Many believe that we are not morally responsible for what we cannot help doing. Call this 'the Traditional View of Responsible Agency.' Some forty years ago, Harry Frankfurt (1969) challenged this view, thereby initiating a new stage of the free-will debate. In contrast to the previous stage, in which debate centered on how best to accommodate the Traditional View, contemporary theorists have focused on whether this view should be accepted at all. If the link between moral responsibility and avoidability is severed, an important threat to compatibilism is neutralized.2 [End Page 505]
In the wake of Frankfurt's challenge, a tremendous literature has sprung up, with many ingenious responses matched by equally thoughtful extensions of Frankfurt's original argument. Quite recently, an altogether new line of response has been proposed. This new approach, versions of which have been advanced by Maria Alvarez (2009) and Helen Steward (2009), attempts to support the Traditional View indirectly, by appealing to the conditions for action, rather than to the conditions for moral responsibility per se. The leading idea is that it's essential to something's status as an action that the agent be able to refrain from performing it. Call the view that actions (conceptually) must be avoidable 'the Avoidability Thesis about Action.' If this thesis is correct, it's never the case that someone is morally responsible for an unavoidable action, since there is nothing to be morally responsible for when a piece of behavior is unavoidable. Insofar as Frankfurt's challenge depends on showing that someone can be morally responsible for the very same action (= act-token) she is unable to avoid, the Avoidability Thesis, if correct, seems to bring this challenge up short.
Indeed, by contesting the idea that actions can be unavoidable, this response takes aim at a supposition that all 'Frankfurt cases' (the name for Frankfurt-inspired counterexamples to the Traditional View) seem to share. In this respect, it goes beyond many previous objections, which are directed against particular developments of the counterexample strategy. If successful, this response may well reveal a fundamental difficulty for Frankfurt's strategy, one that cannot be resolved through further modifications of the cases; thus defenders of this strategy cannot safely ignore this response.
Here I argue that this response won't help incompatibilist proponents of the Traditional View. This is because these philosophers seek to defend the Traditional View as a premise of the Traditional Argument for Incompatibilism. According to the Traditional Argument, we are morally responsible only for what we are able to avoid (the Traditional View), and we are unable to avoid anything we are deterministically caused to do. If appealing to the Avoidability Thesis preserves the Traditional View at the expense of the Traditional Argument, the result will be a hollow victory for incompatibilists. I shall argue that this appeal yields precisely such a hollow victory. Since incompatibilists who remain loyal to the Traditional Argument have a particularly clear stake in meeting Frankfurt's challenge, it is significant that the new response isn't available to them.
But why think that incompatibilists can't appeal to the Avoidability Thesis in a combined defense of the Traditional View and the Traditional Argument? After all, Frankfurt's challenge purports to undermine the Traditional Argument by disproving the Traditional View, a premise of that argument. It would seem, then, that countering Frankfurt's [End Page 506] challenge to the Traditional View should also constitute a defense of the Traditional Argument. As I hope to show, however, things are not so simple. Even if we grant for argument's sake that appealing to the Avoidability Thesis answers Frankfurt's challenge to the Traditional View, this response cannot be parlayed into a promising defense of the Traditional Argument. To see this, we may begin by putting this question to incompatibilists who endorse the Avoidability Thesis: 'Are actions that originate deterministically ipso facto unavoidable?' If they answer 'yes,' they find themselves on the first horn of a dilemma. For if every such behavior is unavoidable, and if no action is unavoidable, it follows that nobody...