Many years ago at the Garden of Forking Paths, Neil Levy would joke (at least, I think he was joking) that he was an compatibilist about moral responsibility on Mondays and Thursdays and a skeptic the rest of the week. Lately, I’ve wondered whether this is a reasonable position to adopt. Can it be justified to shift back and forth on the question of whether agents can be morally responsible for their behavior?
Here’s why it might be:
1. We decide what MR theory to accept by employing something like wide reflective equilibrium. Fischer and Ravizza are explicit about their use of WRE, but it seems that all theories are implicitly committed to this methodology for justification.
2. We have conflicting intuitions about the
all-important TNR principle. On a
general level, it seems inappropriate to blame agents for actions caused by
factors that trace back beyond the agent’s control. And generalization or manipulation strategies
like the 4-case or zygote arguments provide more intuitive support for the
principle.. On the other hand, it also seems intuitively inappropriate
not to blame offenders who meet
compatibilist conditions. This is
especially true in concrete cases, real life cases, and especially cases where
we know the offended parties.
3. Neither of these intuitions can be explained away. Compatibilist debunking strategies, like the one that Eddy and colleagues have developed, have the potential to explain away the results of certain studies suggesting that we have incompatibilist intuitions (e.g. Nichols and Knobe). But they do not apply to my intuitions in something like the 4-case argument—Pereboom makes it very clear that the agent’s intentions, desires, deliberative processes are not being bypassed. Skeptical debunking strategies like my own in “The Illusion of Freedom Evolves” and Greene and Cohen (2004) might explain away the retributive intuitions of people who embrace a dualist or agent-causal view of agency. But they won’t apply to those concrete, real life cases where (a) we know that the agents are not agent-causes, and (b) we know that our retributive intuitions are in some sense the result of our evolutionary and cultural history, but (c) we still find it intuitively inappropriate to withhold assignments of moral responsibility.
4. ‘All-things-considered,’ it’s a really close call.
Mike Patterson, a grad student here at Houston, recently described my position like this: “He’s a skeptic except if someone hurts his daughter—then he’s a compatibilist.” There’s some truth to this. And as I’ve talked about at the Garden, I only endorse my retributive intuitions upon reflection if the offender meets the normal compatibilist conditions. And of course, I’m aware that whenever an offense is committed, somebody’s daughter/son/wife/husband/partner/friend is harmed—it would be bizarre if someone could be MR for harming my relatives but not the relatives of others. These considerations sway me towards compatibilism. At the same time, it’s only in rare cases that denying moral responsibility seems on balance inappropriate. For most cases of wrongdoing (to friends, relatives, or strangers), moral luck style arguments convince me that the agent cannot deserve blame or praise whether or not they meet compatibilist conditions. So when I ask myself what position makes the most sense, all things considered, to embrace, it's a very close call.
5. The balance of intuitions is a delicate one.
There have been some suggestive studies linking our intuitions about free will and MR with character traits like extroversion. These studies test for links on an interpersonal level, but I imagine that they exist on an intrapersonal level as well. Changes in our environment or life circumstances (like having a child, getting married, being the victim of a crime, losing someone you love, or just getting older) might tilt the balance of MR intuitions one way or another. The closer the “all-things-considered” call is, the smaller the shift in circumstances or temperament would have to be to result in a legitimate change of position. Even our day-to-day mood swings might be sufficient.
So there it is for anyone interested—semi-weekly compatibilism. I’d defend it myself if I weren’t already committed to another crazy theory.