In philosophical discussion of the relation of mind and body, the most heated debate throughout the history of philosophy, which continues today as strong as ever, has been that of free will vs. determinism.
The idea of free will holds that man generally has more than one action possible to him, with his actual action the result, directly or indirectly, of his own choices. Man is in some cases free to choose between several alternatives, all of which are possible, with the choice between them directly under his control and his actions, directly or indirectly, the result of these choices. As against free will stands the idea of determinism, holding that at any time there is exactly one possible action that a person can take consistently with the operation of the laws of nature, and so there is no choice among alternatives. (Historically, some philosophers and theologians have advocated a third position, opposed both to determinism and to free will: divine fatalism, the idea that there often is more than one possible action that man could take, but the choice is not made by man himself but by God or the gods. In modern, scientifically informed discussion, divine fatalism is no longer taken seriously, and I will not discuss it further in this paper.)
The case for free will is clear; it is a self-evident, directly perceived fact. Every reader of this paper can directly perceive that the amount of mental effort he spends on considering and trying to understand it—and then whether he agrees with me or not—is under his own control. The same is true every time any one of us is engaged in any thought process of any difficulty, or makes a decision of any significance in his actions.
In contrast, in reading the writings of determinists, it is often unclear just what their case is for accepting determinism. It is common for determinists to tout determinism as scientifically proven, declare that we must accept determinism in order to be scientific, or compare free will to outmoded scientific theories such as Ptolemian astronomy or phlogiston, without ever stating just what evidence or arguments they believe they have in support of determinism. This often makes the debate highly frustrating, and makes it seem irresolveable.
My purpose in this paper is to try to shed light on the debate by attempting to make a complete list of all the rational arguments that have been raised in support of determinism, and then pointing out the fallacies in each of these arguments. In all my readings of writings by determinists, or personal discussions with determinists, I have come across a total of five arguments that have been raised to support determinism. I will discuss these five in what I consider descending order of intelligence; starting with those arguments that have some plausibility and have been raised by intelligent thinkers; and ending with the arguments that I find the most stupid and obviously fallacious, but which I am still going to discuss for the sake of completeness. I submit, however, that even the former, more plausible arguments can be conclusively answered.
Note that the point of this paper is not simply to address a list of some arguments that have been raised for determinism; it is to address all of the arguments raised in support of determinism. I of course welcome comments on my paper pointing out problems with my answers to the five arguments; but I would welcome even more strongly comments pointing out any sixth argument that's been raised in support of determinism and that I have missed; if any one can point out any such sixth argument, that would be an incompleteness in this paper which needs to be corrected.
Determinism is often held as part of a more general model of the universe, which involves two basic principles:
Mechanism is the idea that the universe is built out of physical particles whose movements are determined by their previous movements and their physical impact on each other. (Note that the term "mechanism" is sometimes used loosely as a synonym for determinism, but in this paper I use it in its more precise sense to refer to this model of the basic constituents of the universe.)
Reductionism is the idea that human beings are complex systems of these physical particles, and that causal laws governing those particles completely determine the actions of the system.
The model of the universe based on these two principles was originated by the Greek atomists, and is associated in modern times with the physicist Pierre-Simon Laplace. It is held by contemporary determinists such as Daniel Dennet, Paul and Patricia Churchland, and many others.
In understanding arguments for determinism, it is important to be precise in identifying just what the argument seeks to prove. Two of the arguments I address in this paper - arguments 3 and 4 - are arguments that cclaim to support, not just determinism, but the entire mechanist/reductionist model of the universe. The other three arguments are more general arguments in support of determinism.
Another idea that is today often associated with determinism is that of compatibilism; the idea that people should continue to use the concepts of free will and of choice, but redefine these concepts so that they no longer involve choice among several possible alternatives, and can thus be made compatible with determinism. Compatibilists have created a lot of confusion, using the familiar terms "choice" and "free will" with new meanings, thus making it harder to clearly explain positions, arguments and evidence on the issue. In order to avoid this confusion, in this paper I use the term "free will" in its plain English meaning, which holds that man generally has a choice among several alternatives all of which are possible; free will in this meaning is in direct contradiction to determinism.
The argument for determinism that I regard as the most plausible and intelligent, is the argument that free will requires irrationality. According to this argument, a rational person will accept as truth whatever is proved to be true by facts and logic, and will take whatever action is rationally appropriate to the situation; his beliefs and actions will therefore be completely determined by antecedent causes, i.e. by the facts he observes; even if he has a faculty of free will, it will be irrelevant to his thoughts and actions. Free will, therefore, if it exists at all, can only be relevant to an irrational person whose beliefs and actions have no reasons, who accepts ideas arbitrarily and acts on whim.
This argument can be used in two directions. As used by the Romantics, or by the Existentialists, it starts from the premise that free will exists and is crucially important, and claims to prove that rationality should be rejected. The more intelligent use of this argument is in the other direction; starting from the premise that rationality is an important virtue, and claiming to prove that free will, if it exists at all, is unimportant and useless.
This argument in itself only provides a motivation for accepting determinism, rather than actual evidence for its truth. A corollary of this argument, however, is that there is a contradiction between free will and the theory of evolution. It is a scientifically established fact that man evolved through natural selection; therefore, to explain free will, we need to explain how it could have developed in the process of evolution, and that requires explaining what selectional advantage free will has. But if the only effect of free will is to make it possible for man to act arbitrarily and irrationally, then it not only provides no selectional advantage, it is positively harmful. Conceivably it could still have developed through evolution as an accidental side effect of some other feature which has selectional advantage; but that seems far-fetched, and makes it unlikely that free will could have developed through evolution.
I think this is the one argument that advocates of free will have, until the past few decades, been unable to adequately answer. Nor were they able to explain what advantage free will provides. The first to adequately answer this was Ayn Rand, with her crucial insight that the center of free will is in man's ability to direct his mental focus.
As Rand identified, man's basic choice is in focusing his mind: whether to focus it, to what level, and what to focus it on. Man is free to keep his mind in full focus, to drift automatically without focus, or to actively evade and refuse to think. Man is also free to focus on all relevant facts and considerations, and make a deliberate, conscious effort to think of additional factors and find anything that might have been missed so far; to limit his thinking to the factors that he notices easily and think no further; or to actively refuse to consider some of the facts. (For more discussion of Rand's theory of free will see Leonard Peikoff, Objectivism: the Philosophy of Ayn Rand, pp. 55-69)
All other choices man makes are results of this basic choice. Man does not make a directly free choice on what ideas to accept; his freedom is in controlling what facts and arguments his mind focuses on, and this selection of facts and arguments determines what ideas he then accepts as truth. Man does not make a directly free choice on what action to take; his freedom is in controlling what considerations relevant to his decision his mind focuses on, and this selection of facts and considerations determines his actions.
This provides the answer to the claim that free will allegedly requires irrationality. Man's free will does not mean that his ideas or actions are without reason; it means that he chooses the reasons, through his control over mental focus. To someone committed to rationality, free will does not lose its relevance; on the contrary, given the effort required for maintaining full focus and for considering all the relevant facts, free will becomes, if anything, of even greater relevance to a person committed to rationality.
Rand's insight also provides a basis for a likely hypothesis regarding the evolutionary selectional advantage of free will. (The hypothesis I present here was presented by David Kelley in his lecture course The Nature of Free Will, presented at the Portland Institute, 1986. As far as I know it has never been presented in any printed or online publication.) The function of free will becomes an extension of the function that William James pointed out for consciousness: allowing higher animals to deal with the greater range of information they receive from the environment and the greater number of possible actions they have to choose from.
Plants and lower animals have a quite narrow range of possible actions that they can take, and their actions depend on relatively simple information from the environment - detecting the presence and direction of sunlight, the presence of nutrients, etc.; they are therefore able to control their actions through direct stimulus-response relations. But higher animals have a much wider range of possible actions, and their actions have to depend on a lot of information from the environment. This makes necessary for the animal to have more complex internal control over its actions, and the function of consciousness is to provide this control.
Man, because of his ability for conceptual thought, has a range of possible actions, and a range of information he can receive from the environment, that is far wider than that of even the most developed of other animals. The range is not only greater in quantity, but in kind; rather than choosing from a pre-existing set of options, man's conceptual faculty allows him to create new kinds of actions and of information, thus making his range of actions and of information unlimited. For example, consider that 20 years ago, reading information on web sites was not yet one of the possible actions man could choose from; and yet now it has become one (in fact, if you are reading this paper on a web site, you are engaged in this new kind of action right now.) And this means that in making decisions, the range of information that could potentially be relevant, that might need to be considered, is also unlimited. For every decision man needs to make, he has to determine what information is relevant, and what information he would spend his time thinking about in making the decision; and because the range he must choose from is unlimited, the choice can't be made by any deterministic mechanism. The function - and the evolutionary selectional advantage - of free will, therefore, is in making it possible for man to direct his decisions based on an unlimited range of possible information; and therefore to make possible the functioning of the conceptual faculty, which creates this unlimited range.
The above is, of course, only a hypothesis; whether this really is the selectional advantage that led to the development of free will through evolution, would need to be proved or disproved by scientific research. But the point is, Rand's identification of mental focus as the center of free will allows us to identify the advantage that free will provides to man. Therefore there is no longer any basis for regarding the development of free will, through evolution by natural selection, as far-fetched or inexplicable; it is just as likely as the evolution of any other useful trait.
The second argument I address in this paper is the claim that determinism is demonstrated by inductive generalization.
For inanimate objects (at least above the level of subatomic particles) and for vegetative biological processes (i.e. all processes in bacteria and plants, and those processes in an animal's body that do not involve consciousness), many deterministic laws have been discovered, and verified by the scientific method, that precisely predict the actions of all these entities under given circumstances. Some of the simpler of these deterministic laws are familiar to us from our everyday experience with such objects; more complex ones are discovered and verified by scientific experimentation. With so many deteriministic laws known and verified, and with so much repeated evidence for each one of them, the evidence is extremely strong that such entities behave by deterministic laws. So, the argument goes, we should generalize from this to the conclusion that all entities in nature, including man, are similarly deterministic; that the knowledge needed to predict man's actions, even though it does not exist today, can eventually be discovered, and that the introspective evidence of making choices among several possible alternatives must be an illusion.
This argument is a classic case of the fallacy of over-simplified generalization. Given the importance of consciousness, and the great differences it makes to the animal's actions; and given that the brains of man and the higher animals are a vastly more complex structure than any other we know of; it is a totally fallacious use of induction to try to extrapolate from the deterministic laws discovered for other entities to the assumption that the actions of man and the higher animals must also be subject to deterministic laws. Man in particular, in addition to consciousness, also has the capacity for conceptual thought, which again makes a crucial difference in his actions; so extrapolating determinism based on inductive evidence from inanimate or vegetative processes is even more far-fetched in the case of man.
Another point to note, regarding the argument from inductive generalization, is that it applies to subatomic particles just as much as to man. Anyone who regards this argument as valid support for determinism should equally regard it as support for the "hidden variables" interpretation of quantum mechanics, claiming that subatomic particles are subject to deterministic laws as well. Any determinist who claims that determinism is proved by inductive generalization, and at the same time is willing to accept non-determinism at the subatomic level, is evidently not thinking clearly.
The third argument is the most mainstream of the defenses of determinism. It dates back to the Greek atomists, and it is still today the main basis that contemporary determinists present for their position. This is the idea that causality requires determinism; that the only alternative to determinism, and therefore a necessary consequence of accepting free will, is to reject, or at least claim exceptions to, the law of causality.
The claim that causality requires determinism is almost always presented as a bald assertion, with no argument. The only support it has is the assumption that causality is a relation between events; the law of causality means that for every event there has to be some prior event which is its cause.
The alternative view of causality, dating back to Aristotle, is that causality is a relationship, not between one event and another, but between an entity and its actions: the way a thing acts (including the way it reacts to the actions of other entities) is a function of its nature. While it is often convenient to refer to some action as the "cause" of a subsequent action, such usage is derivative; primarily, an action's cause is the nature of the acting entity. For example, the motions of atoms or ions are caused by their mass, electric charge, etc., which determine how the forces operating on them affect their movement. If the nature of these entities were different, then they would act differently in response to the same external forces. In the case of living things, whose actions are self-generated (i.e., the action's direction and energy come from sources internal to the acting entity), entity causation becomes agent causation; the contraction of a muscle is caused by the nature of the animal's muscular and nervous systems. This understanding of causality makes it possible to see how human agents, whose nature includes the ability to weigh alternative courses of action and deliberate about them, and consequently the capacity for genuine choice, act in accordance with causality, not in any way in contradiction to it.
What basis is there for the view of causality as a relation between events? That view is accepted by many modern philosophers as unquestioned dogma. In the writings of determinists it is usually left to implication, as implicit support for the assertion that accepting free will would require us to give up causality; but it is never supported by arguments. The idea of causality as a relation between an entity and its actions is not argued against, but rather never mentioned; many determinists write as if the idea simply never occured to them, and seem unable to even comprehend the possibility.
For example, Daniel Dennett, after quoting a statement from Roderick Chisholm explaining the concept of agent causation, responds:
How does an agent cause an effect without there being an event (in the agent, presumably) that is the cause of that effect (and is itself the effect of an earlier cause, and so forth)? Agent causation is a frankly mysterious doctrine, positing something unparalleled by anything we discover in the causal processes of chemical reactions, nuclear fission and fusion, magnetic attraction, hurricanes, volcanoes, or such biological processes as metabolism, growth, immune reactions, and photosynthesis (Dennett, Freedom Evolves, p. 100).
On the view of causality as a relation of an entity to its actions, all causation—including all the processes Dennett lists—involves as cause the entity rather than some earlier event. And all biological processes—including all the ones Dennett lists—are cases of self-generated action. Agent causation, therefore, far from being "mysterious" and "unparalleled by anything," is ubiquitous in nature; it is only Dennett's unquestioning acceptance of the event-event model of causality that makes him blind to this.
The claim that free will is a contradiction to, or an exception to, the law of causality, is thus based on the dogmatic acceptance of an unquestioned, and usually unidentified, philosophical premise: the event-event model of causality. Once that dogma is questioned, this argument for determinism is left without basis.
The fourth argument is again an argument dating back to the Greek atomists.
The argument is that the physical constituents of the human body - atoms, or molecules of water and various chemicals - do not have any power to initiate action, or to choose among different possible actions; the causal laws governing their actions are completely deterministic. Therefore, the same must be true for the human body as a whole.
The argument is seldom stated as clearly and explicitly as I just did; it is much more often stated vaguely or left to implication. When stated explicitly, it becomes a blatant case of the logical fallacy of composition. What is true for the components of an entity is not necessarily true for the entity as a whole. The fact that atoms or molecules do not have any power to choose among different possible actions, does not preclude such powers from emerging in their combination in the human body or brain.
In general, the debate between determinism and free will offers little that is new. It continues today with the same arguments and counter-arguments that have been used for centuries. This is especially remarkable in the arguments for determinism, given the bombastic statements determinists often make about the alleged modern, scientific nature of determinism; such statements are invariably in the form of vague generalizations, while the only specific arguments they offer in support of determinism are the four argument I discussed above, which are the same arguments determinists have used throughout history.
There are only two exceptions to this rule, only two original arguments that appeared in the debate in the past few decades. One, on the side of free will, is Ayn Rand's identification of mental focus as the center of free will, as I discussed above. The second one, on the side of determinism, is the final argument I am going to address in this paper: the appeal to the Libet experiments.
Benjamin Libet performed a series of experiments in the early 80s, in which subjects were asked to occasionally move their hands at unpredictable intervals, and to note and report the time in which they made the decision to move their hand; an EEG measurement, taken during the experiment, showed that the brain waves preceding the hand movement started some fraction of a second before the time the subjects reported as the time they made the decision. These results have been trumpeted by some as proof of determinism, demonstrating that decisions we apparantly make by our free will are actually the result of physical brain processes that occur before we become conscious of the decision.
In fairness to Libet, it should be noted that he is not a determinist, and this was not his own interpretation of his results. Libet's own interpretation, rather, is that his results demonstrate that free will is purely negative. Libet's theory is that our actions are the result of urges created by brain processes that are outside our control, but we are able to consciously override these urges and decline to act on them; this conscious "veto power" is Libet's view of free will. As I discuss below, Libet's interpretation does not follow from his experiments either; but it is not nearly as stupid or obviously fallacious as taking the results as evidence of determinism.
The Libet experiments in fact do not have any interesting implications regarding free will, for two basic reasons. First, the experiments created a situation in which the subject's decision (at what time to move his hand) is necessarily arbitrary; there is no possible reason for the subject to move his hand at one time rather than another. This situation is therefore fundamentally different from actual real-life situations in which people make decisions, and it seems plausible that that difference would completely change how free will operates. Even if the Libet experiments proved anything about decision-making in the laboratory situations created in the experiments - situations in which a decision is made completely arbitrarily, with no reasons - it would be impossible to draw any conclusions from that about the decision-making process in real-life situations, in which a person considers reasons for and against a course of action in order to make his decision. (This problem with the Libet experiments was pointed out by David Kelley, in the Q&A period of his lecture series The Nature of Free Will).
Second, the Libet experiments don't prove anything even about the decision-making process in the laboratory situation they created; both the determinist interpretation of the experiments and Libet's own interpretation ignore the fact that perception takes time. Libet asked subjects to report the time at which they decided to move their hands, with accuracy down to a fraction of a second, by watching a clock-face with a fast-moving dot, and noting the position of the dot at the moment they made the decision. But the perception of the clock-face - like all visual perception - is a process that takes time; it is therefore not at all clear that the dot position reported by the subjects was in fact the position at the moment they became conscious of the decision. Furthermore, generally when we intentionally perform a movement, we monitor that movement with our vision; and so our visual processes when intentionally performing a movement (such as the hand-movement performed in the experiment) will naturally be alert to the moment at which the movement occurs, which will be some fraction of a second after the movement is consciously initiated; this means it is plausible that the dot's position on the clock-face perceived by the subject as being at the same time he became conscious of the decision, was in fact at a time some fraction of a second later. (This problem with the Libet experiments was pointed out by Daniel Dennett in Freedom Evolves, pp. 227-242. While in general I have a very low opinion of Freedom Evolves - see my review of it in Navigator, December 2003 - the section on the Libet experiments is the one section of the book that contains useful and interesting information. The fact that Dennett, himself an avid advocate of determinism, took the time to discuss and expose the problems with bad arguments presented in support of his own position, is the one aspect of his book that deserves respect.)
As I noted in the introduction of this paper, the debate between free will and determinism is often very frustrating and seems irresolveable, mostly because of the failure of determinists to explicitly and clearly state their case. I think much of this failure comes from the fact that many determinists do not accept determinism for any rational reason at all, but because of its emotional appeal.
Determinism is treated by many of its advocates as a dogma, no different from the dogmas of religion, and this gives it much of the same emotional appeal that religion has. For some determinists, their reason for accepting determinism is that in their subculture, denying determinism would be regarded in the same way as denying the existence of god would be regarded in other subcultures.
Religious writers and preachers often insist that lack of belief in god must be a sign of immorality or of some other kind of inferiority. Such claims exert two kinds of emotional pressure on the reader or listener: first, you are threatened that if you don't believe in god, you will be accused of immorality or inferiority; second, if you do believe in god, you are offered the emotional reward of being able to feel superior to the benighted, immoral, inferior unbelievers. The writings of many determinists offer a similar emotional appeal. The writer insists that disagreement with determinism must be a sign of immaturity, of fear of facing the truth, or of some other kind of irrational motive; the reader is thus threatened that if he rejects determinism, he will be regarded as immature or as a coward, and at the same time offered the reward, if he is willing to accept determinism, of an opportunity to feel superior. (See Daniel Dennett's Freedom Evolves for an especially egregious example of the use of this tactic.)
Determinism also offers a third kind of emotional appeal, resulting from the influence of Kantian philosophy. Kant's division between the noumenal and phenomenal worlds has had such a profound influence on modern philosophy, that it has its influence not so much as explicitly accepted philosophy but as deeply seated, subconscious premises; most philosophers, and most people who have taken philosophy in college, come to have an emotional expectation that the world we perceive will have little or no similarity to the world as it really is. And so determinism's inconsistency with the directly perceived, self-evident facts becomes an emotional point in its favor; determinism becomes appealing as a confirmation of the Kantian expectation that our perceptions of the world should be wrong.
My purpose in this paper was to look past what is emotionally appealing to what is supported by logic and evidence, and consider what rational support determinists have offered for their position. I believe my list of five arguments is a complete list. Of these, the first one - the argument that free will allegedly requires irrationality, and its corollary that free will provides no evolutionary selectional advantage - is the only one that advocates of free will have ever had trouble adequately answering; and for the past 50 years, since Rand's identification of mental focus as the center of free will, that argument has also been answered. There is therefore no remaining rational basis for determinism.