philosophy is for everyone
philosophers should know lots
Free Will Versus Determinism
by Stuart Burns
The arguments for Determinism come primarily from the realm of the sciences. The more science learns about the nature of Man and the Universe, the more likely it seems that the future is predictable. If the future is predictable, then it is possible that the decisions you think you make freely, are not so free.
The arguments for Free Will come primarily from the fact that our concepts of morality and personal responsibility for our actions are based on the assumption that the acting agent is able to choose otherwise. We need to be able to assign responsibility, bestow blame and praise, and allocate punishments and rewards.
The essence of "not responsible" is "not able to influence the outcome". The fear is that if Determinism is true, then we cannot choose otherwise than we do, and therefore are not responsible for our actions. Morality and civil law disintegrate into chaos.
The "Compatibilist" argument I will present here maintains that Free Will is not incompatible with Determinism. I will argue that once one understands what we are really talking about when we talk about "Free Will", we will realise that Determinism is actually no threat to Free Will. And that the apparent conflict between the two arises from a fundamental misconception of just what "Free Will" is.
(This is an abridged version of a longer essay that appears on the Web at http://www3.sympatico.ca/saburns/pg0403.htm)
RESOLVING THE CONFLICT
Free Will is incompatible with Determinism. Man has Free Will. Therefore Determinism is false. This is the argument of those who consider the need for Free Will to be more important than the evidence from science for Determinism. There are essentially two different approaches to this way of resolving the conflict.
The scientific approach is to attack the underlying premises of scientific determinism. Quantum Indeterminacy is a favourite escape hatch. As is the "Multiple Worlds" interpretation of Quantum Physics. But scientific libertarians face a serious challenge. If your actions are not the result of your moral evaluation of the desirability of the consequences, then you are not morally responsible for the results. Depending on scientific indeterminacy to escape from Determinism is morally equivalent to rolling the die instead of making your own choices.
The dualist approach is to maintain that whatever it is in the human mind that exercises Free Will (let's call it a "soul"), it is not subject to the constraints of materialist science. But dualist libertarians also face a serious challenge. The dualist must provide a means whereby the immaterial soul, not subject to the constraints of materialist science, initiates the clearly materialist nerve impulses that generate our behaviour. So far, there has been no success at providing such an explanation.
Free Will is incompatible with Determinism. Determinism is true. Therefore Man does not have Free Will. This is the argument of the "Hard Determinist" or "Hard Incompatibilist". The argument is that "Free Will" is an illusion. Perhaps a necessary one, in order for us to function properly. But an illusion none the less. The challenge that the Hard Determinist faces is the relevance of such a conclusion. It seems difficult to understand how such a conclusion matters. The critical social importance of "Moral Responsibility" demands that we ignore the Determinist conclusion, and proceed on the alternate hypothesis that we do indeed have Free Will.
Free Will is not incompatible with Determinism. Determinism is true, and Man does have Free Will. This is the argument of the Compatibilist. It is based on the argument that the concept of "Free Will" is poorly defined and misunderstood.
The remainder of this essay will discuss the Compatibilist concept of "Free Will" and present the arguments for the compatibility of Free Will and Determinism.
COMPATIBILIST FREE WILL IN A DETERMINISTIC REALITY
In order to explore the concept of "Free Will" in more depth lets consider a game of chess between you and a chess program running on your computer. And let's examine what is taking place from the perspective of Daniel C. Dennett's "Stances". (As outlined in his "The Intentional Stance", and "Elbow Room")
It is your move. You examine the chessboard, identify some possible moves, and evaluate their desirability against your knowledge of chess strategy and your projections of how your opponent will respond. You choose a move, and (say) move your bishop. You could have chosen otherwise. It has all of the characteristics of a "Free Will" decision. Yet if we replayed the game tomorrow, and reached exactly the same point in the game, your analysis would be the same, and you would again choose to move the bishop, for the same reasons as you did today. (Assuming, of course, that you do not learn anything new in the interim.)
But now consider the move the computer responds with. The chess program examines the chessboard, identifies some possible moves, and evaluates their desirability against its knowledge of chess strategy, and its projections of how its opponent will respond. The program chooses a move, and (say) moves its knight. Did the computer program not freely choose to make that particular knight move? There was nothing in the situation that would render that particular move forced. It was not coerced. The move was not the result of random quantum events, or the flip of a coin, because a computer is specifically designed to preclude such things. It has clear strategic reasons for the move. There were a number of attractive looking legal moves available to it. It could have chosen otherwise. Doesn't this choice also have all of the characteristics of a "free will" decision? Or is there something different going on here? If this is not an exercise of "Free Will", what's the difference?
The Physical Stance.
At the physical level, the chess program running on your computer is a completely deterministic chunk of software, running on a completely deterministic chunk of hardware. The designers of computer chips and software programs go to great lengths to design out any potential quantum indeterminacies or "cause-less" idiosyncrasies. The process of testing consists of providing the same inputs repeatedly, and ensuring that the outputs are identical to what is expected. The nature of binary coding in the hardware and the software is designed to mask the microvariability of electrical voltages, currents, and charges. A more complete model of a deterministic universe could not be found.
Yet, at the physical level the chess program running on the computer between your ears is also a completely deterministic chunk of software, running on a completely deterministic chunk of hardware. The processes of evolution that designed the hardware between your ears had to design out any potential quantum indeterminacies or "cause-less" idiosyncrasies. The process of natural selection, of differential rates of procreation, would weed out any design where the same inputs did not generate the "expected" outputs from generation to generation. Evolutionary adaptations "work" only so far as they enable marginally more successful rates of procreation over may generations. Biologists have demonstrated that the nature of our genetic coding masks any microvariability of chemical affinities when replicating DNA, translating DNA to RNA, or RNA to proteins. A more complete model of a deterministic universe also could not be found. There is no meaningful difference between the computer chess program, and the human chess player, at the physical level.
The Intentional Stance.
Now think of the computer chess program not as manmade software running on manmade hardware, but as an intelligent agent. Think of the computer program the same way you would think of another person. Here is a self-interested, self-governing, self-motivating agent who wants, thinks, desires, values, contemplates, evaluates, and - chooses. The computer wants to win the game. It thinks that when I move my bishop to that spot, I am threatening its queen. It fears that threat. It hunts for a good move to protect itself from my threat. And, from the available alternatives freely chooses what it judges is a good response. Here now, at this level of analysis, we have something that looks like "Free Will".
Behaviour that looks like "Free Will" appears when we view whoever or whatever we are dealing with as an independent agent, and dismiss as irrelevant detail all of the mess at the physical levels that muddies our understanding of events. What matters to us when we are playing chess against a computer program, is how to win the game. To best accomplish that goal, the details of program design, computer language design, computer chip design, or the physics of electrons through semi-conductors are all irrelevant. What is relevant is how well we can predict the behaviour of our opponent. And it is easier, simpler, and less resource expensive for us to predict what the opposition will do, if we adopt the stance that the opposition is something just like us. We do better at predicting the behaviour of opponents, of friends, and of ourselves, when we treat them as agents who feel like they have Free Will. We invest "Free Will" (among a whole host of other self-like feelings and motivations) in ourselves, our friends, our enemies, our pets, our predators, our prey, even our cars, our folding chairs, and our computer programs. It is an evolutionarily adaptive strategy.
"Free Will", then, is a concept from the Intentional Stance, applicable to Intentional Systems. We feel we have Free Will because we can observe ourselves thinking, desiring, valuing, contemplating, evaluating, and - choosing. Determinism, on the other hand, is a concept from the Physical Stance, applicable to physical systems. We are not aware of the physical processes that take place within us. The apparent conflict between the two concepts disappears once one firms up our understanding of just what we are talking about when we discuss "Free Will". It is a "frame error" to compare or contrast the two. Both can be, and are, true. A completely deterministic device - be it a mind or a computer program - can have and exercise Free Will.
Just what exactly is "Free Will"? We can start by taking a minimalist stance, and define "Free Will" as whatever aspect of the mind would be necessary to enable "Moral Responsibility". That means that "Free Will" is that capacity of the mind that chooses and could choose otherwise.
The verb "to choose" means to evaluate the alternatives and select that alternative that appears to be the "best" according to some standard. The verbs "to decide" and "to judge" mean the same thing. If you do not deliberate over your choice, and have no reasons or justification for your choice when you make it, then you have not "chosen".
"Free Will" cannot mean a "cause-less" source of choice. How can such a "cause-less" source for your choices constitute your choices. Any attempt at such an explanation would have to address how a "cause-less" source becomes an evaluation of alternatives.
"Free Will" cannot mean a "random" source of choice. You are not normally held directly responsible for the "accidental" results of the roll of a die, because they are not your choices.
"Actions are, by their very nature, temporary and perishing; and where they proceed not from some cause in the character and disposition of the person who performed them, they can neither redound to his honour, if good; nor infamy, if evil." - David Hume (An Enquiry Concerning Human Understanding, Section VIII, Part 2). In short, how you choose between alternatives is determined by who and what you are. So if your choice is caused by anything, it is caused by the sum total of your past history because that is what you are. If you make a deliberate choice, you do so because you have reasons. And given those reasons, you would not have chosen otherwise.
For any choice to be your free choice, and not the result of someone else's control of, or influence on you, you have to make the choice based on your own understanding of your options, as evaluated by the values and priorities you have learned through experience. A choice is not freely yours if it is not based on your beliefs and your character, your experiences and your goals. Yet with all of these constraints, you still feel that the choice is a free one. That is because these constraints that I have listed are what you are. In many cases, your friends can predict the way that you will choose, because they know who you are. And if you were to choose other than you would normally choose, your friends would look for some unusual reason for this unusual choice. If these constraints are not operational when you make your choice, then you are not making the choice.
What does it mean to be able to "choose otherwise" than you did? The cause of your choice was your reasons and justifications. Given those reasons and those justifications, and given your character, beliefs, values, and experiences, you judged the choice you made to be the best of the available alternatives. Unless you learn something new to change your evaluation of the situation, you would be making the wrong choice to choose otherwise. Yet it is certainly possible that you would have chosen otherwise, had your reasons and justification been different. It is, therefore, an error in conception to presume that Free Will must involve an ability to choose otherwise, given the reasons and justification that exist. But it is not at all extraordinary to understand Free Will as including an ability to choose otherwise than you did, had the reasons and justifications been different.
The Fallacy of Fatalism
The doctrine of Determinism merely asserts that nothing happens without a cause, that every state of affairs is the outcome of a preceding state of affairs. But more importantly, Determinism does not assert that a given future will unfold regardless of what you or I may do to promote or prevent it. Yet this is the assumption implicit in Fatalism. Just because a Compatibilist believes that Determinism is true does not imply a Fatalist attitude about the uncontrollability of the future. The future is what we choose to make it.
Moral Judgements from a Compatibilist
The primary challenge facing the Compatibilist, is explaining the purpose and function of moral judgements. The answer is that we are all learning machines. Unlike the chess program, if we replay the match, we will bring to the rerun all the things we have learned in the interim. We learn from all experiences, and we can never exactly rerun a "test scenario" and expect exactly the same output.
We learn to improve our processes of evaluation and choice by distinguishing "good" choices from "bad" ones. Therefore, as an aide to learning - both for ourselves and for others - it pays to advertise those decisions that are notable for being "good" and "bad". Passing judgement on one's own choices, and the choices of others, serves the purpose of making plain how such choices "ought" to be made. Hence the purpose of praise and blame, punishment and reward.
Does the morally culpable miscreant deserve to be punished? What is the meaning of "deserve" in this context? The dictionary says it means "be worthy of, merit, earn". Surely then, the miscreant's behaviour is worthy of, merits, has earned the punishment? The objection that the miscreant could not have chosen other than he did, and is therefore not morally responsible, is invalid. What is deserving of punishment is the behavior observed, not the myriad of causes that generated it - determined or otherwise.
So the answer is "Yes!" For three reasons:
The human brain is an incredibly complex, and deterministic, biological computer. "Free Will" is just exactly that mental process that evaluates, deliberates, and chooses the most appropriate response to the current situation.
Free Will is not represented by choices, decisions, or judgements that are uncaused, or caused by random or indeterminate events. Free Will is represented by choices, decisions and judgements that are caused by your character, beliefs, values, and experiences, and the reasons and justifications that you perceive at the time.
Given the character, beliefs, values, and experiences, and the reasons and justifications that are perceived at the time, you could not have chosen other than you did. But given any difference to this long list of inputs, and you might have chosen other than you did.
The judgements we make, and the emotional reactions we feel, about the choices we and others make, serve the purpose of training the evaluative processes of the mind. We cannot learn to choose more wisely, unless we can recognise when someone makes a particularly good or bad choice. We react the way we do because we are learning machines, and that is how this particular kind of learning machine learns.
© Stuart Burns 2004
Web site: http://www3.sympatico.ca/saburns/