Blog Coordinator

« Mind Over Masters: The Question of Free Will | Main | You Are Not an Event (Are You?) »



Feed You can follow this conversation by subscribing to the comment feed for this post.

I like Gregg's paper a lot. It builds on Derk's quarantine model, but by adding the public health material (a very nice resource) it moves the emphasis away from quarantine of individuals to changing social conditions to prevent the development of criminal behavior, and thus has a less individualistic focus. Of course Derk acknowledges the importance of such efforts, but by placing the question in a larger public health context Gregg gives it a more systematic and basic focus. I'm still not sure I like talking about "forward looking moral responsibility" (would rather not talk about moral responsibility at all), and I have some concerns about any model that relates/compares/analogizes criminal behavior to problems of "poor health"; nonetheless, this seems to me a very promising and rigorous effort to deal with a vexing problem for moral responsibility skeptics. Am eager to know what Derk and Thomas and Tom Clark and other "Flickers" (Flickerers?) think of the approach.

Thanks Bruce! I too am eager to know what Derk, Thomas, Tom Clark, and others think!

Hi Gregg: Interesting paper! I'm going to be totally annoying and ask a meta-question about what you think we should take away from investigations like this (sorry!:).

The way you present this paper, it's a kind of big conditional, viz. "*if* free will skepticism is true then this would be the right kind of approach to criminal punishment."

Okay, but now what you think--as I and others do--that it is an open question whether free will skepticism is true? As you know, I have defended a new version of libertarianism that I argue is consistent with all known physics, and there are others who have argued in a similar vein that we have *no* good evidence against libertarianism (see e.g. Mark Balaguer's )

Now, I'm not expecting you or any other free will skeptic to agree with us (only science will confirm or disconfirm my theory's predictions). What I do think is reasonable, however, is for free will skeptics to ascribe some non-zero probability to libertarianism being true. Of course, you might think libertarian freedom is incoherent, or magical, or whatever, but there are really smart people who argue the contrary: that there's nothing incoherent about the notion of a cause-sui, etc.

Anyway, here's my question. If you accept that there's at least some epistemic uncertainty about which theory--skepticism or libertarianism--is true, then don't we have to answer the following moral question before we implement anything: "How should we weigh the moral costs and benefits of implementing your scheme given that uncertainty about the truth?"

Here's why I think this matters. Suppose you think, as I do, that if we have libertarian free will, it is of immense moral importance (indeed, categorical moral importance) to hold people retributively responsible for their actions because they *did* freely choose them. Suppose you then think, as I do, that it's not so bad to hold people retributively responsible if free will skepticism is so true, since we are naturally inclined (thanks so evolution, let's say) to hold ourselves and others retributively responsible (we do it naturally, derive satisfaction from it, and so on).

If this is right--if it would be a moral disaster not to hold people retributively responsible if we have libertarian free will, but not a disaster to (incorrectly) do so if we do have free will--then couldn't someone accept your argument here, think free will skepticism is probably true, but *still* think we should use retributive punishment (to guard against the greater moral disaster that might obtain if, contrary to one's skepticism, it turned out that we are libertarian-free)?

I realize a skeptic could run the argument the other way, arguing that it's not a moral disaster if libertarianism is true and we fail to correctly hold people retributively responsible, but *would* be a disaster if skepticism is true and we do. But this just brings me to my final point: that this what needs to be settled. We cannot simply read off that we should implement your quarantine model from the mere expectation that free will skepticism is likely to be true. We need to settle another issue before we can (legitimately) derive any conclusions about implementation: namely, which of the two would be a greater moral disaster--holding people retributively responsible in a world in which we shouldn't (a world with no free will), or failing to hold people retributively responsible in a world in which we should (a libertarian world).

I ask this ridiculously long-winded question because I am entirely on board with your claim that this is the punishment model that would be best if we *knew* free will skepticism is true--but I am far less sure that it is a good idea to even countenance implementing the model given that libertarianism might be true (if it *is*, after all, we would be doing everyone an immense moral disfavor by merely quarantining them).

Marcus, thanks for your question! This paper, as you know, doesn't attempt to argue for free will skepticism. It tries instead to address a frequently voiced criticism of free will skepticism--that is it unable to adequately deal with criminal behavior.

That said, on your meta-question, I think we clearly have different background assumptions. I agree with Richard Double that given the lack of epistemic justification *for* libertarian free will, it would be "hard-hearted" to hold people morally responsible in the basic desert sense. (For those who haven't read Double's article, it's really worth a read: I don't think the mere possibility of libertarian free will is enough to justify retributivism or just deserts.

I also think we disagree about which approach to criminal justice would be "a greater moral disaster." I personally think retributive punishment is far more disastrous than my alternative, hence the epistemic justification for libertarianism would need to be rather high to justify such treatment.

Of course, the public health-quarantine model I am defending in this paper doesn't depend on anything I just said. And I'm glad to hear that you think it makes a good non-retributive alternative for those who are seek such an alternative. While that might not be you, I'm hoping the account appeals to my fellow skeptics ;)

If that previous link to Double's paper didn't work, try this:

As you know from earlier this year I think that the kind of Fischerean (is that a word?) dialectical stalemates we find in the literature, along with Levyian (is that a word?) doubts about the meta-issues of action theory, should point to another out: a pragmatic resolution of the matters of responsibility and punishment. And yours here I think meets a lot if not all reasonable criteria for being pragmatic. I approve.

Now if there were only some thought-experimental way to see how we should treat criminals who were indistinguishable behaviorally from libertarian transgressors in both thought and action, but were empirically and demonstrably not indeterministically motivated. . .

maybe free will zombies. . .

I do need to get that paper in print sometime.

"which of the two would be a greater moral disaster--holding people retributively responsible in a world in which we shouldn't (a world with no free will), or failing to hold people retributively responsible in a world in which we should (a libertarian world)."

I'd say it's about a wash, although I would probably lean more toward the latter than the former. Our retributive sentiments should be overridden, or in any case constrained by by consequentialist considerations. I am of the mind that not much really rides on the truth of determinism in terms of our practices of rewards and punishment.

The greatest moral disaster would be holding people responsible (or not) without any respect to consequence. Once consequence is factored in, the fact that we're justifying our practices in part by retributivist considerations seems practically irrelevant.

"Your [account] I think meets a lot if not all reasonable criteria for being pragmatic. I approve." I'll take what I can get Alan! ;) That's all I was hoping for.

Izzy, I don't thinks it's a wash. Blame causes harm. Harming someone requires justification. If your justification for harming someone is that basic desert moral responsibility exists you need good epistemic reasons for thinking that. Again, though, the goal of this paper was more narrow: it was to argue that free will skeptics can, in fact, deal successfully with criminal behavior.

Hi Greg,

We don't disagree. I am on board with your general project.

I was responding to the dilemma posited by Marcus. It's true that blame causes harm, but failing to blame and withholding praise also causes harm. He suggested that (perhaps) it would be a moral disaster if we didn't blame people who have free will, but not so much a moral disaster if we blamed people who don't have libertarian free will. I think both are morally unacceptable approaches. In other words, I think the quarantine model ought to be implemented quite apart from what turns on the question of whether or not free will skepticism is true. I don't think a prior question about free will skepticism needs to be settled since I largely think basic desert should be irrelevant to our practices. Any kind of strict basic desert account of -holding- people responsible strikes me as a moral disaster.

Gregg: Thanks for your reply! I expect we'll just have to agree to disagree.

For example, when you say, in response to Iggy's reply to my comment (suggesting the two moral risks are a wash), "I don't thinks it's a wash. Blame causes harm. Harming someone requires justification", I respond: *not* blaming people in a world in which we have libertarian free will is a harm. It is a harm not only to those who made the wrong choices they made (since they are not being held accountable for wrongs they *deserve* to be held accountable). More deeply, it harms those *they* harmed through their wrongful (including criminal) behavior, by not holding the person who harmed them retributively accountable.

In other words, I don't think you can so quickly argue, "Merely quarantining is a better moral risk because it doesn't run the risk of harming people, which retributivism does." For, people with retributivists such as myself think that failing to hold people retributively responsible in a libertarian world is a harm: (1) a harm to the person who isn't properly held retributively responsible, and more importantly, (2) a harm to their victims (by failing to hold the person who harmed them appropriately accountable).

Izzy, I got you now! Glad to know we are in agreement. And I agree with you that there are good independent reasons for rejecting basic desert moral responsibility.

Marcus, I never intended (at least not in this paper) to having anything riding on the moral hazards of retributivism. I was just trying to answer your meta-question by pointing out that you need to contend with Double's argument. You write: "retributivists such as myself think that failing to hold people retributively responsible in a libertarian world is a harm." Okay, but how certain are you that we actually do live in a libertarian world? Are you 90% certain? 50%? 20%? If we have low epistemic justification for believing libertarianism is true, is blaming and retributively punishing people justified? I would think not--but maybe I am just soft-hearted at Double would say ;)

I'm still curious to hear what people think of my actually proposal. I think it goes a long way in responding to a common practical objection to free will skepticism.

"failing to hold people retributively responsible in a libertarian world is a harm: (1) a harm to the person who isn't properly held retributively responsible, and more importantly, (2) a harm to their victims (by failing to hold the person who harmed them appropriately accountable)."

Wrongdoers would would still be held responsible in terms of consequentialism. It's hard for me to see that it would be especially injurious for us to give up basic desert if we're still responsible on a forward-looking consequentlist account. The harms of retributive blame seem vastly more costly than the harms of quarantining, especially since the latter is modeled on the maximization of welfare.

Hi Gregg: Thanks again for your reply. I'll bow out after this, as I don't want to derail the discussion further--but I suppose I should address your question (about how certain I am we live in a libertarian world).

Personally, I think that everyone--including you!--should ascribe it at least a 50% probability. Here's why.

The amount of confidence one should put in a hypothesis should depend on the extent to which the hypothesis in question has actually been tested. At the present point in time (recorded history up until now), science has only studied fundamental physics *outside* of brains (in particle colliders and the like). This means that all of our evidence, both for and against libertarianism, has been gathered in precisely those parts of the world in which--prior to engaging in inquiry--one would not expect to find it (namely, in conscious brains). As such, because science has never actually studied the one part of the world where you would otherwise expect libertarian free will to exist, you should be agnostic: take it to be a 50/50 proposition, and wait for the actual evidence to roll in.

Now, you might think this is a bad argument--that the progress of physics (everything it has studied outside of human brains) gives us reasons to think that physics probably works similarly within human brains. But actually, I want say, careful epistemology and scientific history are on my side. Allow me to give some examples where people have made moves similar to the free will skeptic's and been wrong precisely because they extrapolated further than the actual hypotheses the sciences of their time had ever tested.

First, prior to the discovery of quantum mechanics, what should someone's level of credence have been in the proposition that the world is deterministic? We all know that prior to QM's discovery, just about everyone believed the Newtonian/Laplacean dogma that (obviously!) the world must be deterministic. Why did people believe this dogma (and it was a dogma, for reasons I will now explain)? The answer is that the science up to their point in time had *only* studied macro-sized objects, and since those macro-sized objects (appeared) to obey deterministic laws, they extrapolated a very high level of credence that of course the microphysical world is probably deterministic too. But this was a horrific epistemic mistake. The science of their time had *never* carefully tested the microphysical world they were making assumptions about on the basis of the behavior of macro-sized phenomena. They made an epistemically invalid assumption--that the microworld is probably like the macroworld--and they paid for it (voila, once we actually studied the microworld, it turned out to be indeterminisic--exactly what no one expected!).

Anyway, I want to say that they made a mistake all along. No one had *any* reason to believe one way or the other--in favor of microphysical determinism or indeterminism--because the microworld had never actually been studied. Prior to actually studying the microphysical world, the epistemically correct thing to do was to withhold belief--ascribing a 50/50 probability to each hypothesis--and then wait for the actual evidence to roll in.

This was not the only time people made this kind of error. Prior to the discovery of General Relativity, what should someone's credence have been in the "luminiferous aether"? No doubt many scientists *believed* it was likely to exist...because they were under another dogma that seemed reasonable to them at the time: the dogma that space and time must (obviously!) be absolute. But, of course, in retrospect their belief that the aether was likely now looks foolish. They had no reason to believe one way or the other because, up to that point in time, science had never *studied* that phenomenon either.

In short, I want to say (A) for purely epistemological reasons, and (B) for historical reasons, the epistemically correct thing to do, right now, is have a 50/50 credence on libertarianism.

It's epistemically correct because science has never measured the one area of the world in which libertarian free will is thought to exist (conscious brains)--and, I want to say, history bears this out. All too often, the things people are confident "must" be true, but which their science hasn't yet studied, turn out to be false.


(And I hope I'm not interloping Gregg.)

I must strenuously protest your point about the credence for belief in libertarianism.

First it's unfair to analyze credences for false past scientific beliefs after the fact. Of course any false belief updated by new data and theory will look credentially bad by comparison. That's just a feature of learning from mistakes.

But second, to say that present circumstances should be evaluated from that perspective loses all proper context of what constitutes ongoing epistemic evaluation. When, prior to Michelson-Morley, lots of experiments and theory favored belief in the ether, it was quite reasonable to believe in it. So reasonable that Lorentz, despite being a good friend of Einstein and really getting his physics, never could let go of it. We must remember that credences are an overall function of what we take to be proper contexts and sets of evidence.

Third, science's mirror of nature on the brain has returned a lot of evidence that it functions by bio- and electro-chemical means that are reflected in lots of non-human biological processes that connect all life by evolution. While genomic differences show that roses and human minds function in very different ways, there is no good evidence that they function in ways that specifically depend on subatomic indeterminism--in fact there are many very good reasons to believe that human minds are closer to a kind of computer wetware than reflective of electrons-uncertainty-measured for position and momentum. Not to mention that the meta-explanation of quanta phenomona themselves is not uncontested, and thus adds to epistemic uncertainty applying indeterminism as a credence explanation of how minds work.

Given that evolution connects us to the rest of nature, and there are no good examples of life working by anything other that bio-electro-processes that are compatible with causal and nomic descriptions (even ones that employ some sens of chance as a factor, as with lucky adaptations), a kind of rational, indeterministic libertarian choice is far, far less than 50% an apt reasonable belief about how the human mind functions.

Hi Gregg

That's an interesting paper, and a lot to consider. Just a couple of brief comments:

With respect to Marcus's challenge, it seems to me it doesn't need to be limited to skepticism vs. libertarianism. If libertarian free will does not exist but free will does (and is compatibilist), and so do moral responsibility, desert, etc. (as I believe), then the same questions may be raised.
On that note, while I personally (and for reasons not related to physics) don't agree with Marcus's assessment that we should all ascribe at least 50% to the hypothesis that we live in a libertarian world, I think we should ascribe more than that to the hypothesis that we live in a free will and desert world, i.e.., a world in which ordinary assessments of free will (e.g., "she did it of her own free will"), blame and desert, are at least often true.

On a different note, I'm not sure I understand your characterization of desert moral responsibility, but as I see it, it is not required that the person who deserves to be punished knew (or understood) that his actions were immoral.
For example, it may be that a person targets and kills innocent people believing that his behavior is morally praiseworthy or obligatory, perhaps in the context of his improper belief in some ideology (political, religious, both, etc.). Maybe he bombs them, or maybe he stones them to death for adultery, etc., depending on the case. I'm not sure he knows or understands that his behavior is immoral - though I think he should know -, but I think he deserves punishment regardless.

Marcus, you appear to reason as follows: we should assign a 50/50 credence to libertarianism because we should assign a 50/50 credence to the proposition that the brain is indeterministic. But that only goes through if we should be 100% confident of the conditional: if the brain is indeterministic, libertarianism is true. Most opponents of libertarianism (I think all of the bette known opponents) do not base their opposition on the claim that the brain probably doesn't work like that (here I can speak for myself, of course). Are *you* really so confident of the conditional? Do you really think that everyone should be so confident of the conditional? Again, speaking for myself I am very confident (somewhere short of 100% confident, but very confident) that libertarianism is false, though I am not confident that the brain is deterministic.


With regard to the second objection to deterrence theories - namely, that sometimes they would justify punishing the innocent -, I don't see why punishing the innocent would be morally worse than punishing the guilty given the same expected future consequences, assuming that backward-looking (or even present-looking) justifications for punishment are rejected. In particular, it seems to me that if only forward-looking accounts of punishment could be acceptable, whether a person behaved immorally in the past is not relevant to whether other people should punish her. I may be mistaken, but my impression is that the second objection in question implicitly brings back backward-looking considerations for punishment, even if in a limited manner.

As for the quarantine model, I'd like to raise the issue of people who committed serious crimes but are no longer dangerous.

For example, we may consider the cases of former dictators or enforcers in a dictatorial regime who engaged in acts of mass murder, mass torture, etc., in the past, but no longer pose a threat. Among them, one might include many former military and police officers in some dictatorial regimes in Latin America, or old former Nazi war criminals, etc. Under the quarantine model, it seems to me they would not be punished.
Similarly, it seems to me that, say, men who engaged in many instances of child molestation in the past but are now old and haven't done anything like that for, say, over two decades, would not be detained, as they seem to pose no significant threat anymore - or at least, that would be the conclusion if some psychological or other medical tests indicate they're almost certain not to engage in any such behavior again (in some jurisdictions, people who engaged in such crimes are not punished even on a partially retributivist system, due to statutes of limitations).
There are also cases that don't involve crimes committed long ago, and in which a person seriously harms or kills innocent people, but the harm is neither intended nor expected by the criminal. For example, someone is celebrating New Year, and decides to fire a few shots in the air with his pistol, or assault rifle, or some other firearm. As it happens, one of the bullets lands on the head of a person out of sight, killing her instantly. When the shooter finds out, he feels terrible for what he did, and there is clearly no realistic chance - court psychologists can tell - that he'll fire shots in the air again.
It seems to me that applying the quarantine model, he would get away with no punishment, at least after the psychological evaluations are done - at most, he would be quarantined for a very brief period so that the psychologists can assess the matter.
Moreover, if another person does the same (i.e., fires a few shots in the air, in similar circumstances), but the bullets land harmlessly, he's much more likely to do that again. It seems to me he would incur more punishment under the quarantine model, all other things equal, since he is a greater threat, even though his action was equally but not more immoral, and caused no harm.

Based on those examples and similar ones, one might raise the following issues:

1. Even if the law rejects desert, it seems very unlikely that the population at large will, as human retributive feelings will not go away. The victims and/or their families - and probably others - will usually demand that the guilty suffer for what they did. But their demands will not be met, and many of them will likely suffer psychologically as a result.
2. If the demands in question are not met, some - perhaps many - people may be willing to do violence. That's a problem for social peace.
3. The system might be inherently unstable. In a democratic system, candidates may campaign on a platform promising to bring desert back, to punish those who deserve it, etc. They may well be successful.

But perhaps, potential outcomes like 1. or 2 might be considered a threat to society that justifies quarantining someone?
A worry is that taking an approach like that might open the door for punishment of people who aren't guilty (though as I said above, I'm not sure why that would be a problem if only forward-looking considerations for punishment are accepted), because many people want to punish behaviors they mistakenly believe to be immoral.
Another worry is that in a way, accepting the threat of 1. or 2. as justifications for quarantine indirectly brings back desert - since someone might be punished not because she deserves it, but because some or many other people believe that she deserves it, and they might suffer psychologically and/or do violence if they see her get away with what she did.

Regarding Marcus' point, I agree with Neil 100%!

Angra, thank for your comments! They are very helpful. I will only respond to your points about my public health-quarantine model since those are most relevant to this post. I acknowledge that once we give up the concept of just deserts, punishing people who commit crimes but are no longer dangerous becomes a problem. But is is only a problem if you believe people justly deserve to be punished in the first place. Consider, for instance, Katrina Sifferd's Bert example from last month. If we conclude that Bert made an honest mistake (it was a one-off oversight) and that his children are not at continued risk of harm in the future, I think it would be inappropriate to punish Bert. If, on the other hand, we think Bert will provide appropriate care for his children in the future but we have some concerns about his forgetfulness, perhaps we could recommend oversight or monitoring of his visitation for a short period. You can imagine escalating sanctions from here based on different assessments of risk of harm in the future. I readily acknowledge, however, that the ex-Nazi war criminal is a different story--our retributive intuitions are extremely high here (and for good reason). To try to manage those intuitions, though, let me flip the case and consider a U.S. Private who commits a war crime in Vietnam. Here too I think we have strong retributive intuitions. But despite the retributive impulse, perhaps the most appropriate course of action would be to consider the circumstances of the crime, take active steps to prevent such situations from occurring in the future (my focus on prevention), and provide the private with needed psychological counseling and treatment. I understand that many will find this unacceptable because they strongly believe the private deserves to be punished, but my intuitions are a bit different. The Stanford Prison Guard experiment, the milgram experiment, etc. reveal that we are not immune to the influence of authority. I think many of us are capable of such crimes and it is often luck that decides whether we are placed in exactly those circumstances with exactly those psychological forces acting on us. Of course we need to deal with the individual in cases like this (and I think my model allows for that), but I think preventing the circumstances that allow for such crimes is perhaps more important here (which my model also highlights). [I'm of course working under your assumption (your assumption) that this is case were we deem that the private is no longer a danger. In the real world, however, it's not always going to be so black or white and some sanctions will likely be required.] I acknowledge, though, that I need to think more about such cases. [I'm curious what Derk would add ;)]

As for how easy it will be for the population at large to adopt a non-retributive approach to criminal justice, I don't know. I'm hopeful that it's possible. It seems that countries like Norway have gone pretty far in that direction with pretty good results. I don't think we will ever lose the retributive impulse on a personal level (and perhaps that's good), but from the prospective of policy I think we can successfully be moved by philosophical and pragmatic arguments and concerns.

As for your third point: "In a democratic system, candidates may campaign on a platform promising to bring desert back, to punish those who deserve it, etc. They may well be successful." I'm afraid that might be true. But it doesn't justify retributive punishment! I says more about our broken political system.

Hi Neil (and Gregg): No, I absolutely reject that conditional. I think it is an absolute mistake to identify libertarianism with indeterminism in the brain (which is why I think the "libertarianism" of people like Kane is wildly mistaken, not libertarianism at all). As I have argued in my own work, *only* a very special kind of indeterminism would be evidence for libertarianism--indeterminism which is non-random and systematically improbable, violating in a manner of speaking the normal quantum wave-function probability.

As such, let me ask you a question: suppose we found when we looked in human brains "indeterminism" of this sort--quantum collapse values that are consistently and systematically improbable given the normal quantum wave-function, with *no* further physical explanation at all, which correspond to the person's choices. When confronted with this kind of "indeterminism", would you really be inclined to say that it's *not* evidence for libertarianism? It would be (A) unique quantum-collapse values in each individual's brain (their own personal "wave-function"), (B) not predicted or explained by any physical law or mechanism whatsoever (including the normal quantum wave function), (C) corresponding directly to their (unpredictable) choices, and (D) which might change in response to new choices (with the person's "personal wave function" changing in response to new choices). Question: how would this *not* be evidence in favor of some kind of libertarianism?

V. Alan White: It is not at all unfair to analyze credences for false scientific beliefs after the fact--not if they made a mistake. If they made a mistake, it is a mistake we should think about so that we do not keep making it! Let me explain why.

One thing that bugs me about the history of scientific progress is that we treat people as "geniuses" who *weren't* geniuses, but who were the only ones who bothered to question the scientific dogma(s) of their time. Consider Einstein. Einstein was a very, very smart man. But was he a genius? Let us consider what his first relativity paper *actually* did. Many people prior to Einstein (Poincare and others) saw way before him that existing observations might lead us to question the absoluteness of space and time. But *none* of them thought the idea worth pursuing. Why? Because they were "certain" that space and time "had" to be absolute. Why? Because that is the dogma that the science of their time taught them--with no real evidence--to believe. Einstein was the only one who took the possibility seriously enough to pursue to its logical conclusion. His relativity paper does nothing more than tease out the implications of two very simple ideas: (1) the observation that the speed of light is the same in all reference frames (which was known in science), and (2) the observation that the laws of physics appear the same in every reference frame. That is *all* Einstein did, and we now consider him a "genius" for doing it.

This is the problem, then. Repeatedly, throughout scientific history, people become locked into a dogma--one that lacks any evidence whatsoever. I find your remark, "lots of experiments and theory favored belief in the ether", very instructive here. Actually, there were *no* experiments or theory that favored belief in the ether. There never were. People who believed in the ether *invalidly* inferred the existence of something that no experiment or theory at the time gave them any good reason to believe it. They went *beyond* their evidence, instead of sticking to the actual evidence they had. It was only the "genius" Einstein who said, "Wait a minute. Maybe we don't have any evidence for ether, and maybe we should question all of the *assumptions* that led us to invalidly believe in it", who we now recognize as a genius for doing so.

This is why, far from "[losing] all proper context of what constitutes ongoing epistemic evaluation", I think it is the *failure* to appropriately reflect on past mistakes that does this. In my view, scientists (and philosophers) make the same epistemic mistake over and over again, never learning from the past: attaching high credences to things that we do not have good evidence for. This holds science *back*. People wasted decades "theorizing about the ether" with no observational evidence (and no theoretical evidence supported by observation) supporting it at all. It took one *day* for one person who simply questioned that dogma (Einstein) to change the world. We should learn from this, not continue to ignore it.

Finally, you say, "there is no good evidence that they function in ways that specifically depend on subatomic indeterminism." This is not true. There is emerging evidence that certain biological functions depend critically on quantum processing (even in plants, thanks to evolution!), as well as emerging evidence that brain-processing does. See:


Part of the concern I have with your hypothetical is that it strikes me as a bit implausible. It's not merely that we can't seem to give up our retributive intuitions when we contemplate Nazi war criminals, but it's that it would take an awful lot to convince us that someone truly capable of mass murder and that actually has committed mass murder is no longer a threat. Because of how we are wired as human beings, it would take our witnessing the very human process of contrition to truly satisfy our intuitions that a person is no longer a threat. We will always be suspicious of a person who hasn't expressed genuine remorse for their actions, and our moral sentiments will consequently always remain unsettled. I imagine this would be the case no matter how much scientific evidence you could throw at us that might otherwise indicate a person is guaranteed to no longer be a threat. This is probably just a mere fact of human psychology and the results of evolution.

Now, imagine a situation where you have a lowly SS agent who is culpable for his complicity in various war crimes. Although he didn't conceive of the crimes himself, he followed orders against the nagging demands of his conscience. He found a way to rationalize what he was doing and block out the evil he was committing, but soon after the war, his horrific crimes began to haunt him as the full moral weight of his actions became ever more present to his mind. He then rapidly begins to feel true remorse and moral anguish for his actions. As a result, he spends the rest of his life attempting to atone for his crimes, seeking to pay penance to the families of those he harmed and to publicly decry the crimes of Nazism.

It may turn out that the families of those he harmed may still wish to see him rot away in a prison cell, but my intuitions are that this is not a man that still needs to be imprisoned. In fact, I think we should even encourage his agency rather than diminish it. I think he at least now has the opportunity to use his position to reach out to other criminal agents of authority, and spend the remainder of his life promoting welfare.

It's true though that we may never be able to fully give up our retributive impulses, but it's important to note that for many grievers, letting go of one's anger and desire for retribution is often an important part of the healing process. Indeed, for many people, to see a man broken by his crimes and express true remorse may very well be sufficient for the victims in letting go of their own preoccupation with the past in a way that helps mitigate, or in any case helps them come to terms with their own suffering.

