Blog Coordinator

« Passing the Torch! | Main | A Boring Episode of *Always Sunny in Philadelphia* »



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

Hey Justin,

Nice suggestion. But on this view, how would desert differ say from fittingness or appropriateness or even "ought to be blamed"? As you write, if one accepts a buck-passing approach, all of these might be analyzed in the way you suggest (prima facie, at least)

Thanks John,

I'm not sure there would be a difference. No doubt, people characterize what it is to be morally responsible by appealing to apparently different relations that can obtain between agents, their actions, and morally significant responses like praise or blame. G. Strawson uses the notion of "making sense"; Wallace uses "appropriateness"; you and Ravizza (as you know) use "apt target"; Pereboom uses "basic desert"; "fittingness" and perhaps even "ought to be ..." could also be options ("ought to be blamed" strikes me as much too strong, however).

If y'all aren't just all talking past each other, it seems to me that you must all be getting at something like the same relation. And *that* is the relationship I'm interested in.

Suppose Bill Gates says he will give me a billion dollars if I blame Jimmy. But he will do so only if Jimmy x'd. Sure enough, Jimmy does said x. It seems then that the fact that Jimmy x'd is a reason for me to blame him. But surely the existence of *this* reason to blame doesn't imply Jimmy *deserves* to be blamed.

This is an instance of the "wrong kind of reason" problem and I don't think buckpassers can solve it.

My own view is that the order of explanation goes the other way: The fact (if it is a fact) that Jimmy deserves to blamed for his x'ing explains why there is a reason--of a particular sort--to blame him. The fact that Bill Gates will pay me if I blame Jimmy provides another kind of reason, one grounded in a different kind of fact.

Hi Justin,

I have a few potential worries about this approach, but it might be that there's nothing here to worry about at all.

First worry. Let's suppose that Agnes Vs and it's a clear case in which we'd want to blame Agnes. You say that what it is for Agnes to deserve blame is for Agnes' V-ing to be a pro tanto reason to blame her. Because it's a clear case in which we'd want to blame her, I was thinking of the case as one in which her V-ing manifests some sort of ill will or failure of proper concern. Now, whether her V-ing manifests ill will or failure of proper concern depends upon the agent's knowledge, beliefs, intentions, and motives and it would seem that the agent could have V'd having had different knowledge, different beliefs, different intentions, and different motives. If we vary these psychological states, we can vary the case so that in one situation Agnes Vs and is blameworthy for it even though there's another where she Vs and is excused. That makes me think that even when she deserves blame for having Vd, the fact that she Vd isn't going to be a good candidate for the reason to blame. If that's right, it looks like the account excuses too much. All we know about the Ving is that it was supposed to be a paradigmatic case of culpable action.

There's a second point that's potentially related to the first and that has to do with this idea of a pro tanto reason to blame. I'm sure some people will want to say that the Ving should include a specification of Agnes' mental states. I think this is a mistake, but let's just talk about thin and thick Vings and think of thin Vings as the deeds done without specification of motive, intention, knowledge, etc.

* Thick. If we think of Vings as thick Vings it seems that the specification of the Ving will build in all the stuff that matters to determining whether the agent's actions manifested ill will or not. If all that stuff is already packed in, why do we have 'pro tanto' on the right hand side of your identity claim?

* Thin. If we think of Vings as thin Vings, you might put the pro tanto to work, as it were, and try to combat my first objection as follows:

Look, CL, the first worry isn't a worrying worry because the fact that Agnes Vd is merely a pro tanto reason to blame her. It's not an all things considered reason to blame her. Whether there's such a reason or not depends upon further things, such as facts about Agnes' state of mind at the time of action. If she had the bad attitudes (badittudes?), we'd have an all things considered reason to blame her. If she didn't, we wouldn't have the all things considered reason to blame her.

I don't think this response will do, however, not if by 'pro tanto' we mean a reason that's something like Ross' notion of a prima facie reason. In the standard case where you have such a reason but it isn't an all things considered reason to Y, there's still some ground for regretting that you didn't Y. This residue or remainder seems to be missing in the relevant cases. Agnes, let's say, takes something of mine without asking. Let's now add that she had every reason to think that what was mine was actually hers so that her action doesn't manifest any sort of ill will at all. It seems that in this case there's not a pro tanto reason to blame that nevertheless fails to be an all things considered reason to blame. There's just no reason at all to blame here and there's nothing at all that would be a kind of residue that makes it regrettable that you don't blame.

Alright, those were the worries.

Hi Chandra,

Fair enough about wrong kinds of reasons, but I'm happy to add a rider of the sort Pereboom uses to characterize basic desert. The revised proposal would then be something like: What it is for an agent to deserve praise or blame is for it to be the case that the fact that she x'd is a reason to praise or blame her independently of consequentialist or contractualist considerations.

FWIW, given my own interests in figuring out the connection between basic desert and reasons (which I'll say more about in the next post), I'm happy with a weaker claim like:
An agent deserves to be praised or blamed for her action iff the fact that she performed the action is itself a reason to praise or blame her for that action independently of consequentialist or contractualist considerations. It seems like you could accept this weaker proposal since it's neutral with respect to the order of explanation.

Hi Clayton,

In the case of Agnes taking something of yours (even though she has every reason to think that it's hers), I agree that the fact that she took some of yours is not a reason to blame her (since there would be no residue left over from not blaming her). But this is presumably what we want to say about this case since Agnes doesn't deserve to be blamed for her accidental theft. If the original proposal commits me to thinking that the fact that she has taken something of mine is a reason to blame in such a case, then I'm happy to reject the original proposal. But I'm not quite seeing how it has this commitment (or is the Agnes case supposed to be the response that doesn't work?).

Also, I don't mean to propose that mental attitudes, knowledge, intentions, quality of will, etc. aren't relevant to whether you are blameworthy for your actions. Those things are relevant, in particular, because they put constraints on when the fact that you've acted badly is itself a reason to blame you. However, it's because the fact you've acted badly is a reason to blame you that you deserve to be blamed. At least, that's what I meant to claim with the original proposal (Chandra was right to point out that as stated, it needed fixing).

I'm not sure this completely answers your worries, though. So if I've missed the point, let me know.

Hi Justin,

Great to have you blogging!

I want to press on the same issues that Chandra raised.

About the wrong kind of reason worry: I suppose a rider of some sort might solve the issue, but I'm not sure that Pereboom's is sufficient. Pro tanto reasons are cheap, on many views, and it doesn't seem to me that the only ones we need to rule out are ones that would be picked out by contractualist reasoning or that bear on consequences. So I am inclined to think that the rider you would need to handle the full range of problem cases would end up being quite unwieldy--potentially uncodifiable. But I suppose I could be talked out of this.

As for the order of explanation, I was surprised about what you said at the end of your reply to Chandra. The original proposal seemed designed to explain desert in terms of pro tanto reason to blame. But then you gave a biconditional and said that it was neutral wrt order of explanation. But then it is not clearly the same kind of proposal as the original one. Rather, we are being told, in the second case, that there is a tight connection between desert and reason to blame. But we are no longer being given a b-p-type analysis of desert. And I thought this was a core part of what you were after in the original post. Am I misunderstanding what you were up to?

Interesting stuff.

Thanks Ben,

I agree pro tanto reasons are cheap. I guess I'd need to hear more before I could say whether this means that there is no way of reducing normatively significant concepts like value, obligation, desert, etc. to them.

As for the order of explanation stuff, it's true that what I said in response to Chandra gives up the b-p account of desert. But as I'll argue in the next post, the weaker claim is enough to serve as the basis for an argument against incompatibilism. So it gets me what I want (even though I'm still sympathetic to the idea that one can provide something like a b-p account of basic desert).

Hi Justin,

I'm not sure that the cheapness of pro tanto reasons makes it impossible to provide a reduction of desert, value, etc. to them. But I have my doubts about the general project of buck-passing accounts.

One doubt is the one already expressed: to get the reduction right, one would have to rule out wrong reasons. And this may require adding so many conditions to the analysans that it cannot be fully stated.

Perhaps, however, the fix is to plug in a reference to "reasons of the right kind." This makes things tidier, but I have doubts about it making for an informative analysis. We are left wondering what the right kinds of reasons are.

A general worry I have about buck-passing accounts is this: it just doesn't seem clear to me that the notion of a reason is any better understood than the notion of desert or value or what have you. To take an example, Scanlon introduces the b-p acct of value in opposition to a Moorean acct of value. This is, in effect, to trade one unanalyzable non-natural property for another. If the claim that values are such properties is unsatisfactory, then what does it help to analyze values in terms of the notion of a reason, where this is taken to be a non-natural relation, the favoring-relation, which we cannot say anything more informative about?

None of these considerations are conclusive. But they cast doubt, in my mind at least, on the attraction of buck-passing accounts. In a case like yours, where the buck-passing part of the account is not necessary for the intended payoff (i.e., stickin' it to the incompatibilist), why not just leave it out of the account? It seems to bring with more trouble than it is worth.

Hi Justin,

Thanks for your responses. Let me see if I can't put the point that worries me a bit differently.

The thesis: What it is for an agent to deserve to be praised or blamed for x-ing just is for the fact that an agent x’d to be a pro tanto reason to praise or blame.

Here's my worry about the thesis. I take it that the fact that Agnes took the book, say, is distinct from various facts about the properties that this action had, including facts about what motivated her to do what she did, what she knew at the time of action, what she believed about the situation, etc. (Not only are these facts distinct, but if we're thinking of the x-ing as something that could have been done for different reasons, the fact that an agent xs doesn't entail much of anything about the reasons that led the agent to x, what the agent knew at the time, etc.) I guess I can put the worry this way: once we distinguish the fact that Agnes took the book from the fact that, say, she took a book she knew belonged to someone else, I can't see any reason to think that the fact that she took the book is any sort of reason at all to blame her. I don't think it counts in favor of blaming her, shows that there's something good about blaming her, etc. The fact that she took the book, after all, tells us nothing about whether Agnes showed the book's owner the right kind of concern.

I share some of the worries that others raised above about priority and the wrong kind of reasons problem, but my worry is just a really simple one, which is that I think that the thesis is focusing on the wrong fact as the facts that would seem to give us reasons to blame have to include some facts about knowledge, belief, and motivating reasons. Without it, it seems that blame is cut off from the things that tell us whether an action showed ill will.

Wasn't it just a couple of years ago that you were attracted to a b-p account of value and I was skeptical? I have a weirdly vivid memory of us arguing about it (but taking the opposite sides) over a pitcher of beer.

As to trading one non-natural property for another, I'm not meaning to do that, since here I want to be agnostic about whether reasons are non-natural entities (as you know, though, I'm a naturalist about reasons, but this is consistent with thinking that in the normative domain, reasons are the most basic entities). Of course, you're right to say that perhaps the proposal is more trouble than it's worth, but that's why I retreated in my reply to Chandra.


Thanks! This helps me to see the issue better, and fwiw, it's a worry that I've had as well. In the version of this stuff that I'm working on, it's not just the fact that the agent performed the action, it's the fact that she knowingly and with some measure of control performed the action that underwrites desert attributions. That seems more in line with how you're thinking. [1]

Now, really, what we should say (if we're trying to reduce desert) is that desert isn't reducible any one reason to praise/blame but to a whole set. So maybe something more like:

What it is for S to deserve praise/blame for x-ing just is for it to be the case that the set of facts that are relevant to S's x-ing in the particular way S actually x's constitute reasons to praise/blame S for x-ing independently of their connection to consequentialist or contractualist considerations.

This, of course, probably does little to assuage 'wrong kind of reasons' worries, but hopefully it better specifies which facts about an agent might be relevant to her deserving praise/blame.

[1] That said, it does seem to me that even in the case you give, the fact that Agnes takes the book is itself a reason to blame, but as I said to Ben, I extremely promiscuous when it comes to reasons existentials.

Oh, my friend, the beer must have gotten to yer head! I am not now, and was not then, a friend of passin' that buck. Values fundamentalism is the name of this game!

Wouldn't have been the first time that happened.

Hi Justin,

Interesting question. I, too, have tried my hand at getting clearer on the desert relation, so I applaud the effort!

One question about your proposal. It seems the prospects for reducing desert relation to reasons will depend somewhat on how we're conceiving of the relata.

For instance, if one took blame to be like an action, then reasons to blame would be reasons for action. That may lead to a more promiscuous set of pro tanto reasons in favor of blaming (as, arguably, there are no reasons of the wrong kind for actions).

If blame is more like an attitude, then the issues will be slightly different, and the sort of attitude it is will affect how we take claims about what count as reasons.

Another way to put the same point has to do with understanding what the blameworthy deserve. Or, as I prefer it, what they are worthy of. Is it a form of treatment (which looks action-like)? A directed attitude? A reactive attitude? Which of these one picks, it seems to me, will affect the trajectory of the resulting account.

I wish to stand up for those of us who are at an age where our brains can be entirely addlepated without the aid of beer!

Here's my residual worry: some have certainly thought that there is some sort of (important) difference between accounts that employ "desert", "fittingness", "appropriateness", etc. (These may be in accounts of moral responsibility or blame/blameworthiness.) But if these all get analyzed in the same way, there would not be any such difference. Is this what you are suggesting? Also, I would have thought that they are different notions; so, desert might just be an element of "appropriateness"--and if so wouldn't a single analysis be problematic?


That's a fair question, and in general I think we need to attend to things like blame in order to help understand moral responsibility, so I'm sympathetic to your concern.

But couldn't some facts be reasons to both act in certain ways towards people and reasons to adopt certain attitudes towards them as well? Like if you bought me a beer, I'd take that fact to constitute a reason for me to say, "thank you." But I'd also take that fact as a reason to come to believe that that Matt King guy's pretty generous and as a reason to feel grateful. I think something like this can go on in cases of blame as well, so I'm not sure that I have to settle the question of what blame is here (maybe later this month).

Plus, isn't it a strength of the proposal that we can plug in a number of different relata? People can deserve ire (an attitude) and also an obscene gesture (an action), but we wouldn't want to say that people stand in a different relation to what the deserve in these cases.

It's certainly possible that the "desert," "fittingness," "appropriateness," relations that folks appeal to as a way of characterizing what it is to be morally responsible are intentionally picking out different relationships. I suspect that *at least* a little bit of what's going on, though, is that it's really, really, hard to precisely specify the relation in question (though we all "know it if we see it"), and so different folks have all tried to find ways of talking about *that* relationship, whatever it is.

This is, of course, consistent with, e.g., you and Pereboom having identified different kinds of moral responsibility (you're concerned with being an apt target of the reactive attitudes, he's concerned with basic desert). But even in cases where this is going on, at my most ambitious, I suspect (i.e., am tentatively hopeful) that we'll be able to reduce both of these relations to sets of reasons, even if the sets of reasons are subtly different. So you're right that it would be problematic to elide distinctions that are really there, but maybe I don't have to (even if I did in the original post).


I like your impulse here; I had talked about "aptness" or "appropriateness", and I'm still not convinced that there is a huge difference between these notions and "desert" or "fittingness". So it is interesting to see how far one can get with your approach.

I was just wondering about this issue; I think others believe that a lot is at stake here.


My thought was that there will be wrong kinds of reasons to hold the attitude of blame toward an agent, but no wrong kinds of reasons to action-blame them. If that's right, then there could be lots of reasons to action-blame them that aren't desert-related.

Also, I worry that your proposal fails to illuminate anything about the nature of desert. Consider this proposal: What it is for it to be good that an agent be praised or blamed for x-ing just is for the fact that an agent x’d to be a pro tanto reason to praise or blame.

Since reasons come cheap, 'the fact that an agent x'd' could easily be a pro tanto reason to praise or blame them for a variety of, ahem, reasons.

Matt, I think that the case Chandra brings up in the third comment might involve a case of wrong kinds of reasons even for action-blame. Also, lots of what we find worrisome about hypocritical action-blame (e.g., Newt Gingrich's public expressions of opprobrium towards Clinton during the Lewinsky scandal) is that he's blaming Clinton for political reasons and not because he's concerned about the "sanctity of marriage," as he presented himself. I'm (to some degree) losing my grip on how this connects to the initial proposal, but it's interesting stuff.

As to whether the proposal (if true) illuminates anything about the nature of desert... Well, it certainly purports to, since is purports to provide a reductive base for desert claims in the same way buck-passers about value or reductionists about deontic categories (like obligation) seek to reduce these moral properties to reasons. Now you might worry that *this* doesn't actually illuminate anything, and perhaps that's right. But it seems to me that just as we'd do better to think that moral obligation reduces to sets of reasons, we'd do better to think that desert similarly reduces to sets of reasons.

FWIW, the weaker proposal I introduce in response to Chandra and invoke in the subsequent post isn't meant to be illuminating per se--just spelling out a conceptual connection between desert and reasons to praise/blame that might make trouble for the incompatibilists.

Hi Justin,

I am coming in late and have not read all the above so sorry if this is redundant - it is basically just following up on John Fischer's comment early in the thread.

It seems to me that deserving talk has a natural home in our philosophic history and one main idea here is that if you are morally bad then you deserve to suffer and not be happy. In this context, it seems to me that deserving involves there being compelling reasons, not just sufficient reasons, for God to see to it that you suffer and are not happy. Maybe this speaks in favor of adding the idea of compelling reasons to your analysis, and for thinking that sufficient reasons are what are in play when people make claims about appropriateness?

This seems to fit, for example, Wallace's appeal to appropriateness when he responds to the Martin Luther King/Ghandi case. If I remember correctly, he holds that they could admit that resentment is appropriate, but not demanded by reason. In my terms he suggests a view on which there is sufficient but not compelling reason for resentment, and we could say he therefore thinks resentment is appropriate but not deserved.

I bet there has to be some interesting theological discussion of this given that on some views God gracefully foregives those who are appropriate targets for punishment. It would seem odd, to my ear anyway, to say that God fails to punish those who deserve it!

Another idea would be to say that your analysis is right but that assertions about desert imply compelling reasons in a way that assertions about appropriateness do not.

Any way, cool topic!


That's a potentially helpful way of framing what's at stake when folks take care to distinguish between "appropriateness" and "desert," but it's not altogether clear to me that I agree that desert should be specified in terms of compelling reasons.

In a case of hypocritical blame, it seems like the right thing to say is that, e.g., Bill Clinton deserved to be blamed for his affair, but that because Newt Gingrich was in the middle of his own affair, he had reason to refrain from blaming. Indeed, in the case of Newt, I suspect that his reason to refrain from blaming in the circumstances was so weighty that it was a compelling reason (to borrow your way of putting things). This suggests that desert isn't connected to compelling reasons. Or would it just be God for whom the reasons of desert are compelling?

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan