Freedom and Belief by Galen Strawson (2010)
An Essay on Free Will by Peter van Inwagen (1983)
Free Will, Agency, and Meaning in Life by Derk Pereboom (2014)
In my previous post, I questioned the conception of free will described by the conditional formulation of the freedom to do otherwise. I was then briefly sympathetic to the categorical formulation. Here, I’ll discuss a well-known class of examples, typically called Frankfurt-style cases, intended to undermine the freedom to-do-otherwise conception as a whole (often in order to motivate a source conception). A typical Frankfurt-style case is something like the following:
Imagine Alex is trying to decide who to vote for. He will likely vote for the Donkey Party candidate, Doug (Alex has always voted Donkey in the past). However, if he happens to consider his relatives serving in the armed forces just prior to voting, he will have a change of heart and vote for the Elephant Party candidate, Elaine, instead. Unbeknownst to him, neuroscientist and Donkey Party operative Brianne has implanted a very special device in Alex’s brain. The device constantly monitors Alex’s brain activity for thoughts about members of the armed forces (and relays this information instantly to Brianne) and, upon activation, modifies Alex’s brain activity to cause him to vote for Doug. Brianne’s plan is to simply wait for Alex to go to the polls. If he happens to think about the military prior to voting, she’ll activate the device, causing Alex to vote for Doug instead of Elaine. If Alex shows no signs of considering the armed forces, she’ll leave the device inactive, allowing Alex to vote for Doug on his own. As it happens, Alex does not consider the armed forces prior to voting, and votes for Doug without the interference of Brianne’s device.
The argument surrounding a Frankfurt-style case is this. Whether he thinks about his relatives in the armed forces or not, Alex will end up voting for Doug. There is no possible world in which he votes for Elaine. Therefore, Alex does not have the ability to do otherwise. However, the way events actually unfolded, Alex voted for Doug on his own, without Brianne and her device interfering. Intuitively, then, Alex’s decision to vote for Doug was a free one, one for which he is responsible. But if Alex exhibited freedom in a case in which he could not have acted otherwise, freedom cannot consist simply in the ability to do otherwise. The freedom to do otherwise, therefore, does not adequately capture what we mean by free will.
The classic criticism of Frankfurt-style cases comes in the form of a two-horned dilemma. Consider the causal connection between Alex’s failing to think of his relatives in the armed forces and Alex’s voting for Doug. This connection is either deterministic or indeterministic. If the connection is deterministic, then proponents of Frankfurt-style cases cannot expect to convince incompatibilists. If the fact that Alex failed to think of the armed forces deterministically led to his voting for Doug, incompatibilists will deny that Alex made a free choice. The intuition that Alex is responsible for voting for Doug, that he was free in his choice, is crucial to the success of the argument, and so it falls flat. If, on the other hand, the connection is indeterministic, then it is possible that Alex fails to consider his relatives in the armed forces and yet goes on to vote for Elaine. In this case, opponents of Frankfurt-style cases can claim that Alex did in fact have the ability to do otherwise. Either way, according to this dilemma, free will as freedom to do otherwise remains intact.
As you might guess, the arguments surrounding Frankfurt-style cases continue today. There are attempted rebuttals of the two-horned dilemma, then responses to those rebuttals, and on and on. I won’t explore these arguments in more depth here (I’ve started to sense that the success or failure of Frankfurt-style cases will ultimately be irrelevant to where my position stands). Instead, I want to make some methodological points about the argument Frankfurt-style cases attempt to make. Notice that the intuition appealed to, that Alex is responsible for voting for Doug, is an intuition about moral responsibility (praise and blame). I’m fairly skeptical of relying on such normative intuitions, especially when trying to understand something as fundamental as human freedom. But fully investigating the relationship between free will and moral responsibility, and indeed the nature of moral responsibility alone, would take the present overview of the free will debate far afield. Here I simply want to note the reliance on intuitions about moral responsibility, a reliance that would be undermined by a moral skepticism.
For Frankfurt-style cases to apply directly to free will, one would have to identify free will with the kind of control required for moral responsibility. But notice – we seem to have come full circle, logically. I introduced Frankfurt-style cases as part of a broader conceptual inquiry into what free will is, arguments intended to undermine a to-do-otherwise conception of free will in favor of a source conception. But in order to do so, it seems Frankfurt-style cases rely on a conception of free will as the kind of control required for moral responsibility. The success of such cases with respect to the free will debate rests at least in part on the connection between free will and moral responsibility. Perhaps this connection can be assumed (as suggested in my opening free will post), or perhaps not. I won’t argue either way here. Just notice that an implicit conception of free will, the freedom required for moral responsibility, is lurking in the Frankfurt-style case argument.