In sum, it may take more time for us to shake our retributive reactions to these more extreme cases, but on principle, it's not only possible but even ideal. It also seems that even if it will be a more challenging process for these more extreme cases, we can certainly take a quarantine approach for cases of negligence like Bert (as Gregg mentioned), and perhaps one day down the road, cases of a more traumatic nature. The benefits of shifting toward forward thinking punishment practices are manyfold. At the very least, it may allow victims to let go of the past and focus on the future, and repentant sinners to use the rest of their life as a project for redemption and good.

I think I see why free will scepticism is an attractive position. But I'm not sure I see the point in asking e.g. what we should do with criminals, given free will scepticism. I suppose whatever happens to them will be beyond our control, if free will doesn't exist.

"In conclusion, I have argued that the public health-quarantine model not only justifies the incapacitation of dangerous criminals, it also demands that we monitor and redress social and economic inequalities and that we prioritize prevention."

How can we monitor and redress inequalities without free will? I mean, we can find ourselves monitoring and redressing, but we can't decide to start doing it, if we don't happen to be decide it by the influence of some cause.

Izzy, thanks for your assistance and the more detailed example! I share your intuitions.

Marcus, I would say (a) not all libertarians agree on what is required for basic desert MR (as your rejection of Kane makes clear), (b) most libertarian accounts attempt, at best, to show that libertarianism is *consistent* with a scientific workdview, but very few (if any) claim we can currently varify the important empirical assumptions of these different accounts (I think you would admit your account is not yet empirically verified), (c) beyond the questionable scientific assumptions of these various accounts, libertarianism faces non-empirical objections which remain regardless of the fundamental nature of reality (I think that was partly Neil's point). Give these facts, I have a very hard time overcoming Double's argument.

I will now cast the spell which animates Alan's horde of free will zombies. The magic words are: general deterrence. And then I'll also want to raise trouble for some other words ("basic desert" and "retributivism"). First, general deterrence.

Unlike viruses, criminal tendencies can be discouraged by what happens to other criminals. The flu viruses in my body don't care one whit if the flu viruses in your body are quarantined, or even dosed with antivirals. But when we consider the social benefits of imprisonment, or burdensome community service requirements, or punitive fines, most of the benefit comes from general deterrence, not quarantine type effects.

What justifies society in making an example of *you* to deter *me*? Sure, imprisoning *you* in order to prevent *you* from harming (more) people makes obvious sense, regardless of free will skepticism. But making-an-example seems justified only if you are actually(*) responsible: for example if you wronged someone voluntarily or negligently. (*)Or at least free-will-zombie-ly responsible, which I'd call a distinction without a moral difference. So if people don't have any pragmatically workable moral responsibility, then it would be wrong to punish for general deterrence, and only right to punish for specific deterrence. And then punishment will be under-supplied, compared to the social optimum.

So, I'm certainly calling for "negative retributivism" as defined by Alec Walen in his SEP article on retributive justice (esp. sec. 4.6, here: ). But I disfavor the terms "retributivism" (simpliciter) or "basic desert". They seem no clearer than the problems they supposedly illuminate. Unless we are discussing particular philosophers who take an extreme position on those issues, we won't know which way to classify them. And even then, we'll be ignoring philosophers who are closer to the truth, er, I mean, to the middle :)

"The desert at issue here is basic in the sense that the agent would deserve to be blamed or praised just because she has performed the action, given an understanding of its moral status, and not, for example, merely by virtue of consequentialist or contractualist consideration. (Pereboom 2014, 2)" I still don't know what these two strange bedfellows, consequentialism and contractualism, are supposed to be doing together. But maybe it's a different kind of "contractualist" than the one (Scanlon in particular) that I'm thinking of.

My points about general deterrence share some of Angra's concerns, but I wrote them before refreshing the page and seeing his comments. I'll leave them as is, since I still think they provide a useful perspective.

N eem,

The lack of libertarian free will does not imply that we are unable to act according to our wants and plans. It just says that we aren't the ultimate cause of our actions.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan