Figure 1

      On this picture, keeping everything in the universe exactly the same up until a particular moment in time, say time t, the agent’s choice would be causally determined such that they could not have done otherwise in exactly those circumstances.

Of course, some compatibilists have argued that terms like can, power, and ability should be given a conditional or hypothetical analysis. These compatibilists maintain that when we say that an agent can (i.e. has the power or ability to) do something, we mean the agent would do it if the agent wanted (desires or chose) to do it. On this approach, to say “you could have done otherwise” would amount to the counterfactual claim that you would have done otherwise, if (contrary to fact) the past (or the laws of nature) had been different in some way. The problem, however, with analyzing I could have done otherwise as I would have done otherwise if I wanted (or chose) to, is that it only invites the obvious question: Do I have the freedom or ability to want (or choose) differently? For the compatibilist argument to work, it would have to show that the ability to want otherwise is itself compatible with determinism, and here the conditional approach will not help without causing a regress. Furthermore, this analysis would still fail to preserve the unconditional ability to do otherwise – hence, agents would still lack the ability to do otherwise in exactly the same set of circumstances keeping the laws of nature and antecedent conditions fixed. It would seem, then, that understanding the “ability to do otherwise” by means of conditional analysis would amount to nothing more than “the little could that would but can’t so won’t.”

      Setting aside for the moment the question of whether the “ability to do otherwise” is a necessary condition for free will, and without getting into a compatibilist defense of free will quite yet, I simply want to know whether you agree with my summary of determinism and what it means for human action. Do you accept that if determinism is true, then all human behavior, like the behavior of all other things in the physical universe, is causally determined by antecedent conditions in accordance with natural laws? And do you also accept that determinism rules out or excludes an agent’s ability to do otherwise in the unconditional sense?

Dennett: You’ve given an admirably clear summary of the standard definition of determinism, Gregg, and your two closing questions make exactly the right distinction, so let me answer them. Yes, I accept that “if determinism is true, then all human behavior, like the behavior of all other things in the physical universe, is causally determined by antecedent conditions in accordance with natural laws.” (And I have no quarrel with determinism; I see no reason at all to hope that human behavior is undetermined by those antecedent conditions.) But I do not accept your second claim, that determinism rules out or excludes an agent’s ability to do otherwise. You add: “in the unconditional sense” and this clause, as I’ve argued for decades, is a huge mistake – and I’ll say more about it in our debate. If what you are interested in is the ability of anything to do anything, you are never interested in what it does “in exactly the same circumstances.” I’ll try to make that clear as we go along. Let me add that I think your account of how “some compatibilists” interpret “could have done otherwise” in terms of certain conditionals about desires is accurate, but it shows that they, too, miss a key point. If you stick to these (quite standard but nevertheless misguided) readings of “could have done otherwise,” then not only is free will an illusion, as you would claim, but so is life; nothing is really alive. I consider that a reductio ad absurdum, but tastes may vary.

Caruso: Thank you, Dan, for your direct and candid answers. They provide us with a good starting point. It seems we generally agree on the implications of determinism for human action, you just deny that the ability to do otherwise should be understood in terms of the “ability to do otherwise in exactly the same circumstances.” (I assume you agree, though, that if determinism is true, we lack such an ability?) I’ll give you an opportunity to explain your position further, but you may be surprised to learn that my primary reasons for rejecting compatibilism do not depend on the ability to do otherwise or presuppose that it is a necessary condition for free will. In fact, my main arguments against compatibilism focus on the causal history of an agent’s actions and not on alternative possibilities. In our next exchange, I hope to spell out and discuss two of these arguments in detail, the manipulation argument and the hard luck argument. For the moment, though, I’ll provide a few rough reasons for thinking that determinism is a threat to free will and (desert-based) moral responsibility.

      First, consider the following argument for incompatibilism, which is independent of considerations regarding alternative possibilities and the ability to do otherwise. It comes from the American philosopher Peter van Inwagen, Professor of Philosophy at the University of Notre Dame. As he puts it:

      If determinism is true, then there is some state of the world in the distant past P that is connected by the laws of nature to any action A that one performs in the present. But since no one is responsible for the state of the world P in the distant past, and no one is responsible for the laws of nature that lead from P to A, it follows that no one is responsible for any action A that is performed in the present. (van Inwagen 1993: 182–183)

      This argument captures one of the incompatibilist intuitions I have regarding