Blog Coordinator

« Ahoy! | Main | What a Rare Mood I'm In (A Weekend Post for Weekend Thoughts) »



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

Hi Justin. I must say that you've started sprinting out of the blocks if (as I see it) the month is more of a marathon of guest authoring!

First a question. Even granted your entire line of argument, does this count against FW incompatibilism as well as MR incompatibilism, which I take is your target? I'm trying to accommodate John's semi-compatibilist distinctions here to be as clear as possible.

Second--and I think this involves some issues of your first post--does the pro tanto approach attempt to lay down baseline minimalist conditions of dessert based on action, here focusing on Mac instead of Charlie? I take it to be so, and so construe the "basic" of "basic dessert" as you use it to be (source, control, etc.) minimalist rather than say, libertarian maximalist (source, control, etc.) "basic dessert." Am I right on that? Please excuse me (if I deserve it!) for being so simple-minded if that distinction seems obvious to others. But I'm wary about people talking about basic desert and meaning wildly different things (as some said in the previous thread).

Third, I wonder if the assumption that Mac and Charlie can be given equal consequentialist weight for blaming is plausible given that, as you say, Mac's involvement is at all discernible as asymmetrically causal to the harm as compared to Charlie's. Any consequentialist I'd think would concede that asymmetry is relevant for determining further action to produce good consequences of blaming, even if only to reinforce Charlie's (non-)conduct against Mac's. If that's so, then even the consequentialist can make reasons-based discriminations between Mac and Charlie in fixing and constructing consequences of blaming, and even if they concede that both parties are subject to utter causal determinism.

Now back to watching my Packers make the Vikings' defense look like Swiss cheese.

Apologies for the auto corrects of "dessert" for "desert" on a couple of instances.


The conclusion that incompatibilism is false follows from the claim that given two agents, A and B, in a deterministic scenario where A performs some morally bad act and B does not, we will have less reason to blame B than A (and holding fixed anything that might figure into a consequentialist calculus, etc.). The only thing that explains the difference, for you, is that A's agency is implicated in the causal genesis of some bad state of affairs in some way that B's agency is not. Since determinism doesn't affect this relation, incompatibilism is false (I know that's rough, but I think it captures the main thread of the post).

I'm skeptical of the initial claim that we have less reason to blame B than A (Charlie and Mac, respectively, in your case; I'll switch to your names to make it easier). I think of it like this. Mitigation is overdetermined in Charlie's case; not only did he not do anything wrong, but he's under the influence of the distant past and the laws of nature. So two independent factors undermine Charlie's blameworthiness. Mitigation is not overdetermined in Mac's case. He did something wrong, but he's under the influence etc. etc. such that he's no longer blameworthy. But if neither agent is blameworthy, then we have no reasons whatsoever to blame either of them, so the 'lack of reasons' is equivalent between the two. An informal way of stating my worry is that Charlie is 'super-unblameworthy' where Mac is just 'unblameworthy'. But in both cases we have no reason whatsoever to blame one or the other, hence the lack of reasons entirely.

I'm completely lost. You said "causal determinism is true" and "incompatibilist". As far as I understood it, if you assume those two things, you end up with no free will and no reasons to blame or praise anyone. What reasons do you have for thinking that there are still reasons for praise or blame in that situation?

Hi Justin,

This is a very interesting challenge to the incompatibilist.

One question about the Schroeder-style argument you run: Is it possible that there could be less reason to X than to Y, not because there is some reason to Y, but rather because there are reasons to not-Y (or no reasons to Y) and even more reasons to not-X? To put the question another way, can it be the case that we talk about proportions of reasons like we talk about proportions of negative numbers? If I am in the hole $10 and you are only in the hole $5, then I have less money than you do. But neither of us has any money.

If this possibility is live, then the argument doesn't quite go through, does it?

[I'm sure someone has thought of the general point already. So please feel free to point me to the literature if you know of a discussion of it (perhaps Schroeder himself addresses it in his book).]

Hi Justin, great to see you featured! (I have to say how proud I am here, since Justin is my first MA student at GSU to have gone on to become a professor, of course only after going to Riverside to learn just a bit more from John...)

I'm convinced by your argument here. But of course, I would be. But what do you say to a skeptic about deserving blame, who says that Mac and Charlie *deserve* the exact same amount of blame--namely, none--and that our consequentialist reasons for "blaming" them can be equal, but only in really weird circumstances, such that typically we'll have very good reasons for reacting to and treating them differently?

And our intuitions might support this response if we think of Mac as someone for whom consequentialist "blame" won't do any good (maybe he died), and perhaps also if we highlight why "blaming" Charlie would be good (maybe it will prevent 1000 people from embezzling and Charlie agrees to pretend to have done it).

Hi Alan,

In the spirit of Dennis Kimetto's record breaking sub 2:03:00 marathon (4:42/mile average) last week, I'm of the mind that one must get of the blocks quickly. As to your questions, though...

I should've been more clear that I'm taking myself to be presenting an argument against the thesis that moral responsibility is incompatible with causal determinism. (Biographically, I guess I'm a kind of semi-compatibilist, except unlike John, who is officially agnostic about the compatibility of freedom and causal determinism, I'm happy to concede that fight to the incompatibilist (at least with respect to the "freedom to do otherwise").) This means that the scope of the argument is limited in this respect.

As to your second question, I'm not quite sure about how you're distinguishing between "minimalists" and "maximalists." There are two plausible distinctions one could make between comaptibilists and incompatibilists that might be along these lines. The first is the kind of response that's tied to moral *responsibility*. Compatibilists like John, Jay Wallace, Stephen Darwall, etc. connect responsibility to the reactive attitudes, whereas incompatibilists have sometimes put things a bit more grandly, in terms of "heaven and hell responsibility." The former conception of responsibility is plausibly more minimalist than the latter. But one could also note that the conditions for moral responsibility are less demanding on compatibilism than on incompatibilism, and so that distinction also tracks "minimalist" and "maximalist" variants. My own view is minimalist in both respects, but the argument I present here isn't, as far as I can tell, committed to that minimalism.

As to your third point, I admit I just stipulate that the goods associated with blaming Mac might, in some contexts, be equal in weight to those associated with blaming Charlie. So perhaps more argumentation is needed there. But suppose that Charlie's generally a bad guy and he's well positioned to be elected to an important post where he'll do lots of harm. By blaming him for the embezzlement, we could ameliorate that harm significantly such that on balance the total good could, in principle, be equal to that of blaming the person who brought about the bad action through his agency. No doubt there are other stories to be told (involving God or the end of the world or some other far fetched state of affairs...).

Eddie Lacy finally showed up; my fantasy team might be alright this week.

Hi Phil H,

I guess I took myself to be providing an argument that even if we assume causal determinism, there will still be reasons to praise or blame that don't depend on the consequences of praising or blaming. In short, I want to claim that if there are no such reasons, as the incompatibilist claims, then it would follow that there would be no reason to blame Mac for embezzling. But plausibly, there is more reason to blame Mac than to blame Charlie, who was completely innocent. This means that there must be some reason to blame Mac. If such a reason can't be given a consequentialist basis, then it must instead be grounded simply in the fact that Mac is guilty of embezzlement.

The general strategy, then, is to show that accepting incompatibilism (and the proposal about desert and reasons I floated in my first post) has some implausible normative consequences.

Does this help?

Really fun posts so far, Justin.

Following up on Ben's comment, couldn't it be that, since Charlie didn't even x, whereas Mac did, it's clear that there is even less reason to blame Charlie for x-ing? After all, Charlie didn't even x - so there's a class of reasons, those depending on x-ing, that can't apply to Charlie though they *might* apply to Mac.

That said, if the incompatibilist denies that anyone deserves blame (in the relevant sense), then shouldn't they just say that none of these reasons in fact apply to Mac? So while it seems like there's more reason to blame Mac, this is just because there are sets of reasons that could apply to Mac that couldn't apply to Charlie (holding everything else fixed). But since incompatibilism and determinism are both true, these reasons don't in fact apply/obtain.


That's the right place to push, since you might think that the fact that Charlie is innocent is itself a reason to refrain from blaming him. And this might lead you to think that it's this fact that explains why there is "more" (in some sense of more) reason to blame Mac than there is to blame Charlie.

But while I admit that being innocent of some wrongdoing is itself a reason for others to refrain from blaming you, I don't think that can fully explain the judgment that in this case there is more reason to blame Mac than there is to blame Charlie. To see this, consider a third co-worker, Dennis, who helped Mac develop the idea and participated in the early stages of planning. Unlike Mac, Dennis got caught up in a different scheme, and so he didn't take any of the businesses money, even though Mac wouldn't have been able to succeed without Dennis's help in the early stages of the crime. Now, even on the assumption that both Mac and Dennis are causally determined, it seems to me that there is more reason to blame Mac than there is to blame Dennis. But notice that you can't explain the difference here in terms of Dennis's innocence, since his agency was also implicated in the embezzlement. This means that we need to appeal to the fact that Mac knowingly and with some measure of control brought about the embezzlement to explain the existence of those reasons to which we must appeal in order to make sense of the judgment that there is more reason to blame Mac than there is to blame Dennis.

Matt, I didn't see your comment when I was replying to Ben. See my response to him (the comment at 9:38 am).

Hi Eddy,

Thanks for the kind words. My time at GSU and your mentorship were certainly important for any success (i.e., good luck) that I've had so far.

As to your question, I guess I'd agree that the consequentialist considerations will be a wash only in rare, fairly exceptional cases. But it's pretty rare that executing an innocent person will bring about the best consequences, and yet many take this rather far-fetched possibility to be a decisive reason to reject act-utilitarianism. Moreover, it's even rarer (assuming certain religious doctrines aren't true) for agents satisfying compatibilist conditions on FW/MR to be subject to original design of the sort Mele describes in the Zygote Argument, and yet that argument is plausibly thought to show something that compatibilists are committed to. So I guess the goal of the argument is to show that incompatibilism has normative commitments that (I think) should strike us as very odd. Of course, they can bite the bullet, but it's worth getting clear on the bullets one must bite.

Justin: I agree with Matt; this is a really fun post.

This particular incompatibilist gets off the train at step one. You say, "For plausibly, Mac and Charlie’s co-workers still have less reason to blame Charlie than they do Mac." This doesn't strike me as especially plausible. Since they've no reason to blame either person, I'd say they don't have less reason to blame Charlie. Nor do I think I'm just pounding my fist on the table here. Suppose Jones, through no fault of his own, doesn't understand that his misogynistic comments are offensive to his female co-workers and that making such comments is morally wrong. To him, it's all just fun and games. Now compare Jones with Smith, who makes no sexist remarks, ever. Do we have less reason to blame Smith than Jones? I think not. While Jones has acted badly, his blameless ignorance of the relevant moral and social norms provides him with an excuse. Smith hasn't acted badly and so needs no excuse. Neither is deserving of blame, though for very different reasons. Now, if all that sounds plausible, as it does to me, why can't we plausibly say similar things about Mac and Charlie?

Why might it seem plausible that there is less reason to blame Charlie than Mac. Here's one suggestion. We may have reason to be critical of Mac and his behavior, since he has acted wrongly and his behavior demonstrates a lack of due regard for others. His actions may be attributable to him as the basis for moral criticism, but, in my idiolect at least, moral criticism is not the same as blame. Or, if you like, the sort of detached moral criticism of which Mac and his behavior are worthy is not the sort of blame that I think is incompatible with determinism.

This is a cool argument. I take the general challenge to the incompatibilist to be that they can't make certain distinctions between people and (the ways we have reason to treat them) that it seems we ought to make, and that seems like a good challenge. It's good not just as an argument against incompatibilism, but as a challenge to encourage the incompatibilist to say more about how they would conceive of human relationships if determinism were true.

I like Justin Capes's analogy with nonculpable ignorance, but that got me wondering: is excused wrongdoing really the same (in terms of how we have reason to treat the wrongdoer) as no wrongdoing? It certainly seems like I would have good reason to distance myself from the misogynistic Jones in various ways, even if I thought his behavior was the result of nonculpable ignorance. An excuse doesn't "erase" the wrongdoing, right?


My comment was like Ben's, true, but it didn't rest on innocence. So, Mac embezzled, and no on else did. So that's why it can seem like there's more reason to blame Mac than any non-embezzler for embezzling. But this is just an appearance, resting on the fact that only Mac is in the ballpark for being responsible for embezzling since he's the only embezzler on the scene. Or so the incompatibilist might argue, no?

Hi Justin,

Let me see if I can push you to stay on the train a bit longer (all the way to the station is probably too ambitious but hopefully past step one).

First, suppose that Frank, a demi-urge who passes the ages posing as a regular guy, is Mac and Charlie's boss. And suppose that he finds out about Mac's embezzlement and gets really upset by it. Now, if he says to you, "Justin, I'm so angry right now that I must punish someone, but I'll let you pick whether it's Mac or Charlie who feels my wrath." I don't think the right response is to say, "well, since causal determinism is true, neither of them deserve to be punished, so if you absolutely insist on punishing one of them, then (assuming the total good produced by punishing either Mac or Charlie is the same) I'm indifferent to to who you choose--just flip a coin Frank." A better response, it seems to me at least, is to say that because Mac knew what he was doing, it wasn't an accident, he exhibited some measure of control over his behavior, etc., Frank has more reason to punish him.

More generally, I think what you say suggests that you might reject a very plausible principle concerning desert: fault forfeits first (FFF). [1] Suppose that B deliberately harms A but through incompetence also harms himself. And further suppose that we can provide life saving aid to only one of the two injured parties. The natural thought (that undergirds FFF) in this case is that we have reason to provide A with aid and that we have reason to not to give B the aid. However, if causal determinism is true, it looks like on your view, B would be relevantly similar to Jones, and so independently of consequentialist considerations, we'd have no reason to prefer aiding A to aiding B. But this sort of indifference looks morally problematic.

[1] Shelly Kagan, *Desert*, (2012). What I say here relies on Kagan's discussion of FFF.

Hi Justin,

Thanks for your reply. I think I agree with you when you say that "we need to appeal to the fact that Mac knowingly and with some measure of control brought about the embezzlement to explain the existence of those reasons to which we must appeal in order to make sense of the judgment that there is more reason to blame Mac than there is to blame Dennis."

But I am not sure that I agree with they way you put things in the original post:

"S deserves to be praised/blamed for x-ing if and only if the fact that S x’d (knowingly and with some measure of control) is itself a reason to praise/blame S for x-ing independently of consequentialist considerations. But given that the fact that Mac knowingly and with some measure of control [embezzled the money] is a reason to blame him independently of consequentialist considerations, it would apparently follow that he deserves to be blamed for embezzling the money, even though he was causally determined to do so. Incompatibilism thus appears to be false."

One difference between these two claim is this: making sense of the judgment that there is more reason to blame Mac than Charlie (or Dennis) need not appeal to a *positive reason to blame someone*. Rather, and this is what I was suggesting in my earlier comment, it may appeal only to *negative reasons not to blame someone*. This was supposed to be the force of the indebtedness analogy--you have more money than me because you are less in the hole, but this does not mean that you have any money. (Causal determination might be a negative reason not to blame someone, but it does not seem to rule out compounding those reasons. I might be causally determined and also not involved in the wrongdoing. These may be independent reasons not to blame me.)

If I am right about the possibility of making sense of the comparative reasons claim without appeal to reasons to blame (as opposed to reasons not to blame), then the claim that we have more reason to blame Mac than Charlie does not support the claim that we have any reason to blame Mac. So it also does not support the claim that causal determinism is consistent with reasons to blame. So it does not support an argument against incompatibilism.

This worry leaves in place the nice insight that Mac's involvement explains the comparative reasons claim, and it seems to me to be compatible with what you say about the more detailed case involving Dennis. (We have fewer reasons not to blame Mac than reasons not to blame Dennis, and this has to do with Mac's involvement in the act of embezzlement. But we have fewer reasons not to blame Dennis than reasons not to blame Charlie because of Dennis' involvement in the planning of the scheme.)

Neal, I really like that way of framing things. I'm also sympathetic to what you say about non-culpable wrongdoing in the case of Jones and about excuses more generally: they don't "erase" transgression.

I think I'm missing something because what you say sounds pretty friendly to me. I guess the incompatibilist could insist that because Mac's the only agent who might be in the ballpark of being responsible that it *appears* that there is more reason to blame him than there is to blame Charlie, but that this appearance is mistaken because causal determinism rules out the existence of such reasons. But I'm not sure how this addresses (as Ben put it) the Schroeder-style move in the argument, which are meant to show that our judgments about so-called negative reasons existentials are unreliable.

Thanks for pushing me on this. It's helpful. (This last point is actually true for everyone!)

Can you say a bit more about what a negative reason not to blame is? I know what a reason not to blame would be, but I lose track of things with the first "negative." Maybe this is something basic, but I'm just not sure.

As for the worry, I suspect that it rests on an infelicity inherent to the way I talk of "more" or "less" reason, which are poorly-defined comparative claims. If I cleaned that up in more precise but (much) less intuitive ways (say, by talking in terms of set of pro tanto reasons of varying weights, etc.), the disanalogy between the point I'm making and the case of being more or less in debt would become clearer.

Maybe here's a way of making this point. Abstracting away from consequentialist considerations (broadly construed), there is no reason at all to blame Charlie. That is, no fact about Charlie favors (or, if you don't like "favors," fill in your preferred relation here) blaming Charlie. If this is similarly true of Mac, as the incompatibilist might claim, then it would follow that (among other things) we should prefer a state of affairs in which neither are blamed to a(n irrational) state of affairs in which both of them are blamed. We should also prefer a state of affairs in which at most one of them are blamed to a state of affairs in which both of them are blamed. But if there are no reasons to blame either of them, we should be indifferent as to which is blamed, since the resulting states of affairs would involve equal measures of irrationality. But we have reason not to be indifferent about which agemt is blamed. And I propose that the best explanation for the existence of that reason is the fact that Mac deliberately, knowingly, and with some measure of control brought about the wrongful action.

Not sure that gets me any closer to convincing you, but at least it doesn't rely on the not altogether clear notions of some, less, and more reason.

Hi Justin,

Sorry for the confusing terminology. I didn't put things well. Here's a try to clean the ideas up. (It's fun to think out loud with you about such an interesting argument.)

Let a reason to X be a consideration that *favors* X-ing. Let a reason not to X be a consideration that *disfavors* X-ing.

What I want to press you on is your claim that accounting for the claim that we have more reason to blame Mac than Charlie depends on accepting that there is a consideration in favor of blaming Mac. Rather, and this is supposed to be suggested by the debt analogy, it may be the case that we have more reason to blame Mac than to blame Charlie simply because there are more (or more weighty) considerations that disfavor blaming Charlie than considerations that disfavor blaming Mac.

This way of making sense of the proportionality claim would, if plausible, help the incompatibilist. For the incompatibilist might claim something like this: the absence of causal determination is an enabling condition on any consideration favoring blaming anyone. But the incompatibilist might leave open the possibility that the absence of causal determination is not an enabling condition on considerations disfavoring blaming anyone. (Even in causally deterministic worlds it counts against blaming someone that they were not involved in the relevant action.) So the incomaptibilist can make sense of the proportionality claims (say, by appealing to the fact that Mac was involved in the wrongdoing) without admitting that there is any reason to blame Mac.

That's the idea anyway. I hope it makes sense this time around.

Hi Justin, Thanks for another thought provoking post!! I like your argument a lot, and I am still working out how I feel about it. But for now, here is where I see a problem.

I am inclined to support incompatibilists who get off the train early, i.e. who deny this claim:

(B) Consequentialist reasons to blame being equal, there is more reason to blame Mac than Charlie

A standard incompatibilist move is to offer thought experiments that make the problematic implications of determinism vivid – as Randy Clarke says, they work like a stain to help us see something that was always there, but now much more clearly. There are lots of ways to do this, but to my mind Mele’s zygote case is really one of the best of the bunch.

So suppose we zygotize Mac and for that matter Charlie too: Demigod Diana selected and placed the right zygotes in the right wombs in just the right way so that with absolute certainty it is the case that Mac embezzles and Charlie does not. Given her setting things up the way she did, it is logically necessary, given the laws of the universe, that Mac and Charlie do what they do, i.e., just the things that Diana intended.

With this setup, does (B) strike you still as being obvious? To me it is not. So now we have a live debate about (B). But if that is what we should now debate, then the argument you are offering hasn’t shifted things in the compatibilist’s favor. We are still debating the very same thing that has always divided incompatibilists and compatibilists.


Neal's excellent point about excuses not erasing transgression, which I agree with wholeheartedly, actually helps me respond to your point about FFF. We might have reasons to alter our relationships with wrongdoers in significant ways, but these reasons needn't have anything to do with responsibility.

About Frank choosing between Mac and Charlie, if you've got to punish one of two people but neither deserves it, then you needn't flip a coin. Just pick the one you dislike most. That you dislike Mac the most gives you reason to choose him, given that you've got to choose someone.

That helps, but I'm still not sure that the sort of "disfavoring" relations will be present in the Mac and Dennis version of the story. Or, if they are present, that such relations will provide the best explanation of why there is more reason to blame Mac rather than Dennis. This suggests, then, that additionally to other facts that might be relevant, something about Mac himself provides a reason to blame.

Also, I think your way of framing incompatibilism is really interesting--viz., that the falsity of determinism is (in Dancy's sense) an enabler. That is, without it being the case that determinism doesn't obtain, facts about individuals agential involvement in their actions are not reasons to praise or blame. But notice, this means that the truth of causal determinism seems to be a very different kind of excuse than say, non-culpable ignorance.

Anyway, cool stuff that I need to think about more.

I agree with Alan that the assumption that Mac and Charlie can be given equal consequentialist weight for blaming is problematic. You mention a specific case where there are equal consequentialist reasons for blaming Mac and Charlie. But if we go for a consequentialist measure that considers the general weight of blaming persons like Mac versus blaming persons like Charlie, then a specific extra ordinary case is irrelevant. The motivation for blaming Mac is a consequentialist understanding that it is in the long run good to blame persons like Mac. The conclusion, I think, is that it will be impossible to find a general case with the same consequentialist weight for both of them and the argument fails.

Another way to argue it to turn the argument upside down. A skeptic would not think there can be any desert in a deterministic world (or in an indeterministic either) because we don't have any genuine control. But in a deterministic world we don't have any control of our agency either (compare Chandra's argument) so your solution to the problem is not possible (according to the skeptic). The conclusion is that the only remaining reason to blame Mac must be a consequentialist one.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan