s location within it) dictate the next moment. The person is a kind of little ball.
• Random conduct, according to which the person acts in a manner completely released from influences. This may be described by a lottery model. When the person has to decide what to do, he conducts a lottery and acts accordingly.
• Free choice, according to which the person’s conduct is influenced by the topographical layout, but not determined by it. The decision is made through his own free judgment, which also weighs and takes account of the constraints and influences of the environment.

It is important to understand that, contrary to common conceptions in discussions of determinism, the last two mechanisms are not identical. In both of them there is no cause that dictates the outcome, but in the second mechanism we are dealing with sheer arbitrary randomness, whereas in the third we are dealing with judgment that seeks to realize the person’s values and goals. True, the conduct is not the mechanical result of a cause that produces it, but it certainly should not be viewed as random conduct. One might say that in the first mechanism conduct is determined by causes; in the third, conduct is in light of purposes, or for reasons, but not out of causes (the person can decide to climb the mountain even if that direction runs contrary to the forces acting upon him and does not lead him to minimum energy). And in the second we are dealing with aimless conduct, based neither on causes nor on purposes and reasons. In the second mechanism there is no judgment producing the decision; in the third there is.

The Elections Parable

This threefold distinction can be sharpened by means of a parable. Imagine free political elections in which a person enters the polling booth, freely chooses a ballot slip, and drops it into the ballot box. The ruler is elected by majority vote. These political elections serve here as a parable for the choices of the individual person in his private life. Let us now consider three different applications of this model:

• In Syria all this proceeds as described above, except that in the polling booth there is only one kind of ballot slip. This is, of course, illusory freedom. It is determinism. Despite the illusion of freedom, the current state uniquely dictates the result.
• In Switzerland this indeed takes place with several kinds of ballot slips, but in our metaphorical Switzerland there is no problem that needs solving. It does not really matter who the ruler will be, because nothing depends on it. Here the elections are equivalent to a lottery, for one could in fact simply conduct a wholly random draw and obtain a reasonable outcome. There is no consideration of purposes toward which the voters are striving, and they function as a large lottery machine. Therefore these “free” elections actually express randomness, not freedom. There are neither causes nor purposes and reasons for the choice, and therefore this is not really a choice but a lottery.
• By contrast, in Israel there are also real problems awaiting resolution. In the polling booth there are several kinds of ballot slips from which the voter may choose, and he chooses one of them and drops it into the ballot box.[5] The results will largely dictate policy, and they will have costs for better and for worse. Here there is no cause that dictates the result (the elections are free), but there is judgment in light of purposes, and therefore this is not randomness. This, and only this, is the situation that parallels free choice, or free will.

These three models parallel the three mechanisms described in the previous subsection and help sharpen the difference between the last two.

B. Discussion on the Philosophical Plane: Arguments and Implications

In the fourth chapter of my book I discussed several a priori arguments at length. Because of lack of space, I will limit the discussion here to three in favor of determinism and three against it.

For Determinism 1: Peter van Inwagen’s Dilemma Argument

One of the most common arguments against determinism is usually cited in the name of the philosopher Peter van Inwagen.[6] This is a dilemma argument that can be presented as follows. Suppose I am now standing at moment t, and before me are two possibilities: to do X or to do Y (Y may also be refraining from X). Libertarianism holds that both possibilities are open to me, and I choose between them freely. Suppose I chose to do Y. In principle, God can run the “film” backward, return me and the world once again to the point t, and place me again at the same crossroads, and there may be situations in which I decide to do X. But the state of the world up to moment t is identical in both cases; if so, what caused me once to choose X and once Y?

Van Inwagen presents here a dilemma argument. If there was something different in the world, or within my inner self, that caused the difference, then there were some circumstances from which my choice was derived, meaning that it was not really a free choice. And if the state in both cases is indeed identical, and yet two different outcomes issue from that very same state, then my decision was made without a cause, which is to say that it was random. Thus, either way, whether there is a prior cause or not, my decision is not the result of free choice. Either we have randomness or we have deterministic determination.

It is easy to see where the flaw lies in this popular argument. In dilemma arguments, the flaw usually lies in the hidden dichotomy they assume. In our case, the argument covertly assumes that there are only two possibilities: causality or randomness. Van Inwagen’s assumption is that if there is no difference in the state at moment t, and yet on different occasions the same person may choose differently, then we must be dealing with randomness. But this is a mistake, or at least a begging of the question. The dispute between the libertarian and the determinist is precisely about this point. The determinist claims that the opposite of determinism is nothing but randomness. He does not recognize the existence of a third mechanism. The libertarian, by contrast, holds that there is a third mechanism as well: free choice (see above at the end of section A). The libertarian’s claim is that the “mechanism” of free choice is non-deterministic on the one hand, but also non-random on the other. Van Inwagen’s argument does not prove that this is false; rather, it assumes it and uses that assumption in order to establish itself. It seems that a philosophical misunderstanding lies in the background, because van Inwagen thinks that libertarianism’s innovation is the claim that something can occur without a cause, rather than the claim that there is a third mechanism. The libertarian thinks that indeed there is nothing in the situation at time t that necessarily determines the outcome of the choice, but such a situation need not express randomness; it may just as well reflect free choice (if the action is done out of judgment and not by lottery).

As an aside, I should note that van Inwagen’s original argument is in truth somewhat more complex.[7] He speaks of a statistical experiment of repeating the same choice a thousand times in the same situation, producing a statistical distribution of choices of X or Y. The fact that nothing in the preceding situation determines the outcome, together with the fact that a statistical distribution of choices is produced, indicates to him that we are dealing with randomness. But this formulation too does not hold water. From the description at the end of section A it is easy to see that even a statistical distribution of outcomes, and even if it repeats itself on average in every thousand experiments we conduct (say we always get 300 choices of X and 700 of Y), still does not necessarily constitute evidence of a random process. In the third chapter of my book I explain that even a collection of free choices generally distributes statistically like a collection of random events. The libertarian thinks that all psychological and educational research is based on this assumption.

A Broader Look at the Fallacy of Begging the Question in Discussions of Free Will

As we saw in van Inwagen’s argument, most perplexities and doubts concerning free will stem from the need to ground free will upon one of the other two mechanisms. Even when this is done in different formulations, the central question is basically what causes us to decide as we did. A free-will decision is made without a cause, and this leads many to doubt that such a thing is possible. Since we do not find a cause for our free decision, it would seem that we must reject our intuition about it, or else view it as randomness. But as noted, these are three different mechanisms, and therefore it is no wonder that none of them can be grounded in the other two. All these doubts suffer from the same begging of the question that we encountered in van Inwagen.

Moreover, arguments of this sort implicitly assume that the mechanism of free choice is the unintelligible one, whereas determinism or randomness are intelligible. Therefore they try to ground it upon them. But a second look shows that the only mechanism truly known to us immediately is precisely the third one—free will. We know it from direct self-awareness. By contrast, the philosopher David Hume already showed that the deterministic-causal mechanism is a priori assumption without empirical basis. And the “random” mechanism is also not familiar to us from our experience. Where have we ever seen anything random? The only true randomness known to us from physics belongs to the microscopic world of quantum theory, which none of us has ever seen. This is an abstract theory, and even if it is correct, there is no immediate human experience here. In our macroscopic world we have not truly encountered random occurrences. Thus both randomness and causality are abstractions lacking empirical basis in our ordinary experience. So why try at all to ground the familiar mechanism in two mechanisms that are not familiar? The entire philosophical entanglement surrounding free will seems rather strange.

For Determinism 2: Logical Determinism

Logical determinism is an argument that tries to derive determinism from a priori analysis (without relying on any facts). As Kant already established regarding the ontological proof of God’s existence, such an argument is suspicious on its face. It is hard to accept an argument that derives some claim about the world from purely a priori conceptual analysis. And yet many people become confused and see this argument as a valid one that forces us to accept its deterministic conclusion.[8]

Aristotle already pointed out that the truth value of a proposition does not depend on time. For example, suppose that tomorrow, on Tuesday (23.8.2011), a torch race will take place in Australia. In that case, the statement “On 23.8.2011 a torch race will take place in Australia” is a true statement (its truth value is “true”), because the content of the statement corresponds to reality. If no such race will occur, then the statement is false (its truth value is “false”). Aristotle argued that the truth value of a proposition does not depend on time. Even if we do not know today whether such a race will occur tomorrow, that is merely our lack of information. If a race will indeed occur tomorrow, the truth value of the statement is “true” already today, because it correctly describes reality. Our knowledge about that reality neither adds nor subtracts anything. This analysis is of course true of every factual proposition. We have assumed nothing here about torch races or about Australia in particular.

The next step of the argument is that if this statement is true today, and in fact was already true two thousand five hundred years ago, then it cannot be that when the date arrives the race will not take place. If it does not occur, then the statement is not true today, nor was it ever true. But that contradicts our assumption that it has been true from time immemorial. From this the logical determinist concludes that in any event, whatever the truth value of the statement may be, the occurrence is fixed in advance and cannot be changed. The same is true, of course, of every factual proposition on earth. The fact that I have just eaten a tomato is also a fact fixed since the six days of creation, because the statement “Michael Abraham ate a tomato on date X” was already true in Aristotle’s time as well, and therefore could not have failed to be realized today.

In the fourth chapter of my book I explained the flaw in this argument at length. The essence of the matter is that just as the truth value of a proposition is unaffected by the time axis, so all of logic is indifferent to the time axis. Therefore the existence of a truth value in the past does not determine an occurrence in the future. The race may occur or may not occur. If it occurs, then the truth value will be fixed retroactively as “true” from the creation of time. And if the race does not occur, then the truth value will be fixed retroactively as “false,” likewise from the creation of time. If any reader is troubled by backward influence along the time axis, set your heart at ease. We are not dealing with causal influence (which always proceeds forward in time, that is, the cause precedes the effect), but with logical “influence,” and as the logical determinist himself correctly assumes, logic is indifferent to the time axis.

Against Determinism 1: What Is Judgment?

In the deterministic worldview, human judgment is the product of a rigid mechanism imposed upon him, from within or from without. But we have no way of knowing whether that mechanism is reliable, and certainly no way of justifying our trust in it. The considerations that lead us to determinism are themselves exposed to the same attack. After all, they are the products of those same mechanisms imposed upon us, over which we have no control, and we have no way to examine their reliability. Any examination we make of those mechanisms will itself be made by means of those same tools. For there is nothing in us beyond those mechanisms, and there is no “other” here that performs autonomous judgment. Thus determinism as a philosophical thesis saws off the branch on which it itself sits. I should note that the claim is presented here schematically and not fully, and in order to exhaust it and understand that this is not the usual skeptical claim, one must examine several philosophical twists connected with it. The interested reader should turn to the fourth chapter of my book.

Against Determinism 2: Newcomb’s Paradox

Another a priori consideration against determinism is Newcomb’s paradox. In his 1969 article, the philosopher Robert Nozick presented the following thought experiment. Two people are involved in it: the chooser and the predictor. Before the chooser are placed two boxes: one is open and contains 1,000$, and the second is closed and contains either 0$ or 1,000,000$ (the chooser does not know which). The chooser must decide whether to take the contents of both boxes together, or only the contents of the closed one. Here the predictor enters the picture, and for the sake of the discussion we assume that his predictive power is perfect. The predictor knows in advance and with certainty what the chooser will do, and accordingly he prepares the contents of the closed box one day before the chooser makes his decision. Since the predictor wants to reward one who is satisfied with little, he adopts the following tactic: if the chooser is going to take both boxes, he makes sure in advance that the closed box contains no money at all. But if the chooser is a person who is satisfied with little and is going to take only the closed box, the predictor rewards him by placing a million dollars in it.

The chooser himself knows the predictor’s policy and abilities. The only thing he does not know is this one detail: what the predictor predicted about the choice he is about to make, and consequently he also does not know what the predictor actually put into the closed box. The question is what strategy the chooser ought to adopt. Should he take only the closed box, or both?

Seemingly, he ought to take only the closed box, for then he will win a million dollars, whereas if he takes both he will win only a thousand (because the closed one will be empty). But on the other hand, the closed box is already before him now, and what is in it was fixed yesterday and will not change if he acts differently. If so, why should he not take the closed box as he wanted, and afterwards also take the open one with the additional thousand dollars? After all, his taking the second box cannot retroactively change the contents of the closed box lying before him (our assumption is that the predictor is all-knowing but not necessarily all-powerful). It seems that taking both boxes is the winning strategy in every case. That way he gets the contents of the closed box plus another thousand dollars. This is always better than the contents of the closed box alone. But if the predictor is truly all-knowing, he will ensure that in such a case the closed box is empty. If so, when the chooser takes both, he will win only 1,000$, meaning he chose incorrectly. Alternatively, the predictor did not predict correctly, which contradicts the assumption that he is all-knowing.

To understand the root of the problem, we must notice that the assumption that there is such a predictor contains within it a disguised deterministic position. In the libertarian picture, one cannot speak of such a predictor, because one cannot know with certainty in advance what the chooser will decide. If the information does not exist now, how can there be someone who knows it with certainty? “To know information that does not exist” is a self-contradictory expression; to the same extent, a predictor who predicts in advance the decision of a person who chooses freely is an oxymoron. Predictive power is defined only where the information exists in principle, that is, where it can be derived from current circumstances. But in a libertarian picture the information simply does not exist now, and therefore no one, however great his predictive power, will ever be able to predict the future. The libertarian will respond to Nozick and Newcomb that there is no such predictor. The concept itself contains an internal contradiction, and therefore the problem does not arise at all. The libertarian will of course take both boxes.

By contrast, in the deterministic picture the difficulty remains in full force. There the information does exist. Even if no one can reach it (because the calculation is very complex and difficult; there is no computer powerful enough to predict it), there is in principle no obstacle to the existence of a predictor who knows it. The concept contains no internal contradiction. It follows that Newcomb’s paradox is in fact an a priori argument against determinism, because if determinism is correct then the information really does exist, and such a predictor is possible in principle—and then we enter a logical loop. According to the determinist, the chooser must make a patently irrational decision (not to take the thousand dollars lying before him, and to settle for the closed box alone), or alternatively believe in backward causal influence in time (that taking the box will retroactively change its contents). Which proves, in and of itself, that determinism is manifestly unreasonable.

To exhaust the discussion we would have to continue through several further twists in the plot (such as what “making a decision” means for the determinist), and I cannot do so here. I refer the interested reader to the fourth chapter of my book. For the purposes of the discussion in the final section I will only note that if human beings do indeed have free choice, then such an all-knowing knower cannot exist, even if it were God Himself. This is a logical contradiction, regardless of any assumption about the nature of that all-knowing predictor.

For Determinism 3: The Problem of Causality

I will conclude the philosophical discussion with the main a priori consideration concerning determinism, namely the frontal contradiction between libertarianism and the principle of causality. The fifth and sixth chapters of my book are devoted to this, and here again I will be forced to be brief.

The principle of causality states that every event in the world has a cause. If something happens in the world, the assumption is that there was something that brought it about. A ball begins to roll—apparently someone or something applied force to it. A sound is heard somewhere—we are certain that there was some source that produced that sound. A war breaks out, a political decision is made, rain begins to fall—to all of these we relate as outcomes brought about by some causes. We have no thought at all that any of this might have happened by itself, without any cause. In my book I showed that this is an a priori insight and not the result of observations, but every rational person acts on the basis of this assumption.

On the other hand, libertarianism assumes that there are occurrences that happen without a cause. It holds that given a fixed system of circumstances (internal state and external circumstances), a human being can act in more than one way. Even if the laws of physics determine that he will do X, he may decide to do Y. But the principle of causality states that every occurrence has a cause, and given the cause, the effect should be uniquely determined. Therefore libertarianism appears flatly to contradict the principle of causality.

Let us illustrate the problem raised by libertarianism by means of a description (somewhat schematic, but sufficient for our purposes) of a human decision that ends in a mechanical action. When a person decides to perform a physical act, there is a chain that leads from the decision to the mechanical action that implements it. For example, when a person decides to walk, his leg begins to move. Walking is a physical-mechanical occurrence, and this means that objects with mass are moving (the legs, and with them the whole body). According to the laws of physics, objects with mass do not simply begin to move; there must be a force acting on them. According to Newton’s second law, in order for an object with mass to accelerate (or begin moving), a force must be applied. Moving one stage backward, we indeed find a force that causes the movement of the hand or leg—the force that the muscle exerts on the moving limb. But now we must retreat one further stage: what moves the muscle? After all, it moves, contracts, and relaxes, and apparently that too must have some physical cause. The physiology of muscle action teaches us that the muscle moves because of an electric current that reaches it through the nervous system. Thus again we are dealing with a standard physical process: the movement of the muscle is caused by a physical cause that produces it; the muscle converts the electric current into mechanical motion. But the electric current that moves the muscle is also a physical event, and therefore it too was clearly caused by something. There must be a physical cause that creates this current. Electric current is the movement of charged particles (electrons), and what moves such particles is usually an electric field. We now know that the mechanism that sends the electric current to the muscle occurs in the brain. In the brain there is a physico-chemical process that ultimately creates an electric field that creates the electric current which reaches the muscle through the nervous system and contracts it, and the muscle moves the hand or the leg, and thus we move. Up to this point everything is standard and consistent with the laws of physics.

If so, we have now reached in our chain the question of what generates the neural activity (the electric currents and fields) in the brain, the activity that begins the whole chain of decision we have described so far. If one traces the whole way back within the human brain (which is of course far more detailed and complex than I have described here), one must eventually arrive at some event that stands at the beginning of the chain, and is the initial cause of the whole process. Somewhere all these processes begin. What happens there? Does something there occur out of nothing? The two answers that are offered to this question lie at the root of the dispute between determinism and libertarianism.

The Deterministic Picture

The determinist claims that the initial cause is an event that occurs outside the human being. Some event begins the internal chain we have described, which ends in a mechanical event (the movement of a limb). For example, someone threatens me, and I identify this by means of rays of light reaching the retina of my eye. My brain processes this visual information (this is a physico-chemical action; we shall later come to the question of the mental), and in response creates an electromagnetic field that generates an electric current, which eventually reaches the muscles of my hand, which rises and strikes the attacker or defends against him. This is the deterministic description of the human “decision.” Determinism paints a picture of the human being as an automatic machine driven by external causes. The entire internal chain I described above is nothing more than part of a long causal chain that began with the Big Bang, reached the person who threatened me (who of course also went through processes similar to those described here), continued through my brain processes, and ended (for the time being) in the movement of my hand to hit him or defend against him. Those actions too will of course continue to be further links in the chain of mental and physical events, in me and in others, a chain that will apparently continue until the end of days.

What will determine whether I defend myself, strike first, flee, or perhaps restrain myself and stand still? The external stimulus is one and the same, yet different people react to it in different ways. The determinist claims that the factor is the current structure and state of the person’s brain, which are themselves of course products of previous events he has undergone and experienced, and of course also of his genetics. Different machines can produce different outputs for the same input, and therefore different people, or the same person in different states, react in different ways. There is no choice here, only a wholly mechanical action. Despite the difficulty of seeing a human being as a kind of machine, this picture would seem necessary. It is the inevitable conclusion that emerges from applying the laws of physics and the principle of causality.

The Libertarian Picture

How can the libertarian hold an alternative picture without denying the principle of causality? Does he think that there can be a physical action that has no physical causal factor? And more concretely: does he think that an electron in the brain can move without a physical force field moving it? After all, the electron in the brain is no different from an electron anywhere else in the world; it is an ordinary material entity, just like electrons and other material entities in the universe, and therefore the same physical laws govern it.

From the description so far it follows that the libertarian has no escape; he must adopt the possibility that the electric field that moves electrons in the brain during a free decision is created without a physical cause. The will, or the human decision, is what begins this chain, and from that point a new physico-chemical chain begins, such as the one described above.

There is no need to mention that the laws of physics forbid the creation of a field without a physical cause (electrical charges) that generates it. Here we have reached the main weak point of the libertarian conception. It necessarily leads us to accept the possibility that force fields can be created without sources (charges), contrary to the laws of physics. We find such a possibility very difficult to accept, and this, then, is the fundamental motivation for embracing a deterministic conception.

If libertarianism is based on the strong intuition of freedom that each of us possesses, we can now understand that determinism too is based on an a priori intuition no less strong. These are two a priori intuitions, and neither has an observational basis. This is the root of the dispute between determinism and libertarianism. The contradiction with the principle of causality is the main argument in favor of determinism, and on this point the matter rises or falls.

A Possible Libertarian Solution: Between the Principle of Causality and the Laws of Physics

It is important to distinguish between two terms that I have used thus far somewhat carelessly. When I say that an electron moves, or that a field is created, out of nothing, the intention is that this happens without a physical cause. This does not mean that the libertarian holds that there is no cause at all. There certainly is a cause: the human will, that is, my decision to act. Thus we are not dealing here with an occurrence out of nothing, but with something arising from something—except that the causing thing is not physical. This can be formulated as follows: libertarianism contradicts at most the laws of physics, but not the principle of causality. In the libertarian picture, the physical event occurs by virtue of a cause, but it is a non-physical cause.

But if we now continue one step further, then apparently there is nevertheless something here that contradicts even the principle of causality: the coming-into-being of our will or decision itself. What causes that event? Here there is no cause at all (only a reason, or a purpose), for we have already seen that any cause, physical or not, contradicts free will.

We are thus distinguishing between the principle of causality and the laws of physics. The principle of causality states that there must be a cause, but it does not require specifically a physical cause. Nor does it distinguish between one physical occurrence and another. By contrast, the laws of physics deal only with physical objects, and from their point of view specifically a physical cause is required, not some other kind. Having made this distinction, we can now return to the chain of decision described above, which speaks of the arising of a will, which creates a physical force field, which begins to move particles and ultimately to create mechanical actions.

Libertarianism makes two claims: (a) the emergence of the will is not a physical event, and therefore its spontaneity does not clash with the laws of physics. On the other hand, it does seem to conflict with the principle of causality. (b) The creation of a physical force field by the will is a physical event that has a non-physical cause. This stage does not contradict the principle of causality, but it certainly does contradict the laws of physics (because they require specifically a physical cause, and the will is not a physical cause).

Here we reconnect with the picture described at the end of the previous section. As we saw there, the mechanism of free choice is essentially teleological, that is, conduct directed toward a purpose or goal, and not conduct arising out of a cause. That is precisely what characterizes the free human will (hence it is called “free”). It is therefore no surprise that we find the same thing here as well. The arising of the will is an occurrence that has no cause but does have a purpose. The will is directed toward some goal (when I want X, this does not mean that I “want because,” but that I “want in order to”). The interactionism defined in the previous section underlies the effect of the will upon the physical chain that comes immediately after it. The will creates an electric field, which moves electrons and begins an ordinary physical process. Thus spirit moves matter and acts upon it. I should note that this whole process can preserve energy, because the physical actions use energy that the body stored in the past (sleep, food, and the like). The total energy that the body expends is taken from what it had accumulated, and the rest remains stored within. In the libertarian picture, the law of conservation of energy is not necessarily violated.

What we have seen here is in fact the uniqueness of human will. In the libertarian picture, human will is a highly unique capacity. It succeeds in translating a future goal and turning it into a cause for a physical process (with the help of the energy stored in the body). My desire to meet a friend tomorrow creates an electric field in the brain that moves electrons and causes my hand to move toward the telephone receiver, and then a conversation with the friend in which a meeting for tomorrow is arranged. The meeting tomorrow, which is the goal of the process, becomes through the will a force field that constitutes a physical cause that activates it.

In summary, it is true that particles do not move if no physical force acts on them, but we have no experience teaching us that force fields cannot be created by a mental capacity such as will. We know that electric charges create an electric field, but there is no a priori reason to assume that human will cannot create an electric field as well. Of course, if we assume that a force field cannot arise without a physical source, the conclusion will be deterministic. But this is again a begging of the question, for this is precisely the point under dispute. The libertarian claims that the human will is exceptional and special, that unlike every other physical situation it can take energy stored in the body and create force fields from it without charge sources.

The Idler Argument

Strangely enough, it is very difficult to find practical implications of the deterministic-libertarian dispute. Let us take as an example the idler’s argument, which says that if determinism is correct and everything is dictated in advance, then there is no reason or need to decide or do anything. In any event, everything will happen. This is a mistaken argument, for according to the determinist the decision itself is also dictated in advance (we have no ability not to decide, or to decide otherwise). Moreover, if we do not do something, it will not happen, and therefore for it to happen we must act. If we need to study for an exam in order to succeed, one cannot refrain from studying on the grounds that success on the exam is fixed in advance. Without studying there is no success. At most one can say that both the studying and the success are jointly fixed in advance. And even if I decide to loaf, that is not really a decision but a result known in advance.

Against Determinism 3: Moral Responsibility

Some argue that within the deterministic picture there is no point in judging people, morally or criminally. For example, today in courts, when it is shown that a person acted on the basis of an irresistible impulse, he is exempted from criminal responsibility. In the deterministic picture this has no meaning at all. A person always acts from impulses that are irresistible (internal or external; we have already seen that this makes no essential difference). Imposing moral responsibility on a person who acts within a deterministic framework seems very problematic. This is the main point that arises in the contexts of the discussion of determinism and its implications.

And yet, at least on the practical plane, questions of judgment and moral and criminal responsibility are also somehow resolved, and in fact all these simply receive a different meaning, but practically they are supposed to function in exactly the same way. First, the judgment by which we judge others, and our imposition of responsibility upon them, are themselves events imposed upon us, just as the other person’s act was imposed upon him. Therefore there is really no point in discussing whether to impose moral responsibility or not. Moreover, even if we insist on discussing the logic of all this, there is even logic in punishing and morally and criminally condemning a person for his bad deeds, since the punishment itself can later lead him not to return and do such deeds. His actions are the product of a totality of causes, among which is also the desire not to be punished (or the fear of and aversion to punishment).

In the final analysis, it seems that almost every ordinary human thought or behavior can also be justified within the deterministic picture. True, all our most essential activities are emptied of real content. They become a kind of useful illusion, or shadows of blind forces that move us through the world like marionettes, while also making sure that we live with the feeling that we are engaged in momentous and important matters, and that the human being is the crown of creation and stands at its ethical summit, both in the conduct demanded of him and in the rights granted him. The determinist understands very well that the truth is that all this is nothing but groundless illusion. Human beings have rights and ethical status roughly to the same extent as beetles, polar bears, rocks, clouds, gravel stones, or common sea urchins. The only difference is that in human beings there also dwells the utterly groundless illusion that they possess such rights. The fact that we harbor this or that illusion does not of course mean that there is any substance to it. Moreover, we are all wise and know the Torah, and therefore we also understand very well that this is a groundless illusion, or a kind of manipulation meant to secure the benefit of safe and reasonable lives, and yet we continue to conduct ourselves as though this were a supreme truth, and as though anyone who deviates from it were a heretic to the humanist religion, and we speak of it with impressive moral pathos.

The meaning of all this is that the superiority of man over beast disappears entirely in the deterministic picture. Humanism has become an empty word, which does not in the least prevent many determinists from continuing to use humanist discourse with enthusiasm and great pathos, and even to condemn others who behave in an inhumane manner. Some derive from determinism an ideology of veganism (it is forbidden to use animals, and perhaps even inanimate things, since there is no principled difference between them and us), and others derive from it the opposite—that everything is permitted, since what can a “prohibition” imposed on a gravel stone possibly mean?! They continue energetically to discuss in the courts whether to exempt a person from moral responsibility because of lack of control or some mental illness, without being troubled by the absurdity of such considerations. But a gravel stone is not supposed to weigh considerations that have any point at all. It simply is what it is.

So the world goes on as usual, and yet something very fundamental disappears here. The main price of deterministic conceptions is the loss of humanism, despite the fact that this has no necessary practical implication. Deterministic human beings go on behaving like everyone else; among them too there are good and bad people, just as in every other human group. The only difference is that those among them who behave rationally and morally are simply inconsistent. But a gravel stone is not supposed to be consistent, or to perform justified actions. The blessed moral illusion and the blessed humanist illusion save us from the consequences of this absurd and harmful view.

The Philosophical Decision: Lex Specialis

At the bottom line, it seems that on the philosophical plane two intuitions stand opposite one another: the intuition of freedom and the intuition of causality. The determinist chooses the second and finds himself compelled to reject the first. The libertarian, by contrast, adopts the first and finds himself compelled to reject the second. So which of them is right? How should one decide between these two intuitions? Is this an arbitrary choice, a matter of taste, or can one point a priori to a substantive superiority of one over the other?

There is a simple methodological consideration that gives priority to the libertarian option. In legal systems it is accepted that when different principles collide, various doctrines of decision are applied. One of the most basic of them states that one must always prefer the more specific law over the more general one. This doctrine is expressed by the Latin maxim: Lex specialis derogat legi generali, or in short: Lex specialis. A halakhic example will clarify this. Jewish law forbids killing a person; but when someone deliberately murders his fellow, Jewish law rules that his punishment is death. There is now, seemingly, a contradiction between two halakhic norms: on the one hand a general norm stating that it is gravely forbidden to kill a person; on the other hand a specific norm stating that a murderer must be killed. Which of the two norms prevails? From the rule of Lex specialis it follows that the specific norm overrides the general one, that is, the murderer must be executed. Why? Because if we give up the specific norm, then the law that a murderer should be put to death becomes entirely empty of content; whereas if we prefer the specific norm over the general one, the general norm does not remain a dead letter in the lawbook. There is still a halakhic prohibition against killing a person, except that it has an exception (when one is killing a murderer). Therefore this is the preferred legal interpretation. We prefer the interpretive possibility that leaves both laws meaningful, and sees the specific norm as an exception to the more general norm.

It seems that this is the situation in our case as well. We have a strong causal intuition, but opposite it stands an intuition no less strong concerning freedom of will. Which of them is preferable? How can we apply the rule of Lex specialis here? From this doctrine it clearly follows that the libertarian intuition is preferable to the causal one, because even if we adopt libertarianism we have not emptied the principle of causality and the laws of physics of content; we have merely excepted the human will from them. They remain valid in every physical and other context. By contrast, preferring the causal intuition would leave our intuition concerning free will entirely empty of content.

C. The Scientific Plane

Introduction

Many thinkers and researchers try to escape the physicalist tangle described so far, but to do so within the scientific framework. On the one hand, they adopt the principle of causality without exception: the whole world is physics, and therefore it is all governed by the laws of physics (and chemistry and biology). Yet they are unwilling to give up human dignity, moral responsibility, and the possibility of judging human beings—that is, libertarianism. How can these two be reconciled (we recall that compatibilism has already been removed from the discussion)?

Such thinkers try to find free spaces within physics itself. If the laws of physics allow degrees of freedom in which human choice takes place, then we will have achieved everything we want. The whole world remains physics, and at the same time humanism and free will remain intact. It is very tempting to place free will within such gaps in a way that does not harm the principle of causality and does not require relinquishing humanist intuitions. Thus we are spared recourse to the Lex Specialis described at the end of the previous section. But is physicalist libertarianism possible at all?

For that purpose we must examine three areas in which such a gap supposedly appears within physics, and in which such an argument could therefore arise: chaos, quantum theory, and complexity theory (emergence). In this section we shall see that physicalist libertarianism cannot be built on any of these three, and therefore is probably impossible.

Chaos[9]

The first domain in which we must examine the question of a physicalist gap is the domain of chaos. During the twentieth century it gradually became clear to physicists that there are special physical phenomena whose course cannot be predicted; that is, even if the present state is known, one cannot determine what will happen later in the process. At first glance this is exactly the sort of physical gap we are looking for here.

Doyne Farmer, one of the pioneers of chaos theory and a member of the Santa Cruz group, describes his feelings when he and his colleagues began to understand that they had in their hands a theory with far-reaching physical significance:[10]

On the philosophical level, it seemed to me like a practical way to define free will, in a way that lets you reconcile “freedom is granted” with “everything is foreseen.” The system is deterministic, but you cannot say what it will do in the next moment… There is one coin with two sides here. There is order out of which randomness emerges, and one step away there is randomness underlain by order.

Farmer claims that this gap in physicalism, which enables us to find randomness and unpredictability within hard physics, provides an opening for understanding the phenomenon of free will.

But despite the excitement and the fascinating possibilities that chaos seemingly opens for us, the link Farmer draws here between randomness and free will is problematic, to put it mildly. We have already seen that free choice is not randomness but another mechanism. Moreover, in the chapter on evolution I already noted that throwing a die or a coin are phenomena in which there is nothing truly random, and yet we treat them with probabilistic and statistical tools. And indeed, chaotic systems are not really random. They have characteristics similar to randomness, and therefore it is customary to use probabilistic tools to analyze them. A chaotic system is completely deterministic, but because of its complexity and its strong dependence on complicated factors (or initial conditions), it is very difficult to predict its future behavior.

Farmer’s basic mistake is conceptual. Actions that are the result of free will are indeed unpredictable, but not everything that is unpredictable expresses freedom of the sort involved in free will. Therefore the fact that we have no predictive ability with respect to such systems does not mean that they are non-causal. Sometimes this is merely a matter of complexity and computational difficulty, and nothing more.

In order not to enter into all the details relevant to chaos theory, I will make do here with a simple example. A point particle is located at the top of a mountain with a rounded peak, exactly at the uppermost point, as in the following figure:

This is what is called in physics an unstable equilibrium. In principle, as long as no force acts upon it, the particle will remain there (theoretically, if it is point-like and located at the exact mathematical top of the mountain), but any force, however small, will topple it to one side. There is no way to know in advance to which side it will fall, because the slightest breeze can move it. Lorenz, one of the pioneers of chaos theory, already pointed out that climatic phenomena are extremely complex and very hard to predict (they are chaotic). It is therefore very difficult to predict the directions of the wind, especially in a particular tiny location.

Such a situation is chaotic, because there is here strong dependence on initial conditions (where exactly the little ball is placed, or what exactly the direction of the wind will be at that point atop the mountain), so that a minute difference in the circumstances leads to utterly different results. Therefore this is a situation about which we have no predictive ability at all, not even approximately. But can one say that there is freedom here, that is, principled indeterminacy? Any reasonable person can see that this is a fully deterministic process, in which the present state (of the entire universe) uniquely determines the future to occur (the way in which the little ball will fall), except that it is hard for us to calculate it. This is an epistemic gap (that is, a gap in our cognition or information about reality) in the laws of physics, not an ontic gap (in reality itself). But libertarianism speaks of ontic freedom, and therefore chaos will clearly not be its glory.

Quantum Theory[11]  

In the chapter on evolution I cited Pierre-Simon Laplace, who said: “Probability does not come in place of causality and real determination, but only in place of human ignorance.” And indeed, in the previous subsection on chaos we saw that this is the situation there. Probability does not express real (ontic) freedom, but at most ignorance, that is, lack of information or lack of computational power on our part (an epistemic gap). Nevertheless, during the twentieth century it became clear that Laplace was mistaken. Today we know that there is at least one exception to his sweeping statement: in quantum theory there are ontic gaps, that is, real freedom within physics. There the use of probability expresses not only our lack of information, but freedom within physics itself. Is this the physical gap we were looking for? Is this where the salvation of physicalist libertarianism will come from?

To understand why the answer is negative, we would have to describe and understand quantum theory, which would require a long, complex, and detailed discussion—much more than I can do here. Therefore again I will suffice with a very brief and simplified schematic description, but I think it is enough to clarify why quantum theory too does not provide a basis for physicalist libertarianism. Quantum theory has various interpretations, and we will use the one called the Copenhagen interpretation. But that is mainly a matter of language. The claims do not depend on the specific interpretation given to this strange theory.

In quantum theory we describe a particle such as an electron in a somewhat odd way. As long as we have not measured its position (or its velocity), it has no single definite position (or single definite velocity). The positions (or velocities) are described by a wave function that assigns a probability to each value of position or velocity. Measurement causes the “collapse” of the wave function, that is, the selection of one of the classical states (a state in which the particle has a well-defined velocity or a well-defined position). Accordingly, given the present state of the electron, we have no way of predicting what its state (its position or velocity) will be at the next moment. Moreover, if we know its position with great precision, then its velocity at that moment is “smeared” across a wider range of possibilities (this is the uncertainty principle). Some wished to claim that human measurement is what causes the wave function to collapse, that is, that it produces the value taken by the particle’s position or velocity. This psychokinetic phenomenon (psychological influence on physics) is ostensibly the physicalist libertarian’s dream come true, although it is important to note that most physicists today think this claim is false.

And yet, even here there is no possibility of housing free will within physics. First, quantum phenomena occur on very small scales. On larger scales they are “washed out,” and in effect disappear. There physics is classical and deterministic, as we generally know it. Even a single neuron in the brain is already too large to exhibit such wayward quantum behavior. But beyond that, even on the small scales governed by quantum theory, we are dealing with conduct that is fundamentally random. We have no way of determining what value the position (or velocity) of the particle will take. Measurement may at most cause it to choose some value, but it does so randomly (with probabilities described by the wave function). Therefore there is no mechanism of choice here in the libertarian sense, but at most randomness, and we have already seen that randomness is not freedom in the sense we are seeking.

At most one can say that quantum theory breaks the causal intuition, and thereby greatly blunts the force of the deterministic argument from the principle of causality. We can no longer say with certainty that every physical event must have a physical cause that uniquely determines it. In that sense quantum theory weakens the causal intuition, and thereby certainly improves the libertarian’s position. But there is no free gap in this theory in which one can house free will.

Emergence[12]

One more examination remains in order finally to rule out the possibility of freedom within a physicalist framework, and this one does not concern gaps located squarely within physics itself, but rather gaps that might accompany it or ride upon its back.

The thesis of emergence was born within the discussion of materialism. The fundamental question with which the materialist must grapple is the existence of mental phenomena (called qualia). How can one explain within a physicalist framework the fact that we have desires, experiences, emotions, thinking, memory, consciousness, moods, and the other “psychic,” or mental, phenomena that do not characterize physical bodies? A stone or an organic molecule has no such characteristics; they emerge only when an organic complex is formed in the shape of a cell, or a human being. This is in fact the basic motivation for advocating dualism, which sees the mental as a different substance, one that “rides upon” matter but differs from it. What can the materialist say in response? There are quite a few confused materialist arguments on this matter. Some simply deny the existence of the mental, and the confusion is great. The only reasonable materialist answer to this question is by way of emergence. The claim is that mental phenomena are merely a byproduct (epiphenomena) of material processes. For example (following the philosopher John Searle), the property of liquidity does not characterize an individual water molecule, yet it does characterize a collection of water molecules. In other words, liquidity emerges from the material whole and characterizes the whole, not the individual parts that compose it. So too, say the determinists, the mental phenomena—the qualia—emerge from our material whole (mainly the neural whole, the neuronal networks in the brain). They characterize the whole, even though it is composed only of material-physical components.

It is fairly easy to show that emergence does not really answer the question of the mental within a materialist framework, but for the sake of brevity I will focus only on what matters for our discussion: can emergence, even if it were true, answer the question of determinism as well, that is, make room for free will within a physicalist picture? Quite a few thinkers propose this. They argue that although our material whole is deterministic, the mental phenomena that emerge from it on the macroscopic plane display, among other things and unexpectedly, freedom as well. This freedom emerges from a picture that is wholly deterministic at the microphysical level. I will now show in a simple way why this is impossible.

Without entering into details about the picture offered by neuroscience today,[13] we can say that we have a microscopic physical system, which of course behaves according to the laws of physics. Every configuration of this system is a macroscopic state reflecting either a state of physical action (walking, speaking, and the like) or some mental state (sadness, desire for something, depression, joy, a particular mathematical calculation, an urge to eat, love, frustration, and so on). Such models are used by physicists in the field known as statistical mechanics, which underlies thermodynamics. In this picture, one sees the thermodynamic system as a collection of particles, each with ordinary physical properties (velocity, energy, position), while the whole has thermodynamic properties described in entirely different terms, such as pressure, temperature, entropy, compressibility, state of aggregation (solid, liquid, or gas), and the like. The macro characteristics emerge from the micro characteristics, but they create an obviously different language that is entirely irrelevant on the microscopic plane (at the level of the individual particle it is wholly irrelevant to speak of liquidity and solidity, or of entropy and pressure).

Returning to neuroscience and our mental states, let us take a very simplified model just to illustrate the emergentist perspective. Think of a chain of neurons, each of which can be in one of two electrical states, denoted “0” or “1.” Suppose now that we have a chain one hundred neurons long. Its state may be described by a vector of the following kind: (… 1,1,0,0,0,0,1,0,1,0,1,1,1,1,0,0,1,1). Every time the network reaches such a state, the laws of physics (including chemistry and biology, of course) determine where it will go next, that is, what its state will be at the next moment. Since this is a wholly physical system, its dynamics are rigid and deterministic, governed by the laws of physics. How many such microscopic states are there? In such a chain (far simpler than a real neuronal network in a human brain) there is an enormous number of states. We are speaking of 2¹⁰⁰ different states. To give some sense of scale, this is a number with more than thirty digits. A creature whose brain is so simple and primitive can be in a rather limited number of macro states (probably with no mental states at all), say about 1000 different states. It follows necessarily that there are a great many microstates corresponding to the same macrostate. The same is true in a human being, except that the neuronal network is much larger and more intricate, and the macroscopic states include mental states as well.

Neuroscience teaches us that there is some relation between the physics on the microscopic plane and the mental-psychic phenomena on the macroscopic plane. Just for the sake of example, suppose that a structure in which every prime-numbered position in that chain (the second, third, fifth, seventh, eleventh, and so forth) contains “1,” and all the others contain “0,” expresses anger on the macroscopic mental plane. But as we have seen, there are usually very many other microscopic states of the chain that also express anger. Continuing our hypothetical and simplified example, as long as the chain has “1” in 26 different locations (the number of prime numbers up to 100), the macroscopic state will be anger. As noted, there are a great many microscopic states corresponding to the same macroscopic mental state. In our example, the number of ways of distributing “1” among 26 different positions in the chain is about 7*10²³ states, that is, a 24-digit number.

It follows that if we see an angry person, we have no way of knowing the exact state of his neuronal network. It could be any one of billions upon billions of these microscopic states. This is the phenomenon known in statistical mechanics as “multiple realizability,” because physics can realize the same macroscopic state (mental, in our example) in many different ways. On the other hand, in the materialist picture every given microscopic state uniquely defines the mental state corresponding to it (for nothing besides the microscopic material states is involved in the creation of the mental state). The resulting picture is the following:

The relation between the macroscopic state (described by an uppercase letter) and its realizations, which are microscopic states (described by a lowercase letter). All the microscopic states in the same box have exactly the same macro-properties.

We can now understand why many materialists imagine that we have here a gap borne upon physics, one that ostensibly makes it possible to house free will within the physicalist picture. How does this work? Suppose Reuven provokes Shimon. The provocation creates in Shimon a certain mental macrostate A, say anger. But the microscopic state that is produced could be any one of the states a_i corresponding to anger. The libertarian question is whether he has a choice before him between hitting the one who angered him and restraining himself from doing so. At first glance, such a picture does contain this kind of freedom, for the same mental macrostate (anger) may correspond to two different microstates. Each of the microstates will lead, by the physical dynamics of the chain, to a different microstate, which may correspond to different macrostates, one restraint and the other response. If so, one current state allows two different paths of continuation, and all this within the deterministic picture. Apparently, a person in a given mental state can choose among several different options of response, and it seems that we have thereby housed free will and choice within materialist physicalism.

But this is a mistake. To understand why, let us assume that we are in the initial macrostate of anger (A). What is the microstate of our neural network? There are several microstates corresponding to anger, but at that given moment the person is of course in only one specific one among them, because the brain as a physical object is not simultaneously in several different physical states. It is clear that at that specific moment every neuron in the brain has one definite and unique state, and therefore the entire neural chain is described by one particular definite vector, say a₅. Thus at every moment the person’s state may be described on both planes by a pair of symbols, the left representing the mental state (macro) and the right the neural state (micro): in our example, at the first moment he is in the state (A, a₅). What happens at the next moment? In a physicalist picture, the dynamics of the brain are governed by the laws of physics and chemistry, that is, by the microstates; let us therefore assume that the laws of physics move the brain from microstate a₅ to microstate b₈, one of the states corresponding to the macroscopic mental state B. If so, the physical dynamics of the brain lead it from mental state A to mental state B. The brain is now in the state (B, b₈). Let us suppose that state B is striking the other, meaning that this person’s response to the anger aroused in him by the other was to hit the provoking party. Had the initial state been (A, a₉), whose macroscopic expression is likewise anger (A), then the laws of physics, beginning from microstate a₉ (rather than a₅ as before), could have led the brain to a different microstate, say c₂, whose macroscopic expression is C (restraint), that is, to the state (C, c₂). These are two descriptions of two dynamics that take us from a state of anger to two different responses, and therefore some have seen here a possibility of housing libertarianism within a physicalist worldview. But the crucial point they miss is that the decision between the two states is not entrusted to the reacting person’s choice. It is a deterministic result of the initial microstate and the laws of physics and chemistry. Therefore the freedom that exists among the microstates does not really make room for free will. It follows that emergence, even if it were a correct description of the relation between micro and macro (which is probably true), and even if it solved the problem of the mental (which is probably not true), simply does not do the job with respect to libertarianism. There is no way here to insert libertarian freedom into the physicalist picture.

The Picture of Dualist Libertarianism

How does one who holds a libertarian worldview see this? He too need not deny the empirical findings of neuroscience. For him too there is an emergent relation between the neural micro and the mental macro. He too understands that in a mental state of anger, the initial state might be (A, a₅) or (A, a₉); he disagrees only about the dynamics from that point onward. As noted, here we are no longer dealing with empirical findings but with assumptions. The physicalist assumes that there is no influence from the mental to the neural, only the reverse, and therefore the dynamics of the neural are a result of the laws of physics acting on the microstates. On his view, the macrostates are epiphenomena, passively borne along by the dynamics of the physical system (at the micro level), and as we have seen, this necessarily yields determinism. By contrast, the libertarian advocates interactionist dualism, that is, he holds that there is at times a possibility of intervention by the mental in the neural; the soul or the will, which are not physical entities, affect matter. As we have seen, on the libertarian’s view the will can create an electric field that moves electrons. This means that in ordinary conduct, the person does in fact react wholly mechanically, and then the physicalist description is correct—a neural microstate determines the person’s response. But in cases where the person chooses differently, there the next microstate may not be the one dictated by the laws of physics, that is, (B, b₈), but another state to which the mental has decided to move despite the laws of physics, for example (C, c₂), which on the mental plane means restraint.

Interim Summary

The meaning of the picture we have described in the last few subsections is that one cannot connect libertarianism with physicalism. In fact, we have returned to the identification between the two questions with which we opened the article: the question of materialism (or physicalism) and the question of determinism. We have seen here that libertarianism must go beyond physics, because within physics there are no gaps that allow the kind of freedom it describes. The rifle does not give the hunter the bird, but at most its carcass. As I claimed above, in order to be a libertarian you must be an interactionist dualist.

Neuroscience: Libet’s Experiments[14]

We conclude this section with the only discussion in our issue so far that claims to be scientific and empirical, rather than a priori conceptual-philosophical. At the beginning of the 1980s, the American neurologist Benjamin Libet began a series of experiments that stirred up a stormy discussion around the question of free will. These experiments continue to this day, with technologies and scientific apparatus that examine the matter in an ever more precise, sophisticated, and focused way. This is a lively and varied field, and this is not the place to enter into it.[15] Again we must make do with a schematic description that will explain why these experiments too do not do the job—in other words, why even after Libet and his successors, the question of free will remains on the philosophical plane and has not passed over to the scientific one.

Libet seated his subjects in front of a button and asked them to choose a moment and press it when they decided to do so. At the same time they were supposed to watch a very precise clock and report the moment at which they decided to press. The experimenter checked the moment of pressing, and simultaneously also tracked by EEG the brain processes in which an early electrical signal appears before the button is pressed, what is called the Readiness Potential (RP). To his astonishment, Libet discovered that the RP signal appears not only before the actual press of the button, but significantly even before the moment at which the decision to press is made. The straightforward meaning of this result is that conscious decision-making is nothing but a fiction, for the brain dictates the decision before it has been consciously made. It is imposed upon us, even though we have the illusion that we freely choose when and whether to press. Many see these findings as empirical proof of determinism.

It is no wonder that these results caused an uproar among brain researchers and philosophers dealing with free will, and many of them were persuaded to “convert” to determinism. Libet himself was a clear libertarian, and he was both the first to be surprised by the results and the first to offer them a libertarian interpretation. I will not present here all the criticisms (there are dozens of different types, some of them very technical) that arose following the experiments, or the debates surrounding them. Here I will focus only on the principled critique of the deterministic interpretation itself, and I will do so on four levels of argument, one upon another. In the end, the reader will be able to see that there is not even a shred of evidence here in favor of determinism. Moreover, one can learn from these arguments that even if such experiments continue and become ever more refined, it is doubtful whether this empirical route will yield unequivocal conclusions even in the future.

• The first argument is the veto theory proposed by Libet himself. Libet argued that although an RP signal always appeared before the press, there were cases in which an RP arose and no press of the button was performed. This means that the signal does not dictate the press, but only enables it (it is a necessary but not sufficient condition). In our terms (see above, section A), the signal expresses pressure exerted by the topographical layout (the internal and external circumstances) on the subject to press the button, but the subject can decide whether to comply with it or to resist it—in other words, to impose a veto on the impulse of the topographical layout. The picture we described in section A neutralizes this purported proof of determinism from the outset.
• Now let us suppose, for the sake of the discussion, that a different Libet experiment were conducted in which no exceptions were found. In every case where an RP appeared, a press on the button in fact followed immediately. Do we now have evidence for determinism? Still no. The reason is that pressing a button is an action devoid of value significance. It involves no real deliberation or moral or other decision. In the terms we saw above, this is a trivial action equivalent to a lottery (the elections in Switzerland). If a person is required to perform a trivial action like pressing a button, it is perfectly reasonable that the action will occur when circumstances push him to perform it. Even if in principle he can resist that impulse, he has no reason to do so. By what criterion would he choose the moment of pressing? There is no substantive criterion here (values, or utility), and in such cases it is obvious that the person does not resist impulses. Libertarianism speaks of decision in situations where there are reasons to resist the topographical impulse, and only there can we examine whether we have free will. In English this is called choosing rather than picking, like an arbitrary press of a button.
• Let us now go one step further, and suppose that a Libet experiment were conducted on an action of choosing. Let us also suppose that every time an RP signal appeared, the action it predicted was indeed performed, with no exception. Would such a result finally constitute empirical proof of determinism? Unfortunately, here too the answer is no.

To understand why, let us look at the decision of a person who stands at a value-crossroads, that is, a crossroads of choice between conflicting values. For example, the French philosopher Jean-Paul Sartre recounts that one of his students approached him during the Second World War in occupied Paris. He was torn between remaining to help his elderly, sick mother, or abandoning her and fleeing in order to join the Free French forces and fight the Nazis. Such a dilemma pits against one another the value of helping an elderly mother and the value of fighting evil. How does one decide such a dilemma? Quite simply, one has to build a scale of values, and examine which of the two values stands higher on one’s scale. The question is when that student built his value scale. Is it necessarily built at the moment of decision? Or perhaps he already built his scale sometime in the past, and now he merely performs a wholly mechanical calculation on the basis of that scale. In other words, it is unclear whether Sartre helped him build the scale, or merely helped him perform a calculation on the basis of an existing scale.

If the value scale was built at some point in the past, then there is no obstacle to a Libet experiment accompanying such a decision producing “deterministic” results, that is, the RP signal, which represents a mechanical calculation and not a conscious decision, would determine the action performed. This does not mean that the student does not decide freely; rather, it means that his decision was already made in the past, before we attached him to the EEG, and then it was made consciously and through judgment.

• At the end of the discussion in my book I raise yet another possibility. In the previous stage we dealt with a value-versus-value choice, in which there is a conflict between two values. But there are situations in which a value-versus-impulse choice is required. For example, a person who is torn over whether to rob or murder (for money). One might argue that in such situations no value scale is involved, for there is no moral dilemma here. Even the person who is deliberating knows what is morally right to do, and his value scale is already fully built. His struggle is current: whether to overcome the impulse for the sake of the value or not. There may be room for the view that decisions of this sort are indeed made at the moment of decision and not beforehand (since the scale is already built), and if so, Libet experiments conducted on such situations might perhaps decide in favor of determinism (if indeed there are no exceptions in which an RP appears and yet the action it predicts is not carried out).

It is clear that this proposal contains several nontrivial assumptions. It may be that the strength to overcome impulse for the sake of value is itself “accumulated” at earlier times. Moreover, even here there is a kind of choice and value scale, albeit of a somewhat different sort, and there is no obstacle to claiming that the scale relevant to such decisions was also built sometime in the past. Beyond that, it is almost impossible in practice to conduct such an experiment.

The conclusion is that it is probably impossible to settle the question of determinism empirically, at least not by the route of Libet experiments, however refined they may become. In any event, at least as of our own time there is no doubt that the question has not yet been settled empirically. For now it still stands as an open philosophical question.

D. Diagnostic Tools

Introduction

Thus far we have seen that there is no possibility of finding gaps within physicalism that would make room for the libertarian conception inside the physicalist picture. We then saw that, at least in the foreseeable future, there is no possibility of deciding this question scientifically. Even in our own day it is still a philosophical question. In section B we proposed a way of deciding it on the philosophical plane (by means of Lex Specialis). What can someone do who still wonders and is perplexed by this question, and to whom the scales appear balanced? How, if at all, can he form a position?

In such a situation there is no escape from a personal diagnostic process. Each of us must examine his basic intuitions, and as long as there is no scientific finding or philosophical argument decisively one way or the other, he has no choice but to rely on those intuitions. The burden of proof rests upon whoever wishes to uproot them from me. We have seen that anyone who chooses determinism is supposed to give up judgment and the ascription of moral responsibility (except in the emptied senses described above in section B). How many of us are prepared to give up our moral intuitions for the sake of determinism without any real evidence? In the next and final section I will present the picture that emerges from Torah sources; believing Jews will presumably take that into account as well when shaping their position.

Such self-diagnosis is built from a series of thought experiments that a person carries out by himself in order to clarify his own feelings regarding the question of determinism. I will briefly describe several of them here.[16]

Elitzur’s Series of Experiments

Avshalom Elitzur, in his book Time and Consciousness,[17] chapter 11, illustrates the distinction between the material dimension and the qualia by means of the following thought experiment. Suppose that you, the reader, receive the following proposal from me: let me kill you in a painless manner, with the promise that I will construct a person who is completely identical to you biologically—that is, every atom, molecule, and neuron that exists in you, and in exactly the same physical state, will be placed in his body in exactly the way they are placed in yours. Afterwards, just to make things more piquant, I will give that duplicated twin of yours a million dollars. Would you agree?

Now the person is offered an even more tempting experiment: we will take from you every molecule, one by one, and reassemble those very same molecules, and thereby recreate you. Would you be willing to agree to that? See there in the chapter for further proposed extensions of these experiments, and for the discussion of them.

The consistent materialist should agree to this proposal, because he does not really die at all. His duplicate has the same thoughts, the same personality, the same memories and experiences, beliefs and values, desires and character, and therefore it is really he himself. As a materialist, he thinks that the human being is nothing but the collection of cells and molecules, and therefore also the collection of all the memories, beliefs, emotions, and experiences that emerge from them. From the materialist’s perspective, the meaning of this proposal is simply that you are being offered a million dollars without doing anything. So what is the problem?

I permit myself to guess that none of my faithful readers (or the less faithful ones) would agree to have this experiment performed on himself. Why not? Elitzur argues that the decision not to agree tacitly presupposes a dualistic conception, that is, a conception that sees the human being as containing a dimension beyond his biology. This duplicate is indeed completely identical to me on the biological plane, and yet it is clear to me that he is not myself. His soul is different from mine.

Such an experiment, in itself, proves nothing. A person can say that he would refuse or agree in accordance with his philosophical position. There is no argument here that forces him to abandon it. The purpose of these experiments is not to persuade one of us to exchange his beliefs, but only to place tools in his hand for self-diagnosis. If he refuses such proposals, he can learn about himself that he is probably not a materialist, because in his opinion the phenomena of qualia cannot be identified with the neural processes that generate them. If he agrees to the proposal, he is probably a materialist.

Buridan’s Donkey

Many people do not notice that just as determinism cannot explain free choice, it also cannot explain a situation in which we make some other non-causal decision. If the human being is nothing but a material whole, then all his actions are causal-deterministic in character, governed by the laws of physics. Such a picture excludes not only free choice but also randomness (at least on the macroscopic scale, where quantum theory is not relevant).

Jean Buridan, rector of the University of Paris during the fourteenth century, raised the following challenge to the concept of rationality. Think of a donkey standing at equal distance from two identical feeding troughs. Its entire environment is perfectly symmetrical in both directions, and the assumption is that its body too is endowed with a similar symmetry (we are speaking of a point-like donkey). Our amiable donkey is peacefully grazing, facing exactly straight ahead. When the grass around him is finished, he rests, and when he feels hunger again he wonders which of the two troughs before him he ought to approach. Because of the symmetry he sees no possibility of preferring one of them, and so he stands in place, lost in thought, unable to reach a decision. In the tragic ending of the story, our donkey dies of hunger. Buridan wonders whether this is rational behavior. Seemingly it is, because the donkey is unwilling to act without a good reason (and there is no good reason to prefer trough A over trough B). On the other hand, it does not seem terribly rational to die of hunger, even if one does so in order to be rational.

But now let us ask ourselves what a human being (a point-like one) would do in such a situation. Seemingly he too is a rational creature, no less than the donkey, and therefore he too should die of hunger. On the other hand, intuitively it is obvious to all of us that the human being’s fate would be different. A human being, as a rational creature, would conduct some sort of lottery—in other words, decide arbitrarily on one of the troughs, so as not to die of hunger. But that answer is not so simple. In the mechanistic-physicalist picture of the human being, every action is performed because of a prior physical cause. But the physical state in which this person stands is perfectly symmetrical, and therefore there is no possibility that as a result of such a state the person will turn to one side (that is, conduct a lottery). Here we are dealing not with a question of rationality but of physical causality. It is important to understand that this claim is a mathematical theorem about the laws of physics, not a conjecture. A mechanical creature that acts according to the equations of the laws of physics cannot act in a way that lacks the symmetry of the equation (or of the problem). The surprising conclusion that follows is that determinism excludes not only an act of free choice, but also a random act, such as a lottery. Determinism states unequivocally that even a human being in a Buridan situation would die of hunger.

Now each of us must ask himself whether his intuition is prepared to accept that he himself, when standing in such a situation, would be unable to turn toward one of the troughs and would simply die there of hunger. It sounds absurd, but this is a mathematically necessary conclusion from the physicalist worldview. It seems to me that intuition rebels against this conclusion. A human being is a rational creature, but not necessarily in Buridan’s causal sense; rather, in the sense of reasonable judgment. It is not plausible to assume that a human being in such a situation would indeed die of hunger. But what is the meaning of his decision to eat from one of the troughs? Necessarily it is an action not determined by the physical circumstances that preceded it (for those circumstances were symmetrical and the decision is asymmetrical). It follows that there was something beyond physics that participated in determining the continuation of the process. This compels a departure from physicalist determinism, and the adoption of interactionist dualism. And again, there is no proof here. We are dealing with a thought experiment that each of us should perform on himself in order to discover what he himself thinks.

E. A Torah Inquiry[18]

Introduction

To conclude, I will try to sketch in schematic form the outlines of a Torah position on the issue of free will. For the reader who wonders why there is any need for this at all, I will say that in recent years one hears challenges that try to deny the libertarian conception of the Torah and of our tradition. It seems to me that the roots of these challenges lie in new scientific findings, but they also rely on several sources (esoteric and far from unambiguous) in Jewish thought over the generations.

The starting point of the discussion is of course what the Torah itself says (Deuteronomy 30:19):

I call heaven and earth to witness against you today: I have set before you life and death, blessing and curse; therefore choose life, that you and your offspring may live.

Ever since, it has been accepted that a foundational assumption of serving God and of the entire Torah outlook is that even within our world, which conducts itself according to rigid natural laws, and despite the fact that the human being is a creature made of matter, he has free choice, in thought and in deed.

As one example among many, I will cite Maimonides, who writes in chapter 5 of the Laws of Repentance:

Permission is granted to every person: if he wishes to incline himself to the good path and be righteous, the choice is in his hand; and if he wishes to incline himself to the evil path and be wicked, the choice is in his hand. The Holy One, blessed be He, does not decree upon a person from the beginning of his creation that he be righteous or wicked… Know that this matter is a great principle and the pillar of the Torah and the commandment, as it is said, ‘See, I have set before you today life,’ and it is written, ‘See, I set before you today’—meaning that the choice is in your hand, and whatever a person wishes to do of human actions, whether good or bad, he does… But know without doubt that a person’s deeds are in the person’s own hand; the Holy One, blessed be He, neither draws him nor decrees upon him to act thus… And because of this it is said in prophecy that a person is judged for his deeds—according to his deeds, whether good or bad.

It cannot be denied that there are several Torah sources that express deterministic positions. Among them is Rabbi Hasdai Crescas in his book Or Hashem, who devotes two sections to the issue of choice. Avi’ezer Ravitzky argues that the first section expresses a more moderate position, but in the second section (which, according to Ravitzky, is later) a fully deterministic position is already expressed. One may add several Hasidic sources, mainly from the school of the Mei HaShiloach and his disciple Rabbi Zadok HaKohen of Lublin, which on the face of it also advocate determinism. But their words are also full of completely opposite references,[19] and as is well known, from a contradiction one can derive any conclusion one wishes. I must add a personal remark. Even if there were no contradictions in their words, the fact that one can find some claim in a book printed in Rashi script and bound in brown covers with gold letters is not enough to prove such a far-fetched claim and grant it Torah legitimacy. At most, it is an elementary error and a failure of judgment. One can of course say that the deterministic position is a modern scientific finding, and therefore we must subordinate the sources to the facts (see the words of Maimonides that I cited at the beginning of the chapter on evolution), but to the best of my judgment, bringing such sources as such neither adds nor subtracts anything from a discussion like ours.

A Discussion of Determinism in Light of Halakhic Concepts

At first glance, in the deterministic picture there is no place to punish criminals, for their actions are not the result of choice. On that basis, the very fact that Jewish law punishes criminals would seem to be proof that one cannot reconcile fidelity to Jewish law with determinism. But as I have already noted, the determinist explains that punishment is a matter of “correction” rather than moral desert. That is, the very fact that a punishment is imposed for a transgression itself influences us (deterministically) not to sin (that is, not to do “bad” actions). Therefore the mere fact that Jewish law imposes punishments on criminals proves nothing.

But Jewish law not only punishes criminals; it also exempts from punishment someone who committed a transgression under compulsion. The condition for punishment is that the transgression was committed out of free will. Jewish law recognizes internal compulsion as well for the sake of exemption from punishment (see Babylonian Talmud, Ketubot 51b, it began under compulsion and ended willingly). In a deterministic picture there is no room for such a distinction.

In addition, Jewish law contains a whole system of mental concepts that are involved in the evaluation of criminality. Recall that in the materialist-determinist picture the mental emerges from the material and is necessarily derived from it. So why make anything conditional on mental states at all? For example, a person who sinned inadvertently is exempt from punishment. If the goal is deterrence, why exempt the inadvertent offender? There is no room here for considerations of fairness in punishment, since punishment is not moral retribution but a mechanical corrective device. Another example is the exemption of one who lacked intention. This does not necessarily mean someone who did not intend to violate the commandment (that is the inadvertent offender, or the absent-minded actor), but someone who did not want the transgression. For example, a person drags a bench across a field on the Sabbath, and in the process creates a furrow in the ground (which is forbidden because of the labor of plowing). He knew that he was committing the transgression (or that it was likely to occur), but that was not the purpose of his act. Since he did not intend plowing but dragging the bench, he is exempt. Another exemption is when the result is unwelcome to him. When a person performs an action without intention, such as dragging the bench, in a case where the creation of the furrow is certain in advance—he is in fact liable. But if he does not want the furrow that is created, then according to some halakhic decisors he is still exempt. Why does it matter whether he is pleased with that result? What matters, after all, is that the transgression was committed. Another concept from this family is that of the absent-minded actor, that is, a person who commits a transgression without noticing. For example, a person walking on the Sabbath past a tree branch, and while walking and brushing against it with his body, leaves are torn from the branch. In such a case too the person is exempt from punishment.

All these concepts raise difficult questions in the deterministic picture, because it is not clear why intention should matter to Jewish law if its concern is merely with avoiding acts. Intention is only a mental epiphenomenon, and it should not be part of the considerations governing punishment. In a deterministic-materialist conception, punishment comes to correct only the neural dimension, and it is not clear why Jewish law cares whether passive mental processes of one kind or another accompany it. It is quite evident that Jewish law assumes free will and interactionist dualism.

Thus far we have discussed transgressions, but the question of intention arises also with respect to positive commandments. According to all views, performing the commandment is not enough; intention is also required. True, some hold that this is not indispensable, meaning that performance without intention also has value, but according to all views intention is required ab initio. Again, we find dependence on mental states and choice.

Another conceptual world in which one sees the significance of the mental dimension and of freedom of choice is the world of transactions. Jewish law requires, as a condition for a person’s acquisition of an object to take effect, that there be firm resolve on the part of both parties; that is, both parties must have a clear decision that they want to buy and to transfer the object, and without this no transaction takes effect. In the deterministic conceptual world, all this discussion is completely meaningless. What is the difference between someone who performs the act of purchase without a clear decision to transfer, and someone who performs such an act with a clear decision? The decision is a deterministic outgrowth of a brain state, and the mental dimension is only an epiphenomenon. So why is any of this important at all? The same is true, of course, with respect to every other halakhic context in which a person’s knowledge or decision is required in addition to his act. This is so in conversion, in contracts, in marriage and divorce, and in many other areas—all of these lose their meaning within a deterministic worldview.

Even halakhic concepts that do not belong to the legal sphere are bound up with intention and decision. Concepts such as repentance and regret, which are conditions for the atonement of sins, also lose their meaning in the deterministic view, because these actions too are nothing but a result of neural physics. What does it mean to regret? To feel an illusion of regret?! This is not a decision I made to regret, but a state into which I have been drawn. Even if we accept the existence of a psychic state of regret as a kind of epiphenomenon, why is what occurs there important at all? Why is it a condition for atonement and for exemption from punishment?

And what about a vow that a person vows? Jewish law allows him to go to a sage who will release him from the vow. That release requires regret and an opening; that is, one investigates whether the person made the vow in error, and then the vow is permitted. But if the vow is always made under internal compulsion, what is the meaning of releasing the vow? What is regret? It is nothing but an epiphenomenal illusion accompanying this or that neural activity. In light of this, it is really unclear what the action of the sage who releases the vow means at all, for he too is compelled to do what he does. This whole halakhic picture looks utterly bizarre when viewed through the deterministic prism.

And above all, what is the meaning of returning in repentance in the deterministic picture? To feel an illusory sense of repentance that accompanies this or that neural activity. What importance does the regret I now feel for my past deeds have? After all, I was compelled to do them, and I am compelled now to regret them. So why does the Torah command this? Even if we understand that the Torah’s commandments are meant to influence me so that I will act toward practical goals, what is the command (or recommendation) to repent meant to achieve? Why is it so important to the Torah that I regret?

If we are already discussing punishment, we may also ask what the meaning is of the ultimate recompense—the world to come—and as a corollary, what the meaning is of belief in the persistence of the soul. Why impose punishment on a person who sinned, if he was compelled to do so? As I already noted, in ordinary legal systems one can find logic in this even within the deterministic picture, because a legal system tries to correct human behavior, and punishment is only an additional constraint intended to cause us to behave differently. But punishments in the world to come (as well as punishments such as death at the hand of Heaven, about which nobody knows when or whether they are carried out) do not correct anything, because they are given after our death and nobody knows about them. So what is their function? At most, one can justify speaking about punishments and about the world to come, since such statements may alter our conduct here in this world. But the world to come itself, and certainly punishment in the world to come, do not seem reasonable, and there is no reason that God should actually carry out such punishment.

If so, logic would suggest speaking about the world to come but not actually implementing that punishment—that is, to lie. But then even the talk no longer achieves its purpose, because I know through rational reflection that it will not be carried out. Moreover, all this arises only if we can understand at all who the entity is that is punished there; for our soul does not exist from the determinist’s point of view (and as we saw, he is usually also a materialist), so who is punished after we die? And what is a world to come in a deterministic system at all? Is it a material world?

Alternatively, one can of course allegorize everything, and tell us an absurd story according to which all our actions are shadows of the real world. In truth everything is mechanical, and the illusions are epiphenomena, and the story told to us is a lie addressed to that epiphenomenal illusion, that is, addressed to feelings, insights, and experiences, even though these determine nothing whatsoever about actual behavior. This is only the natural continuation of the same approach that sees the whole world and all human feelings and insights as a collection of illusions. But this is not very convincing. It seems that in order to persuade me to give up all my insights (except determinism itself, of course) and to regard them all as mere illusions, much stronger and more convincing arguments are required. It is somewhat strange that in the name of science and rationality one demands that I give up the results of direct observation and regard them all as illusions.

Back to the Torah

If with the Sages one may speak of a mistaken conception that is the product of their time and place, what shall we say about the Torah’s own words? Above we already saw the command to choose life, and now we shall quote the context (Deuteronomy 30):

For if you obey the Lord your God, keeping His commandments and His statutes written in this book of the Torah, if you return to the Lord your God with all your heart and with all your soul. For this commandment that I command you today is not too wondrous for you, nor is it far away. It is not in heaven, that you should say, ‘Who will go up to heaven for us and bring it to us, and let us hear it so that we may do it?’ Nor is it beyond the sea, that you should say, ‘Who will cross the sea for us and bring it to us, and let us hear it so that we may do it?’ Rather, the matter is very near to you, in your mouth and in your heart, to do it. See, I have set before you today life and good, death and evil. For I command you today to love the Lord your God, to walk in His ways, and to keep His commandments, His statutes, and His ordinances, that you may live and multiply, and the Lord your God will bless you in the land to which you are coming to possess it. But if your heart turns away and you do not listen, and you are drawn away and bow down to other gods and serve them, I declare to you today that you shall surely perish; you shall not prolong your days on the land that you are crossing the Jordan to enter and possess. I call heaven and earth to witness against you today: I have set before you life and death, blessing and curse; therefore choose life, that you and your offspring may live, by loving the Lord your God, obeying His voice, and cleaving to Him; for He is your life and the length of your days, to dwell upon the land that the Lord swore to your fathers, to Abraham, to Isaac, and to Jacob, to give them.

How should a person who believes in Torah from Heaven relate to verses in the Torah that speak in such terms? The assumption of a believing Jew is that the Torah was written by God, so does God too not know the truth? Or is His purpose to confuse us, or perhaps to manipulate us in some way? As noted, even if we are prepared to accept God as such a manipulator, it is still unclear why all this is needed at all. If God created us, and created the whole world and its laws, for what purpose did He do all this? Why does He command us to do good and avoid evil? After all, He programmed us for that. And if His programming is not good enough and He needs help in the form of such commandments, then let Him improve it and do away with the unnecessary commandments.

The determinist continues here too to speak about illusions and shadows, about mistaken conceptions and hidden manipulations. For him, just as our insights about ourselves and about life are not correct, so too our insights about the Torah and the commandments are not correct. Moreover, the Torah itself helps create this great illusion, but with the kind assistance of neuroscience (which alone—and determinism, of course—are not an illusion but pure truth), we now finally know all this. The picture the determinist presents is a kind of caricature of the world, of the human being, of religion, and in fact of God as well. On his view, God constructed for Himself a virtual game in which programmed creatures play before Him, living under the illusion that they decide their actions, that momentous things depend on what they choose to do, when in fact all of this is nothing but a rigged game. The outcomes of it all are fixed from the very first moment. Is that a reasonable interpretation?

Can God Himself Be Found Within a Deterministic Picture?

We have seen in the course of the book that causal determinism is based on physicalism. At its base lies the assumption that the whole world is nothing but matter, and since that is so, the laws of physics are its sole rulers. The physicalist generally also claims that there is no evidence for the existence of anything beyond physics, and therefore there is no reason to assume its existence. As we saw (in chapters 9 and 10), within the physicalist framework there are no gaps, and therefore it dictates determinism.

But if we are indeed looking at the world from a materialist angle, then beyond the existence of souls, it is not at all clear how one can believe in an abstract spiritual being like God. What is the evidence for His existence? Are those pieces of evidence not themselves illusions? Why are they essentially different from the evidence for the existence of a soul or psyche? Even direct revelation, such as that at Mount Sinai (in the eyes of one who believes in it), ought to be interpreted by physicalism in terms of the laws of physics, because voices, lightning, and fire are physical phenomena, and such phenomena are produced by purely physical causes. That, after all, is the essence of physicalism.

Furthermore, it is not clear how, in the physicalist’s opinion, the interaction between God and us is created. After all, one of the basic assumptions that lead physicalists to their view is the assumption that there can be no connection between the spiritual and the material (the psychophysical problem). On their view, the spiritual cannot operate any physical thing, and therefore we have no way of knowing that it exists. But divine revelation is perceived by our senses, that is, an encounter with a spiritual being (non-physical) creates an interaction with the human being (who is a material being) and moves electrons in his brain, and that is impossible in the deterministic worldview. It seems that there is now no escape but to add God Himself, with all due respect, to the determinist’s list of useful fictions.

We will conclude this section with a brief discussion of two further important remarks concerning the Torah picture of libertarianism.

A. Rabbi Ilai’s Rule

In the second chapter of my book I discuss a puzzling saying of Rabbi Ilai (Babylonian Talmud, Moed Katan 17a and parallels):

For it was taught: Rabbi Ilai says, if a person sees that his evil inclination is overpowering him, let him go to a place where no one knows him, dress in black, wrap himself in black, and do what his heart desires—but let him not profane Heaven’s name in public.

Some commentators suggested seeing here a deterministic conception according to which a person cannot overcome his inclinations.[20] This interpretation is refuted by itself, for if we were really dealing with determinism, then the person also could not go to a distant place and wear black. Moreover, every sin would be the result of unavoidable compulsion. So why does Rabbi Ilai speak only of specific situations (sins and exceptional cases)?

It is quite clear that Rabbi Ilai is dealing only with special situations in which a person indeed estimates that he will not be able to overcome his impulse. His basic assumption is distinctly libertarian, except that, as we have seen, even the libertarian accepts the possibility that a person acts under an irresistible impulse. The determinist understands every human action in that way. For him, that is the normal situation. By contrast, the libertarian claims that these are exceptional cases.

B. The Theological Argument: Divine Knowledge and Freedom of Choice

Religious conceptions tend to oppose determinism strongly. True, there are deterministic approaches in Christianity (Calvinism, and Protestantism generally, are particular kinds of determinism), and there are also such esoteric approaches in Judaism, but one may say that the mainstream of Jewish-Torah thought is libertarian. Against this background, it is interesting to note that one of the oldest considerations in favor of determinism is precisely a consideration of religious, or theological, origin, sometimes called the problem of foreknowledge and free choice.

A common assumption in monotheistic theologies is that God is omnipotent, and therefore naturally also omniscient. And if He knows everything, then He should of course know the future as well. But if He already knows the future now, then it follows that the future is fixed. An event contrary to divine knowledge cannot occur, because if such an event did occur, it would prove that God is not omniscient. This is an argument that grounds determinism not on the principle of causality, but on something akin to the logical determinism we encountered in section B. But here it cannot be rejected in the same manner. There we saw that logic is indeed indifferent to the time axis, but knowledge is not a logical phenomenon; it is a phenomenon containing information. Information cannot exist before it is created.

This is a broad topic, and here I will confine myself to two comments. First, the very fact that the commentators struggle over this point shows that their basic assumption is libertarian. I would add that the commentators’ suggestions for reconciling the contradiction suffer from many and varied flaws. To the best of my understanding, divine foreknowledge really cannot be reconciled with our freedom of will. In section B above I described Newcomb’s paradox, and noted that it also rules out the possibility that God Himself knows in advance what will occur. But on further reflection we will see that there is no need to reconcile this contradiction at all, because the assumption that God knows everything in advance applies only to natural events that occur by deterministic mechanism. By contrast, events that involve the free choice of human beings are indeed not foreknown. God Himself does not know them in advance. He too cannot know information that does not exist. Moreover, there is no theological or other reason to assume that He does know it. See on this my book Two Wagons and a Hot-Air Balloon (in note 28 and its vicinity).

[1] See, for example, Richard Taylor’s book, Metaphysics, translated by Yael Cohen, Open University Press.

[2] The most comprehensive and precise description of compatibilism known to me is found in Eliezer Malkiel’s exemplary book, Will, Freedom, and Necessity, Interpretation and Culture series, Bar-Ilan University Press, 2013. In my review article, ‘On Ethics, Semantics, and Psychology,’ published on the Mida website on 21.2.2014, I point out that the book contains no philosophical mistakes (which are extremely common in this literature), and yet in my opinion there is a grave error in the conceptual framework around it, and that is what trips it up, as it trips up the compatibilist project in general.

[3] In several places in my books (my book Two Wagons and a Hot-Air Balloon is devoted entirely to discussion of what I call “analyticity” and its implications), I have pointed out that this is the character of analytic “solutions” to philosophical problems, in the style of the “theory of types” proposed by Russell and Whitehead in the introduction to their book Principia Mathematica, in order to prevent the formation of instructional paradoxes or paradoxes of self-reference. They construct a language in which there is no possibility of expressing the paradox, and from their point of view it is thereby solved. Compatibilism belongs to that philosophical tradition, and like many analytic treatments it is an empty semantic trick.

[4] This is only a parable, of course. In a real case like this, the libertarian can agree with the determinist that the weighing of the effort required to climb the mountain or hill also enters into the deterministic calculation that will determine whether the person climbs or not. For our purposes, all the external influences are already incorporated into the topography. From this point onward, we are speaking about a free decision.

[5] Some would say that there is no difference among them—at the end of the day, the same thing always happens. For the purposes of the present discussion, we will ignore that aspect.

[6] See:

Van Inwagen, P. ‘Free Will Remains a Mystery’, in R. Kane, The Oxford Handbook of Free Will, Oxford: Oxford University Press, 2002, pp. 158-177.

[7] This simple version is taken from Raz Shpeizer’s article, ‘Is There Such a Thing as Free Will?’ on the e-mago website (dated 29.5.2007); see there at the end of the article and in footnote 6.

[8] See a survey in the Stanford online encyclopedia (SEP), under the entry “fatalism,” by Hugh Rice, from late 2010. See also Richard Taylor’s article:

Richard Taylor (January 1962). ‘Fatalism’. The Philosophical Review, Duke University Press, 71 (1): 56-66.

[9] For a fuller treatment, see chapter nine of my book.

[10] See Chaos, p. 253.

[11] For a fuller treatment, see the discussion in chapter ten of my book.

[12] For a fuller treatment, see chapter thirteen of my book.

[13] See the discussion in chapter twelve of my book.

[14] For a fuller treatment, see chapters fourteen and fifteen of my book.

[15] One may consult Ariel Ronen’s master’s thesis, which deals extensively with Libet’s experiments and their significance. The thesis, written under the supervision of Prof. Benny Shanon, is titled The Freedom to Want Even When the Brain Wants: A Critical Analysis of Libet’s Findings (Libet), Department of Psychology, Hebrew University of Jerusalem, April 2008.

[16] See a fuller treatment in chapter sixteen of my book.

[17] University Broadcast Press and the Ministry of Defense, 1994.

[18] See a fuller treatment in the appendix to my book.

[19] See the appendix to my book on this matter.

[20] See, for example, ‘Explanations of Aggadot According to the Plain Sense,’ printed at the end of the book Kovetz He'arot by Rabbi Elchanan Wasserman.