I am an incompatibilist about moral responsibility. That is, I maintain that causal determinism and moral responsibility are logically incompatible. (Two propositions p, q are logically incompatible just in case they cannot both be true. Hence, logically incompatible propositions are logical contraries, not contradictories.) Here is an argument for incompatibilism:
P1. Causal determinism rules out alternative possibilities. For in a causally deterministic world W there is exactly one nomologically possible future at any time t given the laws of nature and the events that have transpired prior to t in W.
P2. Moral responsibility requires alternative possibilities (e.g., the ability to decide, choose, intend otherwise.)
Therefore
C. Causal determinism rules out (is incompatible with) moral responsibility.
Please note that this argument is not an argument for moral responsibility or for free will or for determinism. It is an argument for the logical incompatibility of (causal) determinism and moral responsibility. The argument is plainly valid, and P1 is true by definition of 'causal determinism.' So the soundness of the argument rides on P2. P2 is just a version of the principle of alternative possibilities (PAP). PAP is intuitively very appealing and has been endorsed by distinguished thinkers. So I hope it will be granted that the onus probandi rests on those who would contest it. The main threats to P2 are the Frankfurt counterexamples (FCEs).
A FCE is a possible set-up in which an agent who is assumed to be morally responsible for an action is nonetheless not able to act otherwise. He is not able to act otherwise because some feature of the set-up prevents him from doing so. Suppose the nefarious Mr. Black wants John Lennon dead because Black hates the dangerous utopian nonsense promoted by the song "Imagine." He gets wind of the fact that Mark David Chapman has been having murderous thoughts about Lennon in part because of the hypocrisy and religion-bashing stupidity of the song in question. So Black decides to use Chapman to carry out Black's design, but not by interfering in the actual course of events if he can help it. In the actual course of events, as we know, Chapman shoots and kills Lennon and acts alone from his own reasons and motives. It is for this reason that we consider Chapman morally responsible for his deed. Had Black intervened in the actual course of events, by messing with Chapman's brain or in any other way, then Chapman would not have been morally responsible for the deed. Black is not an actual intervener but a counterfactual intervener. He merely stands ready to intervene should Chapman suffer a failure of nerve and change his mind about shooting Lennon. The Frankfurtian point is that Chapman is responsible for shooting Lennon (because he acted on his own, not under duress, while in his right mind, etc.) but that he could not have done otherwise. Why not? Well, had Chapman given any prior sign of changing his mind and renouncing his intention to commit murder, Black would have intervened to ensure that he stay the course.
This scenario, the possibility of which we do not dispute , seems to show that (P2) above is false. It seems to show that moral responsibility does not entail the ability to act otherwise. FCEs thus appear to refute the above argument for incompatibilism.
The trouble with FCEs such as the foregoing is that they beg the question against the incompatibilist as Kane, Widerker, Goetz, and others have argued, and thus they do not give the incompatibilist a good reason for rejecting (P2) above. A solid refutation of PAP/P2 must be a counterexample that is neutral on such issues as determinism versus indeterminism and compatibilism versus incompatibilism. I should think that this neutrality is a condition of adequacy for any genuine counterexample to PAP/P2.
Either determinism is true or it is not. If determinism is true, then Chapman's decision, his maintenance of the decision up until the time of the shooting, and the shooting itself are all determined. In that case, the incompatibilist will insist that Chapman is not morally responsible even if he acts on his own. What divides compatibilist and incompatibilist is precisely whether or not causal determinism in the actual course of events excludes moral responsibility. Incompatibilists insist that it does. So if the truth of determinism is built into the FCE, then the example cannot be used to refute PAP/P2.
If, on the other hand, determinism is not true, then Black, even if he has the powers of a Laplacean demon, will not be able to discern in Chapman's mind any sure sign of a failure of resolve. In order to avoid begging the question against the incompatibilist, the Frankfurter cannot construe the prior sign of a change of mind that Black discerns as a causally sufficient condition of a change of mind. He must take the prior sign as merely raising the probability of a failure of resolve. If the prior sign is not causally sufficient, however, it is false that Chapman could not choose otherwise.
So the Frankfurter faces a dilemma. Either determinism is true or it is not. If true, then no alternative possibilities but also no moral responsibility. If not true, then moral responsibility but at the price of alternative possibilities.
What the Frankfurter needs is a possible set-up in which an agent is morally responsible but cannot act otherwise. Now if compatibilism is true, he can easily get what he wants -- and indeed without these far-fetched stories of psychosurgery, etc. But if incompatibilism is true, then the Frankfurt cases are impossible.
The upshot appears to be a stand-off. The Frankfurter has not refuted PAP/P2, and with it the above argument for incompatibilism. His attempted refutation begs the question against the incompatibilist. The incompatibilist in his turn is saddled with an unpalatable alternative: either reject moral responsibility as an illusion or reject determinism.
"What the Frankfurter needs is a possible set-up in which an agent is morally responsible but cannot act otherwise. Now if compatibilism is true, he can easily get what he wants -- and indeed without these far-fetched stories of psychosurgery, etc. But if incompatibilism is true, then the Frankfurt cases are impossible."
Michael Bergmann thinks otherwise (and, BTW, there are plenty of attempted FSC's that attempt to do what you claim needs to be done).
Here's Bergmann's paper:
http://web.ics.purdue.edu/~bergmann/frankfurt.htm
Posted by: Paul | Friday, June 05, 2009 at 02:18 PM
Paul,
No doubt they *attempt* to do it. Thanks very much for the Bergmann link. I'll check it out.
Posted by: Bill Vallicella | Friday, June 05, 2009 at 03:25 PM
Hi Bill,
Yes, you're right. Whether they are successful can be debated. However, a possible impression that your post gives is that no such attempts have been given. You make mention of what the Frankfurter needs. But apparently he has this (and I'm granting that indeterminist world objections even work).
You claim that "If indeterminism is true, then Frankfurt cases are impossible." But there are many (see Mele, Robb, etc) FSCs that make arguments while assuming indeterminism is true. It certainly doesn't seem impossible. At least it's not clear that it's impossible.
What's more interesting is that there are many libertarians (some who specialize on the metaphysics of free will) who deny that PAP is necessary for responsibility. They usually base this on FSCs. So, you have William Lane Craig, David P. Hunt, Linda Zagzebski, for example. You also have those like Bergmann, though not a specialist on the metaphysics of free will, a very fine philosopher.
So, it's not only determinists who find FSCs persuasive, there's indeterminists too. And, many of those indeterminists are secialists in the metaphysics of free will and aware of indeterminist world objections.
When you have people on both sides denying PAP, that might indicate that matters aren't so simple.
Posted by: Paul | Saturday, June 06, 2009 at 06:31 AM
Bill & several other Mavericks have regularly been paying obeisance to the PAP in this discussion of Frankfurt counterexamples. Let me say why I think this is a bad idea. (Yes, Bill, this is going to be the mother of pap smears!)
Many of you have heard of a Trojan Horse strategy in other contexts. One way to wreck a deal or contract is to quietly attach a condition that that cannot be executed or satisfied. So, if we say a agent shall be responsible for doing x only if conditions a and b and c obtain, and it turns out that c cannot obtain, or we can never ascertain or be sure that c obtains, we have rendered it impossible ever to hold someone responsible for doing x. Yes? The PAP is a Trojan Horse if we attach it to moral responsibility. Some defenders of the PAP sees this very clearly and want the conclusion that no one (themselves especially) is ever really responsible for the bad things they do. The PAP opens the barn wide open to the excuse that “I could not have done otherwise.”
“I cannot understand how I did those horrible things to that nice lady. Yes, I knew it was terribly wrong, but it felt so good, and I just could not stop myself. I guess I could not have done otherwise, and the philosophers tell me that means I’m not responsible, and so I'm not going to feel guilty about it or when it happens again.” This is the kind of disgusting and dangerous self-excusing that a PAP principle aloows. It is the kind of excuse psychopaths have been praying for —and if only it could only be converted into a legal defense! Then no truly evil monster could ever be held criminally responsible for his horrible deeds.
The criminal law wisely and steadfastly refuses to allow a Trojan Horse like the PAP into the criminal codes, despite the fact that some benighted philos recommend it. Think about what would happen if a PAP clause—“he could have done/chosen otherwise”-- became a material element of a criminal charge. Then the prosecution would have the burden of proving beyond a reasonable doubt that the perpetrator could have acted/chosen otherwise. First of all, this would be the biggest payday ever for criminal psychiatry, as defense and prosecution experts lined up and debated how deviant the brains of Gacy & Bund had become. There almost certainly is no conclusive answer at this point, and so we could not convict them.
The PAP is a Trojan Horse designed to wreck the concept of moral responsibility by requiring undecidable calls as to what someone could or could not have done, especially when there are genuine issues of psychopathology involved. Some people would very happy with this result, both for reasons of excusing their own misdeeds and for the sake of pressing their idealogical opposition to capital punishment, etc.
Just as the law refuses to allow any PAP condition to undermine criminal responsibilty, we must not allow a PAP condition to undermine our own personal sense of moral responsibility.
Posted by: Philoponus | Saturday, June 06, 2009 at 09:15 AM
On the technical issue of whether Frankfurt-style counterexamples like Bad Bob require/presuppose causal determinism, I don’t see the beginning of an argument that only in a causally deterministic world could Evil Genie monitor and influence Bad Bob’s actions. Keep in mind how strong a claim causal determinism makes about the universe. According to causal determinism, EVERY causal sequence is fully deterministic and governed by non-probabilistic laws. There is no way to deduce that from the fact that Evil Genie can reliably anticipate some last minute aversive reactions Bad Bob might have and response to them. (If you can produce this argument, I will drive to Gold canyon with a CASE of Fat Tire as your condign tribute.)
Conscious human decisions, even very simple ones like “soup or salad”, arise from complex neural processes that take time. No neural processes, no decisions or even consciousness. If the libertarian wants to pretend that “decisions” can spring into existence instantaneously without neural processes, he is riding lunicorns again and he is not talking about decisions that human beings make. Now we have no idea how deterministic all this neural processing is, and there are certainly no deterministic laws that anyone has a clue about. What we can be sure of is that high amygdala activation is NOT good for someone continuing to do what he doing. So if we are watching and can react quickly, in one or another way we can overwrite this aversive reaction and keep Bob feeling confident and happy about the embezzling he is doing. Show me how you can deduce causal determinism from these very modest predictive and interactive claims, and I will be very surprised.
Posted by: Philoponus | Saturday, June 06, 2009 at 09:52 AM
Paul,
You are right: it is not simple at all, and what I say above may contain mistakes. These posts, like most of my posts, are first-draftish and exploratory: I am working through the recent literature and trying to understand the free will problem in all its complexity. There is a tangle of issues: modal issues, questions about causation and determination, metaethical questions, the mind-body problem . . .
You say that there are libertarians who deny that PAP is necessary for MR. Would these be the 'source' as opposed to 'leeway' incompatibilists?
Posted by: Bill Vallicella | Saturday, June 06, 2009 at 03:17 PM
Phil,
Son of a gun, that is indeed the mother of all PAP smears! Your comment reminds of something Sartre says deep in the bowels of Being and Nothingness (if memory serves): determinism is a bottomless well of excuses.
But surely that a principle like PAP can be misused by some turkey doesn't mean much.
Posted by: Bill Vallicella | Saturday, June 06, 2009 at 03:24 PM
Phil,
As for you second point I refer you to Kane, Widerker, and Goetz.
Posted by: Bill Vallicella | Saturday, June 06, 2009 at 03:25 PM
Bill,
The libertarians who deny PAP are diverse. 'Source' incompatibilists are broken down further into 'wide' and 'narrow' source incompatibilists. Wide ones, like Timpe (for instance), claim PAP is necessary, the narrow ones, like Zagzebski (for instance) claim that it is not. But some libertarians seem to deny PAP for reasons other than 'source incompatibilism', issues like God's foreknowledge work for some, like Hunt (for instance). Others just find FSCs too doggone persuasive :-)
Posted by: Paul | Sunday, June 07, 2009 at 01:38 PM
Paul,
Thanks. You say that Timpe claims that PAP is necessary while Zagzebski claims that it is not. Do you mean that Timpe claims that PAP is a necessary truth while Zag claims that it is not a necessary truth, but a contingent truth?
Posted by: Bill Vallicella | Sunday, June 07, 2009 at 01:57 PM
Bill,
Sorry for the confusion. Narrow source incompatibilists (like Zagzebski) do not think PAP necessary to ground ascriptions of moral responsibility. An agent can be responsible without having any alternatives. Wide source incompatibilists (Kane, Pereboom, etc) claim that sourcehood is ultimate but that a weak form of PAP is *entailed* by a right view of sourcehood. Timpe thinks narrow collapses into wide--i.e., we can only be the ultimate source onlyif we have alternatives. So, in that sense, for wide incompatiblists, since ultimate sourcehood is what is most relevant for ascriptions of responsibility, and since we can only be the ultimate source only if we have alternatives, PAP is necessary that way.
Posted by: Paul | Monday, June 08, 2009 at 07:15 AM