Many of you will be interested in a paper by Patrick Todd that was just published online with Philosophical Studies, "A New Approach to Manipulation Arguments". The paper is short and sweet, and makes an intriguing point: proponents of manipulation arguments against compatibilism need only maintain that the manipulated agent's blameworthiness is mitigated, not necessarily eliminated.
It's a quick read, so you should read what Todd has to say for yourselves, but here's the basic idea. Although many compatibilists are attracted to a hard-line response to Pereboom's 4-case argument -- namely, that Plum is blameworthy even when manipulated -- this response is actually even harder than has previously been noted. Why? Simply because the compatibilist must say not only that Plum is still blameworthy when manipulated, but that the manipulation in no way even diminishes Plum's blameworthiness. The compatibilist must say this, Todd argues, because to allow mitigation in response to manipulation is to take the first fragile steps down a slippery slope that ends with compatibilism's demise. To allow mitigation is to commit oneself to finding a distinction between manipulation and determinism (if one is a compatibilist), but what about the manipulation setup could possibly mitigate blameworthiness which wouldn't also mitigate blameworthiness in a simple deterministic world? The bottom line, Todd argues, is that the compatibilist has to be even more hard-core than we used to think. And she was already pretty darn hard-core.
A couple of my own thoughts are below the fold.
First, I wonder whether any compatibilists would be happy to say that the truth of determinism does mitigate blameworthiness. Here are two ways the compatibilist might be able to make that bullet a bit easier to bite.
a) To accept the slogan "determinism mitigates blameworthiness" is just to accept that agents in deterministic worlds are slightly less blameworthy than agents in suitably indeterministic worlds. This could mean either that we should blame the agents in our world less if we were to find out that we live in a deterministic world or simply that we should blame agents more if we were to find out that we live in a suitably indeterministic world. That is, mitigation is always relative to some standard, and as far as I can tell, you could be a consistent compatibilist and believe that most of the blame we dole out in the actual world is perfectly justified -- it's just that we would be justified in doling out even more if it turned out that people were "indeterministic initiators" (Mele 2006).
b) To accept the slogan "determinism mitigates blameworthiness" isn't to accept the slogan "determinism mitigates moral responsibility". That is, you could be a compatibilist who thinks that acting from one's own moderately reasons-responsive mechanism (for example) is sufficient to get you into the MR game (which is an all-or-nothing game), but it takes more than that to be blameworthy (or, what might come to the same thing, to be justifiably held responsible).
Second, I wonder whether our intuitions about mitigated blameworthiness in the case of Plum (if indeed we do have them) track facts about us more than facts about Plum. That is, it's one thing to say, as Todd does, that Plum's blameworthiness is clearly mitigated when he's manipulated, but it's quite another to say that the mitigation intuitions are due to Plum's not satisfying some control condition on blameworthiness. Perhaps finding out about Plum's circumstances makes us seriously entertain, for the first time, the possibility that we are all just victims of similar sorts of manipulation, which makes us feel less entitled to feel so resentful toward Plum. (Considering the Matrix hypothesis might have a similar effect on us with respect to attributions of knowledge: not that we think people know less than they used to -- the Matrix hypothesis is outrageous, after all -- just that we feel less entitled to assert that they do. Maybe that's not the best analogy, but I'm sure there's something analogous in epistemology.)