Elbow Room

Background on Dennett
Dan Dennett is in the philosophy department at Tufts University. He is a philosopher who attempts to pay attention to recent advances in science and deal with the philosophical implications of science, and as such, his philosophy appeals to me, a working biologist. You may have had a chance to see him on TV in the series A Glorious Accident. He is fairly well known for his early work which centered on an analysis of people as complex systems with intentions (summarized in his book, "The Intentional Stance", 1987) and for his work which admirably explains biological evolution to non-scientific academicians (such as his book "Darwins Dangerous Idea"). He first came to my attention through his association with artificial intelligence researcher Douglas Hofstadter (author of Godel, Escher, Bach, 1979) when they published a book about the self and the soul called "The Mind's I" in 1981. More recently, Dennett helped provoke some biologists into starting to talk about consciousness by publishing his book "Consciousness Explained". There is an email-based discussion group for Dennett's work (see the Thinknet BBS DENNETT discussion at ) and his home page is at (http://www.tufts.edu/~ddennett/).

In 1983 Dennett delivered the John Locke Lectures at Oxford on the topic of Free Will. In 1984 these ideas were published in the book "Elbow Room: The Varieties of Free Will Worth Wanting" (review). In this book, Dennett explored what it means for people to have Free Will. The title, Elbow Room, is in reference to the question: are we deterministic machines with no real freedom of action or do we in fact have some elbow room, some real choice in our behavior?

A major task taken on by Dennett in Elbow Room is to clearly describe just what people are as biological entities and why they find the issue of Free Will to be of importance. In discussing what people are and why Free Will matters to us, Dennett makes use of an evolutionary perspective. Dennett describes the mechanical behavior of the digger wasp Sphex. This insect follows a series of genetically programmed steps in preparing for egg laying. If an experimenter interrupts one of these steps the wasp will repeat that step again. For an animal like a wasp, this process of repeating the same behavior can go on indefinitely, the wasp never seeming to notice what is going on. This is the type of mindless, pre-determined behavior is what people can avoid. Given the chance to repeat some futile behavior endlessly, people can notice the futility of doing so, and by act of free will do something else. We can take this as an operational definition of what people mean by free will. Dennett points out the fact that as long as people see themselves as able to avoid futility, most people have seen enough of the Free Will issue. Dennett then invites all who are satisfied with this level of analysis to get on with living while he proceeds into the deeper hair-splitting aspects of the Free Will issue.

From a biological perspective, what is the difference between the wasp and a person? The person can, through interaction with its environment, construct an internal mental model of the situation and figure out a successful behavioral strategy. The wasp, with a much smaller brain and different genetic program, does not learn from its environment and instead is trapped in an endless and futile behavioral loop that is strictly determined by its genetic program. It is in this sense of people as animals with complex brains that can model reality and appear to choose among several possible behaviors that Dennett says we have Free Will.

The deeper philosophical issue of Free Will can be framed as a paradox. On one hand, we all feel like we have Free Will, a multitude of behavioral choices to select among. On the other hand, modern biology describes humans as mechanisms that follow all of the same deterministic rules as wasps or inanimate objects. How do we reconcile our feeling of Free Will with the idea that we are mechanical components of a mechanical universe?

What about determinism? When we say that a person chooses among several possible behaviors is there really a choice or does it just seem like there is a choice? Do people just (through the action of their more complex brains) simply have better behaviors than wasps, while still being totally mechanical in executing those behaviors? Dennett gives his definition of determinism on page one: all physical events are caused or determined by the sum total of all previous events. This definition dodges a question that many people feel should not be dodged: if we repeatedly replayed the universe from the same point in time would it always reach the same future? Since we have no way of performing this experiment, this question is a long-term classic in philosophy and physicists have tried to interpret the results of other experiments in various ways in order to figure out the answer to this question. A related current fantasy game for physicists is to imagine that there are multiple universes and every time there is quantum indeterminacy each possibility occurs and new universes branch off. Since the 1920s physicists have been trying to convince themselves that quantum indeterminacy can in some way explain Free Will. Dennett dismisses this idea as silly. How, he asks, can random resolutions of quantum-level events provide people with any control over their behavior?

Since Dennett wrote Elbow Room in 1983 there has been a futile, but still on-going attempt by some physicists to answer this question by assuming that the brain is a device for controlling quantum indeterminacy so as to construct behavioral choice. Dennett argues that such efforts to salvage Free Will by finding a way out of the prison of determinism are wasted.

Dennett discusses many types of Free Will. Many philosophers have claimed that determinism and Free Will are incompatible. What the physicists seem to be trying to construct is type of Free Will that involves a way for brains to make use of quantum indeterminacy so as to make choices that alter the universe in our favor, or if there are multiple universes, maybe brains can choose among the possible universes. Dennett suggests that we can have another kind of Free Will, a type of Free Will which we can be perfectly happy with even if it does not give us the power to act in more than one way at any given time. Dennett is able to accept determinism and Free Will at the same time. How so?

The type of Free Will that Dennett thinks we have is finally stated clearly in the last chapter of the book: the power to be active agents, biological devices that respond to our environment with rational, desirable courses of action. Dennett has slowly, through the course of the book, stripped the idea of behavioral choice from the idea of Free Will. How can we have Free Will if we do not have real behavioral choice? Dennett tries to substitute control for choice. If our mechanical brains are in control of our behavior and our brains produce good behaviors for us, then do we really need choice? Is an illusion of behavioral choices just as good as actual choices? Is our sensation of having the freedom to execute more than one behavior at a given time really just an illusion? Dennett tries not to beat his readers over the head with this issue, but I think he should have.

If all people have is an illusion of behavioral choice, if people are just machines behaving in the only way they can, then what about personal responsibility? How can we hold people responsible for and punish them for their behaviors if they have no choice in how they behave? Dennett gives a two part answer to this question. First, we hold people responsible for their actions because we know from historical experience that this is an effective means to make people behave in a socially acceptable way. Second, holding people responsible only works when combined with the fact that people can be informed of the fact that they are being held responsible and respond to this state of affairs by controlling their behavior so as to avoid punishment. People who break the rules set by society and get punished may be behaving in the only way they can, but if we did not hold them accountable for their actions, people would behave even worse than they do with the threat of punishment. This is a totally utilitarian approach to the issue of responsibility, there is no need for moral indignation when people break the rules of proper behavior. Is it, then, moral to punish people who are unable to do other than break a rule? Yes, people have the right to come together and improve their condition by creating rules and enforcing them. We would be worse off if we did not do so. Again, an argument for utility.

One final issue, if people do not have real behavioral choices, why not collapse into fatalism? Again, Dennett's argument is that we may not have behavioral choice, but we do have control of our behavior. Dennett asks us to look around at the universe and ask, can I even conceive of beings whose wills are freer than our own? For Dennett, the answer to this question is, no, not really. In Elbow Room he tries to explain why all the attempts that people have tried to make to prove that people have behavioral choice have failed and are, in the final analysis, not really important anyhow. As humans, we are as much in control of our behavior as anything in the universe. As humans, we have the best chance to produce good behavior. We should be satisfied with what we have and not fret over our lack of behavioral choice.

As usual, I find it very hard to disagree with Dennett. My largest complaint about Elbow Room is that it does not satisfactorily deal with the issue of why we feel so strongly that we do have behavioral choice. I agree with Dennett that we do not have choice, but why do we feel like we do? My answer to this question is that our sensation of having behavioral choice has been carefully selected by evolution. The well developed human sensation of having Free Will and being able to select among possible behaviors has strong survival value. People who loose the feeling that they can plan alternative behaviors and execute their choice of possible behaviors tend to become fatalistic and stop struggling for survival. As Dennett writes, Belief in Free Will is a necessary condition for having Free Will. When we are planning for the future and thinking about possible actions to take in the future, we are expending considerable amounts of biologically expensive resources (brain power). Evolution has designed us to feel strongly that all of our effort of planning pays off, that we control what we do. If this connection between our brains efforts to model reality and predict the future and so make possible good outcomes is disconnected from our sense of self and our will, then fatalism and self-destructive behaviors are close at hand.

So at the end of our philosophical hair-splitting, we reach the same conclusion as the average man on the street, but we have some additional baggage. If we accept Dennett's arguments, then we recognize that we have no real behavioral choices, but we continue to behave as if we do. I would say that when we feel like we are making a behavioral choice, this is a very convenient way for a brain to make sure that it keeps planning and struggling for survival. Our conscious thoughts never see the detailed working of our mechanical brains, we can never directly sense that we never really have behavioral choice, that our brains are deterministic machines. Our brains are designed to present us with a tantalizing array of apparent possibilities and the sensation that we have choice, while in reality there is only one way things will work out in the end. So, at the end of our philosophical journey we must be satisfied that our brains are in control and we must content ourselves with behaving as if we have behavioral choice, even though we know we do not. Nature has played a devious trick on us. Grin and bare it. It could be worse, think what it would be like to be a wasp.

Go to John's Book Page.

Go to John's Home Page.