Blog Coordinator

« Free Will Skepticism and 'Ought' Judgments | Main | On a Freedom Requirement for Forgiveness »



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

Another nice post, Justin. That said, I think it's important to first note that when people are exposed to anti-free will primes, some studies have shown that they are less retributivist and more consequentialist when it comes to their punishment judgments (see Shariff et al. 2014). And while I have concerns about much of the empirical literature on folk retributivism (see Nadelhoffer et al. 2013), I nevertheless think several things are clear at this point: (a) there is a strong strand of folk retributivism (at least in American populations), (b) folk retributivism is tied in various ways to people's beliefs about free will, (c) the less people believe in free will, the less retributivist they become, and (d) there are important individual differences in people's baseline beliefs about retributivism, free will, moral desert, etc. that often get ignored by psychologists and philosophers alike. The fact that these individual differences exist can serve as a way of examining claims about what would happen if we gave up on desert as the grounds for punishment. After all, we can look and see whether people who have already given up on desert (or who never felt it's pull in the first place) are worse off than those who have strong desert intuitions. Part of what I tried to argue in my "darkside of free will" paper (Nadelhoffer and Goya Tocchetto 2014) is that while there may be some positive correlations/associations between people's free will beliefs and their other moral and political beliefs, there are also some worrisome correlations as well. At least in my case, while I think that our beliefs and practices would change if people were to give up the belief in free will, some of these changes would be for the better even if other changes might not. In this sense, I suppose I am a mixed skeptic (rather than an optimistic or pessimistic skeptic).

That said, I want to say a few other things about your post:

First, just because someone gives up on the belief in free will, it doesn't mean they have to give up on the distinction between voluntary, involuntary, and non-voluntary action. It also doesn't mean that they have to give up on the basic mens rea distinctions that are already in play in the criminal law. Because these can all remain intact, I think there are straightforward ways of dealing with your elevator scenarios. I don't need free will and deep (or even basic) desert to explain why it makes sense to be angry with those who voluntarily and intentionally cause harm and why it may not make sense to be angry with those who involuntarily or merely recklessly cause harm. I can ground my beliefs, practices, and even attitudinal responses in a purely consequentialist way (without appealing to free will).

Second, I think prepunishment is a bit of a red herring here (and elsewhere). Consequentialists have had the theoretical machinery to deal with these kinds of cases since Bentham laid down his beliefs concerning when punishment is justified and when it is not justified. I think what he says on this front is sufficient to address the often discussed problem of punishing the innocent as well. It both cases, it is difficult to construct a realistic case where pre-punishing (or intentionally punishing an innocent) would really maximize utility. I could say more if pressed. For now, I just wanted to point out that neither of these two issues depend on whether we have free will. Or at least that's how it seems to me. After all, one could be a pure consequentialist about punishment and nevertheless believe in free will. So, it seems to me these are issue best treated separately.

Yeah I'm a little confused about how the case of pre-punishment relates to what I thought was the topic of the post: optimism about having a strictly forward looking account of blame. Pre-punishment, even if it could have a consequentialist justification, doesn't seem to be an example of blame. It seems more like quarantining people who have been exposed to a contagious disease. And pre-punishment has nothing to do with Strawson's central concern: maintaining deep human relationships.

The more serious worry as I understand it involves more common (and much less dramatic) instances of blame. Can we retain relationships with our friends, family, wives, husbands, if we think it's only appropriate to blame them when it leads to good consequences? Let's say I forget to go to my daughter's piano recital. Is it only appropriate for her to blame me if the blame will make me less likely to miss one in the future? Or is the blame appropriate because I let her down...period? The idea that she would have to engage in calculations of any kind to feel that her blame is appropriate seems wrong. That's not how close relationships work. Those cases push my Strawsonian buttons (and challenge my earlier optimism) much more than pre-punishment cases because, again, they don't seem nearly as connected to blame and maintaining deep interpersonal relationships.

Thanks for the comment and the references, Thomas. I haven't read your (2014) on this so I'll be sure to check it out. Also, I think the mix of pessimism and optimism that you seem to endorse about the ramifications of FW skepticism sounds more plausible than a strictly optimistic view. That said, let me offer a brief reply.

I realize now that I should have more explicit about the pre-punishment concern. The concern I was trying to point to was with regards to blame and not punishment proper (I fixed the concern and added a sentence to clarify that). So, much of what you say refers to punishment and not blame but let's assume for a second that they are the same (there might be good reason to think they're not but let's bracket that for the moment) in an effort to see the root of our disagreement.

I'd be curious to hear how that internal narrative would go re: spelling out the justification for overtly blaming or directing your anger at another without grounding that anger or blame in desert. Mckenna's view is nice but I think his view is best understood only when we understand it alongside basic-desert. He seems to agree on this point, though I may have misread him.

You say: "First, just because someone gives up on the belief in free will, it doesn't mean they have to give up on the distinction between voluntary, involuntary, and non-voluntary action."

I would want to resist this claim. Assuming by FW you mean the FW that grounds desert-based attitudes like blame, anger, etc.

But, given that the conversation between Ryan and I went down this path I think it would be fruitful to post on this very topic sooner rather than later, so I likely will. So, maybe we can save our disagreement here for that post?

Suffice to say that I REALLY like Helen Steward's agency incompatibilism, thus I'm hesitant to accept that even the basis of criminal law (as well as a meaningful distinction between voluntary and involuntary) is on firm ground in the wake of FW and MR skepticism if we put forth a purely consequentialist framework.

For those not familiar with Steward's work I will grossly summarize. Steward argues that action, properly conceived of as a kind of input into the world, an input that is genuinely a settling *of the agent* is inconsistent with determinism. If one cannot settle (here she argues *consistently* for an incompatibilist rendering of settling) whether one would A or not, then one is not acting when one does A. It's best to see A as a happening in the same way a rock happens to roll down the hill if it's windy. Now, Steward doesn't say this, these are my words, but this is how I understand her view and I think it should be clear why I am hesitant to accept that the distinction would hold between happenings on the one hand and voluntary action on the other.

To your 2nd point re: prepunishment I guess I would like to hear more about why you don't think prepunishment is a concern. Or, to be specific, why preblame wouldn't be a concern? At first glance, (and even after deep reflection) it seems easy to conjure up examples where one could maximize utility by blaming someone for something they never did. Run of the mill utility monster scenarios are easy to fire away. See here:

But seriously, I think there are other counterexamples as well.

Lastly, I agree that there are views about blame that need not invoke the FW that grounds desert claims. I admit that in the post, at least I tried to. The point of the post was to analyze how to make sense of these new justifications for blaming. So, given that I endorse a mix of retributivism and consequentialism (I think *in most cases* that we should blame only folks who deserve it and the amount of overt blame that one should receive should be proportionate to what produces the best consequences -I'm being quick here) I wanted to try and make better sense of the purely consequentialist or forward-looking account of blame that has seemed to pick up some steam as of late.

To be clear, my project is to look at what is left in the wake of not having free will. So, I grant that one could be a pure consequentialist about punishment and have a belief in free will. But given that skeptics cannot hold a belief in free will it helps me, for my purposes, to set the stage for the discussion by being clear on what conceptual space is available. This is why these issues might be best discussed side by side rather than separately as you suggest.

Thanks for the comment, Tamler.

I made the change to say preblame instead of prepunishsment.

The point in bringing that example up was to show that intention or action need not even be invoked to blame on a forward-looking account. And this seems worrisome.

FWIW, the more serious worries that you mention I invoke in my diss (ch 3) and was trying to hint at in the post. I chose the prepunishment/preblame focus to just get the conversation going. Maybe I should have just went for the sort of example you describe right off the bat.

Anyway, hopefully that clarifies things a bit.

Perhaps this is not the place to debate Helen Steward's intriguing views. (I've done a bit of this elsewhere, in which I've tried to argue that just as moral responsibility doesn't require alternative possibility, so action or agency does not as well--it seems to me that the same considerations apply.)

But allow me also to mark this point. I don't see why "settling" requires indeterminism! Right, you point out that Steward *argues* for an indeterministic reading, but it just doesn't seem plausible to me that settling some issue requires indeterminism. It seems to me that it is a matter of figuring out what my best reason is (or all things considered reason is)--and that process of identifying and weighing reasons is consistent with determinism.

So, ahem, although I respect Steward's and your position on this, I think it is quite contentious!


While I have a pretty good sense of what people mean when they discuss pre-punishment--although I share Tamler's concern that the cases used in this context are better viewed as preemptive quarantine rather than punishment proper--I am not entirely sure what you mean by pre-blame. Clearly, I can't have *reactive* attitudes preemptively or else they wouldn't really be reactive. They would be proactive attitudes instead, no? Other ambiguities arise in this context as well. For instance, I could pre-blame someone for something I know they will do (which is the analog to pre-punishment) but I could also pre-blame someone for something I know they won't do (which is the analog to the problem of punishing the innocent). It's not clear which type of pre-blame you have in mind. In both cases, I could justify the pre-blame based on forward-looking considerations. But I am still not clear why this is an issue for free will skepticism. As I said earlier, the same problem arises for those who defend consequentialist theories of punishment (or who attempt to ground the reactive attitudes in purely forward looking considerations) but who nevertheless believe in free will. So, it's not clear what role the belief or disbelief in free will is playing here. Perhaps you could say more.

For now, here is a case of pre-blame that I find unproblematic (assuming, of course, that I have understood what you have in mind correctly!): My wife recently brought a dog back with her from Brazil (bring our total to four dogs!). He is a stubborn but smart little Westie. So far the only problem we've had is that he really likes to bark at cats--which is a problem because the neighbor across the street has several outdoor cats. As a result, if we didn't intervene and retrain poor little Boris, he would drive us crazy with never-ending barking. So, we've been spraying him with a water bottle whenever he barks. He is a quick learner. But the temptation is strong in the little guy. Sometimes, I see him staring intensely across the street at a cat. Despite the training, sometimes he still barks. So, if I catch him staring intensely, I spray him even though he hasn't barked and even though I know there is a chance he may not bark at all (in which case, the spraying would turn out to be unnecessary). But by engaging in a (pre)blaming practice, I err on the safe side and make it more likely he learns the house rules sooner rather than later.

At no time, however, do I need to appeal to free will to justify my actions. The fact that he is a neuro-typical adult dog who has the capacity to learn the rules suffices. Whether I spray him prior to barking or during the act of barking doesn't seem to matter much to me so long as he has the capacity to understand why he's being sprayed. Contrast this with an imaginary case involving a neuro-atypical dog with obsessive-compulsive behavior that is not responsive to either blaming or pre-blaming practices. Under these circumstances, spraying the dog would be unjustified because fruitless. Yet, here again, I need not appeal to free will in explaining why this is the case. The mere fact that my behavior is pointless suffices.

Notice that the distinction between these two cases still holds even if we assume for the sake of argument that determinism is true--that is, the difference between the two dogs and what I am justified in doing to them is orthogonal to the issue of determinism. Now if we further assume that incompatibilism is true, I can still draw the relevant distinction I need between the two dogs--which suggests to me that the worry about free will skepticism here is overblown.

Hi Justin

Interesting post. I guess a consequentialist/optimistic sceptic could just say that best consequences are achieved by *not* prepunishing/preblaming people. This is of course going to depend on the facts. But it seems quite plausible that better consequences would follow from not prepunishing/preblaming people.

Great post Justin. I don't have too much to add, save to note that I am also sympathetic to the Strawsonian account of why it is a significant loss if we give up on the traditional reactive attitudes (and supplant them with purely consequentialist considerations or whatever), especially for our personal relationships - in much the way that Tamler suggests. Whether praise/blame/etc in personal relationships has good consequences or not, it just seems like the wrong sort of justification to be looking for.

Along those lines I'm very much looking forward to your post on forgiveness; I talked about that a bit in my dissertation, and it's something I've been thinking about recently. I'm tempted to start in on agency incompatibilism here, but I'll save that for when you do a post focused on that topic - which I'm also very much looking forward to.

(also as long as we're sharing utility monster comics, here's another one: )

Isn't the onus on Strawsonists to explain how exactly MR skepticism undermines love (attraction to each other's body/personality) or compatibility (sharing passions and interests). In fact, removing moral blame from the equation only allows those two aspects smoother sailing. As for the much ballyhooed loss of interpersonal moral praising, how many mature relationships actually consist of a stream of acknowledging each other's moral virtuosity?

Empathy, affection and compatibility is all there is to meaningful interpersonal relationships and they are all enhanced by MR skepticism.

Brent, although Strawson himself thought love was incompatible with MR skepticism, defenders of his broader position aren't committed to that view. A more plausible line is that MR skepticism misunderstands blaming emotions in the same way that "love skeptics" might misunderstand love if they thought it was necessarily tied to something non-natural. The crucial quote from F and R on this front is:

"Even the moral [responsibility] sceptic is not immune from his own form of the wish to over-intellectualize such notions as those of moral responsibility, guilt, and blame. He sees that the optimist’s account is inadequate and the pessimist’s libertarian alternative inane; and finds no resource except to declare that the notions in question are inherently confused, that ‘blame is metaphysical’. But the metaphysics was in the eye of the metaphysician. It is a pity that talk of the moral sentiments has fallen out of favour. The phrase would be quite a good name for that network of human attitudes in acknowledging the character and place of which we find, I suggest, the only possibility of reconciling these disputants to each other and the facts."

Justin, I agree with Thomas that pre-blame is hard to make sense of. But even if you could make sense of it--say, my daughter blaming me because she somehow knows I'm going to miss her piano recital--then it seems unproblematic. She should blame me for that.

(In my defense, those piano recitals are brutal. It's one thing when a cute little 6 year old to come up and play jingle bells. But Eliza's recitals also feature adults who are just learning how to play. So I have to watch a 55 year old suck at piano and then applaud like it's inspiring. Why is that OK? I don't make people watch me try to learn Spanish. But somehow with piano it becomes acceptable for adults to force other people watch them learn a new hobby. But this is a separate issue.)


Is that with the Suzuki method? (The piano recitals.) I found those torture. I took both of my kids for years--and now my daughter hates piano and my son doesn't play much (although he's pretty good).

We survived ballet, too.

Now all I have to do is go to philosophy dept meetings and colloquia! (Sometimes it makes me nostalgic for "Twinkle, Twinkle, Little Star..."

I'll be sure to get to everyone's comments tomorrow. My oldest son is prepping for the science fair and I've been giving him a hand. His tentative title: "Brain Malfunctions", exciting! Speaking of which, I hope to get to a post re: psychopathy later this month. I think psychopaths are morally responsible, most don't, at least this is the sense I get. Should make for a fun discussion. Anyway, I better get posting if I hope to get to most of these topics in the coming weeks: forgiveness; agency incompatibilism; psychopathy; degrees of moral responsibility; the authenticity and epistemic condition; MR and history; and an epistemic argument against FW skepticism.

I'd just like to say some general comments and offer a response to Tamler (that’s relevant to some of the conerns Thomas has raised) before jumping back into discussion tomorrow and answering comments specifically.

First, the example of pre-blame was to point to the fact that one *could be blamed* WITHOUT performing an action. This is radical! That seems odd doesn't it? FWIW, I'm not optimistic about forward-looking blame because I believe that a world where we try and figure out who to blame based on some sort of calculation is problematic (as you eluded to earlier Tamler) though I still think this line re: blame is worth pursuing. I'll try and give an example tomorrow to better articulate the concern. That said, I am in complete agreement with Tamler that the other concerns are deeper than prepunishment (or preblame), I am simply trying to pile it on so to speak. Which leads me to my next point.

Thanks for playing along with the preblame/prepunishment connection that I am trying to draw. I wanted to test it out here to see how folks felt about it. Your questions and comments have been very helpful already. Tomorrow I'll try and better explicate the source of my concern re: forward blame with a couple of examples and talk more generally about the reactive attitudes.

Tamler (relevant to Thomas as well), you say "But even if you could make sense of it--say, my daughter blaming me because she somehow knows I'm going to miss her piano recital--then it seems unproblematic. She should blame me for that".

Let's say she doesn't know (she need not know on the forward-looking model). All she would have to know is that the consequences of blaming rather than not blaming are better. What she does know, however, is that when she blames you for not going she gets tons of pleasure from it as do you do when you’re 60 when you reflect on her irratic blaming (for whatever reason), maybe she reminds you of yourself at an earlier age, doesn't matter. So though it infuriates you to be blamed time and time again until she is 21 (every single day and she's dead serious about it) the consequences of her doing so are positive in the long-run. Though you feel a little annoyed with her in general, from your end things aren't better overall. Let's also say you get divorced and your ex-wife also enjoys the daily blaming she hears when she overhears your daughter blaming you on the phone. Hopefully you see where I'm going with this. Let's assume that overall happiness is our metric here (measured in terms of pleasure), just to get my worry off the ground. Now, let's also assume you’re divorced because your wife was a bad person not because you were an ass. It seems weird to say that your daughter blaming you everyday (even though you never do anything that you are being blamed for) is justified on grounds that the consequences produced by her behavior produce more pleasure overall, doesn't it? Seems like we lose something deep by losing basic-desert. At least basic-desert limits who we can justifiably blame, among other things. Also, just because we screw up with basic-desert it doesn't follow that we wouldn't screw up with forward-looking blame. It's also worth noting that epistemic access to how folks will respond to forward-blame at certain times is sketchy so optimism should be cautious at best! But I’m getting ahead of myself…

Tamler, you also said "she should blame you".

But, were assuming you *would be* missing the recital which is maybe why she should? I think that's important (and seems itself be desert-related).
Anyway, I'm posting this without a re-read so apologies in advance for any issues with my writing (not that a read over would have necessarily helped). Hopefully this long-winded attempt was fruitful to see my concern a bit more. Apologies if it only made it worse.

Justin: Thanks for another great post! On the defense of optimistic free will skepticism, I endorse virtually everything Thomas says in reply, and I won’t add anything to what he says. On settling, I agree with John that settling doesn’t require indeterminist free will. One might think of settling by an agent as a kind of difference-making, and there are good compatibilist accounts of this notion. In Carolina Sartorio’s view (“Making a Difference in a Deterministic World,” Philosophical Review 122 (2013): 189-214), moral responsibility requires difference-making in the sense that the agency-involving actual sequence leading to the action makes an agent responsible for the action only if the absence of that actual sequence would not have made the agent responsible for the action. In a recent paper (The Phenomenology of Agency and Deterministic Agent Causation, on my website) I tweak this account to yield a compatibilist criterion for settling. In that paper I also argue that settling requires agent-causation, but that the deterministic accounts of agent causation that Ned Markosian and Dana Nelkin advocate will do. Here is the criterion:

(S-AC) An agent settles whether an action occurs only if she agent-causes it, where the absence of her agent-causing the action would not have caused that action.

Put in terms of David Lewis’s semantics for counterfactuals, the idea is that an agent settles whether an action occurs only if she agent-causes it and in the closest or most similar possible worlds in which she does not agent-cause the action the absence of the agent-causing would not have caused that action.

John, no, not the Suzuki method as far as I know. That sounds horrific. And I think incompatibilists, compatibilists, semi-compatibilists, revisionists, metaskeptics, libertarian compatibilists, and even "willusionists" can all agree that pre-blame is perfectly appropriate when it comes to department meetings.


I agree that your particular pre-blame scenario is incompatible with deep interpersonal relationships. But isn't that primarily because in your scenario my daughter is making me suffer daily just for the fun of it? The pre-blame aspect plays a marginal role at best...


How can we skip the specification of what basic-desert entails, and still make any headway? What we can still do, depends entirely on what we're giving up. Imagine you're on a cross-country trip with a friend, and he opens the hood and says "I'm going to remove the rattling thingy. I think the vehicle will operate just fine without it." Everything depends on what he means by "rattling thingy." No?

Fair point, Paul. Didn't want to open that can of worms, though. Luckily, it seems that the discussion got brewing without it.


Thanks for the references! Clarifying your position is very helpful for me in seeing where our views diverge.


I think that would be tough to prove. With only consequences as a metric it's quite easy to put forth counterexamples once you get clear on what your axiology is, IMHO.


But that's my point. The tortue is justified on the forward-looking account of blame. Such blame would be justified. So, many gravitate toward this forward looking blame because they think basic-desert based blame is the root of so much that is wrong with the justice system (and our interpersonal relationships). Though I am in agreement that sweeping changes ought to be made with regards to how we treat prisoners and those who do wrong by us, I do not think that we get there on a strictly forward-looking account for many reasons, one of which it allows folks to justify atrocities in the name of "better consequences".


Looked at your paper; very interesting and well argued. I have one concern: does your position on (S-AC) (and perhaps Carolina's) require a necessitarian view of determinism to be most forceful? If one were a Humean compatibilist qua Beebee and Mele, then the resultant dual-ability would dismiss this account of agent-causality in all close possible worlds yet retain it for an actual (Humean) deterministic world. But then the same could be said of a libertarian actual-world agent-cause. So at least for Humean compatibilism and libertarian agent-causal views, there would be no reason by (S-AC) to prefer either in terms of settling. But that sets the scene for other considerations to favor a necessitarian-based account of settling over indeterministic and Humean actual-world settling by these being subject to stronger factors of metaphysically open-future luck than just past-based matters of constitutive luck that still might affect necessitarian accounts.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan