Blog Coordinator

« Desiderata for a Theory of Responsibility | Main | From Russia with Love »

09/06/2014

Comments

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

Although I grant that an similarity in conditions can give us at least prima facie grounds for suspecting praiseworthiness and blameworthiness are related, I have two worries:

1 - What of those (perhaps like Dana Nelkin) who think that responsibility is required for both praiseworthiness and blameworthiness but who also think that the conditions required for responsibility for praise differ from the conditions for responsibility for blame? You aren't, I take it, here denying accounts like that. But there's something attractive about Dana's argument, and, at the very least, I wouldn't want to say she was talking about something other than responsibility just because she buys an asymmetry.

2 - The examples you offer all seem suspiciously alike. They would all be examples where I'd be inclined to say, of the blame cases, there is no wrongdoing (though a final verdict would require more development), and I'd probably say something similar for the praise cases. Yes, the muscle spasm resulted in harm or benefit, and yes, had I done 'the same thing' knowingly (or recklessly, or whatever), it'd have been wrongdoing or right doing. But, given my spasm or ignorance or whatever, I'm just inclined to deny that there's action in the relevant sense.

If you think that responsibility is required to connect action to praise or blame, these cases don't show parallelism for responsibility, they show parallelism for action.

(Also, thanks so much for these posts! I really enjoyed your `Traction with Tracing' -- and these posts have been great too.)

Hi Matt,

I too lean against what you are calling the "practice-based" approach (what others have called the "normative" approach, I think your name is better as “normative” is such a broad term). I like your Symmetry Argument quite a lot. But it seems to work only against Wallace's specific version of a practice-based approach, which emphasizes expectations and obligations.

There is another version of the practice-based approach that emphasizes the *reactive emotions*; expressions of these emotions are, on this view, the crux of the relevant practices. This view says that to understand *being responsible* for something, you must first understand the conditions under which these reactive emotions are apt or fitting. I think this view can treat blame and praise symmetrically. This is because the reactive emotions aren’t restricted to blaming; there are praise-related emotions as well. So these views would say that the cases you put forward involving accidents and spasms are ones in which the conditions aren’t met for the fittingness of the relevant reactive emotions—praise and blame emotions alike.

Hi Matt,

I too lean against what you are calling the "practice-based" approach (what others have called the "normative" approach, I think your name is better as “normative” is such a broad term). I like your Symmetry Argument quite a lot. But it seems to work only against Wallace's specific version of a practice-based approach, which emphasizes expectations and obligations.

There is another version of the practice-based approach that emphasizes the *reactive emotions*; expressions of these emotions are, on this view, the crux of the relevant practices. This view says that to understand *being responsible* for something, you must first understand the conditions under which these reactive emotions are apt or fitting. I think this view can treat blame and praise symmetrically. This is because the reactive emotions aren’t restricted to blaming; there are praise-related emotions as well. So these views would say that the cases you put forward involving accidents and spasms are ones in which the conditions aren’t met for the fittingness of the relevant reactive emotions—praise and blame emotions alike.

Just to be clear, I don’t support this reactive emotions view myself. We could provide an account of a strike in baseball in terms of the conditions under which certain responses from the umpire are fitting; this approach would certainly be extensionally adequate. But the approach is wrong-headed because there is a well-behaved property that the umpire is tracking in virtue of which his or her responses are fitting. Until someone shows how reactive emotions are fitting without invoking their tracking some well-behaved property (such as quality of will), I am going to remain deeply skeptical of the practice-based “inverted” approach to the metaphysics of responsibility. But despite these flaws, and returning to the topic of this thread, these reactive emotions views can apparently handle your Symmetry argument I think.

Thanks, Craig. I think BW and PW are symmetrical, and the proposal suggests that it is a basic responsibility relation that is shared between them. Technically, this proposal is consistent with accounts, like Nelkin's or Wolf's, in which an ability to do otherwise is required for blame only. That's because the features I'm pointing to could affect a shared component of each, but there could still be other components to each, which are left unaffected. So if I were to argue for full symmetry between BW and PW, I'd have to tackle those asymmetrical view directly. As it stands, I do think they're false -- but I don't take the evidence for symmetry I'm giving here to show that. I just don't think the arguments for asymmetry are persuasive.

I'm not sure I fully appreciate your second claim. I agree that in all the cases there isn't wrongdoing. But I think that wrongdoing often (if not always) requires responsibility (at least for action).

And thanks for your kind words about my tracing paper. I'm glad you liked (reading) it!

(1) Here's an available view: moral responsibility is fundamentally about the appropriateness of the reactive attitudes. The appropriateness condition in the case of blame is this (maybe others, but never mind): the agent's action manifested a bad quality of will. The appropriateness condition in the case of praise is this: the agent's action manifested a good quality of will. The thread running through your cases is that the following conditions each undermines the claim that the agent's bringing about C amounted to the manifestation of a good or bad quality of will: C is an unintended and unforeseen consequence of ones actions, or C was caused by spazzing out, or C only happened because the agents has a mistaken belief about the facts.

That is not a responsibility-first view, but it's a plausible account of your examples.

(2) McX says: "to be MR is to be BW or PW. And to be BW (PW) is to be worthy of blame (praise). And to be worthy of blame (praise) is for such-and-such response to be warranted / fitting / deserved / true / etc." McX's view has the explanatory priority or dependence going the other way. (I actually think there is no metaphysical dependence here, just a stipulation about the meaning of 'MR', but....). A major virtue of his view is that I understand it or at least I'm very close to being in the neighborhood of understanding what McX is talking about. (I might still have a lot of questions: what exactly does McX have in mind by 'resentment' or 'pride' or 'fittingness' or whatever other terms he uses to explain his account.)

This contrasts for me with your claim: "A prime candidate for what they share is that one must be responsible in order to be either BW or PW." I do not think I have much independent grip on the responsibility notion at play here in the way that I do with the notion of blame, pride, etc. So when people talk this way, I feel just completely lost at sea.

Responsibility is surely not primitive and so there must be some account of what it is. Presumably, since we're going responsibility-first, responses-second, it will be something like: for S to be responsible for X is to for S to stand in such-and-such action-theoretic relation to her behavior. All I know about what that action-theoretic relation is supposed to be is this: the agent's in your examples don't bear it to the relevant actions and the fact that they don't is part of what explains their lack of PW/BW. So, if I had the above quality of will view for the appropriateness conditions of PW/BW, then I would just say that that's what responsibility is in your sense, i.e. that to be responsible in your sense is for the agent's action to manifest a good or bad quality will. But now the responsibility-first versus responses-first dispute seems totally non-substantive.

Hi Chandra - thanks for the comment!

You're right that there are ways of trying to respond on behalf of the practice-based theorist. And I like you're example of the umpire and thinking about our responses "tracking" certain world features.

Let me say why I think the reactive attitude views (RA views) can only go so far in addressing the symmetry. First, we should note that being able to give an explanation of the symmetry isn't really enough, we want a good explanation. So, suppose the RA theorist says, "Look, the emotions constitutive of blame and praise would all be undermined in these cases. So my view gets the right result!"

Now, it does get the right result. But it if we're not appealing to something that all the cases share, the explanation here begins to look like a coincidence. Why should the various emotional responses all be rendered inappropriate across the board? Here's an attractive conjecture: that they're responsive to a prior fact of responsibility. But that's not an available option for the practice-based model.

Second, things get worse of the practice-based account, because the symmetry isn't limited to cases of *moral* blame and praise. It's also present for athletic, artistic, and academic blame and praise. But here we won't have the same reactive attitudes at play, if any at all. So the coincidence grows. And given that there is an enormous multitude of non-moral, but normative domains, which I think will operate the same way, that's a big set of data to rule out as being irrelevant to (moral) responsibility.

(It's worth noting, however, that even a non-"inverted" account on which the RAs are dependent on something like a quality of will face the challenge, since these non-moral domains plausibly won't involve the expressions of qualities of will, at least not as they are usually understood. So I also think that any moralized account of the responsibility relation faces something like this challenge.)

Sorry, my previous post duplicates some of what Chandra already said.

On Chandra's 10:40 AM comment: I agree with you about strikes. To be a strike, perhaps, is just to cross the plate in such-and-such a way. But the "reactive" approach to the theory of strikes wouldn't be inverting anything if we started out wanting to explain correctly calling a strike, instead of strikes. I understand the reactive-attitude MR views to just start by asking a question like that: what are the conditions under which B/P are appropriate (oh, and by the way, appropriateness of B/P is just what I mean by 'MR')?

Thanks, Nate, for pushing me on this. (I wrote this up before your second comment...but I think I address both.)

I'm sympathetic to the idea that the basic responsibility relation isn't full of intuitive content like blame or pride. I think it's probably a relatively "thin" notion in that respect. (I suppose it's fair here to report that whenever anyone mentions "qualities of will" I feel lost -- I don't think anyone has given a satisfactory independent characterization of what the heck they are.)

You may be right that, at the end of the day, we can substitute talk of the responsibility relation with something else. For me, for S to be responsible for x is for A to be evaluable by x's lights. This is different from thinking that the responsibility relation just is the manifesting of good or bad will. So I think there is a disagreement here.

But I should also say that one who says that to be responsible is to express a quality of will does seem to be proposing a prior-fact model. The main purpose of the symmetry argument is to argue for a prior-fact model, so I'd be happy with that result. We can argue over which prior-fact account is better later.

As for that disagreement, however, I'll just echo what I said in response to Chandra. I think we have evidence for the responsibility relation in a broader range of cases then the quality of will theorist can make sense of. So, those views are left treating the symmetry as a coincidence.

Hi Matt,

Thanks. I certainly agree with you that more needs to be said about it is to possess a good / bad quality of will. My use of 'quality of will' was intended to be a gesture towards more worked out accounts, like Arpaly's (which I think is a pretty clear view).

On the non-moral cases:

There is some sense in which sports fans hold certain sorts of failures against players or in which teachers hold academic failures against their students. Sometimes these attitudes are appropriate (Kobe misses a dunk and the Lakers lose; the student doesn't try and so fails the exam); but sometimes they aren't (Kobe misses the dunk because he had a hallucination that the basket was 3 feet to the right; the student went temporarily blind). Your point (I think) is: we need to explain the fact (i) that the blindness and the hallucination are blame-appropriateness-undermining conditions and (ii) that the explanation here should be parallel to the case of moral blame. And, you claim, the quality of will theorist has no story to tell. But here is a story: missed dunks are (non-morally) blameworthy when they manifest bad-player-qualities, like lack of practice or concentration. The hallucination condition shows that Kobe's miss wasn't due to his lacking these qualities. Failed-exams are (non-morally) blameworthy when they manifest bad-student-qualities, like insufficient concern to do ones best. Temporary blindness shows that the failed exam wasn't due to insufficient concern. In short, we address your non-moral cases by embracing the non-moral qualities of will.

Nate,

Good. This is helpful. The space between the two views is radically shrinking. I'm going to be brief here, because I don't want the thread to become about which of these two views is correct. That's because they both are prior-fact views, as I understand it. There is some prior-fact on which the appropriateness of our practices depends. Do you think that's right?

Since that was the principal aim of the symmetry argument, I'll be pleased with that result.

While I have my reasons for being pessimistic about the quality of will approach succeeding along these non-moral lines, all I want to make clear here is that I don't think a quality of will view *cannot* explain the symmetry. If developed along prior-fact lines (which, e.g., Wallace doesn't do), it can provide a kind of explanation. What I think is that such explanations will still be problematic. But that conclusion relies on a critique largely independent of the symmetry stuff.

Thanks for this post, Matt. I have a question for you regarding how you conceive of praise in these practice-based theories. So, I take your argument to be as follows:

1) Mistakes, spasms, and accidents can undermine both praise and blame
2) Practice-based theories can't explain why they undermine praise because praise does not amount to "holding responsible"
3) Therefore, prior-fact theories are preferable because they can accommodate praise-undermining phenomena in a way that practice-based theories can't.

If that's right, then I just don't understand premise 2. You say that we simply don't hold PW folks responsible and *certainly* not in the sense of holding them to some standard or expectation. So my question is this: why does praising someone not amount to holding her to some standard or expectation? As I see it, when I am holding you to a standard, then by your actions or attitudes you can a) fail to meet the standard, b) meet the standard, or c) exceed the standard. If (a) then I'll blame you, if (b) then I'll respond with neither blame nor praise, and if (c) then I'll praise you. So, why, on this way of thinking about it, am I holding you to a standard in (a) and (b) but not in (c)?

Thanks, Nathan.

This is one of those occasions where I didn't say enough since I was trying to be brief. So, to elaborate...

Part of my thought here is just a linguistic intuition. We don't use the phrase "holding responsible" to talk about praise like we do for blame. So we say things like, "We will seek out and hold responsible the perpetrators of this terrible deed;" or, "I hold you responsible for the divorce!"

But we don't really say, "I hold you responsible for throwing me such a nice birthday party" or "We will hold responsible those firefighters that risked so much to save so many". So talk of 'holding responsible' often just looks like talk of blame.

Now, that's just a bit of linguistic data. Maybe it's just that we tend to do more blaming than praising, or something like that. I'm inclined to think it actually points to a way in which our practices of praising are different -- but there's room for disagreement here.

You're right, of course, that in setting standards for conduct, one can fall short, meet, or exceed the standard. So why aren't we holding someone to the standard when we respond to their exceeding it? You might have in mind a kind of metaphor, where the standard is a yardstick, and we hold it up to see where you fall on it. So we're holding you to the standard.

I agree that we can do something like this. I just don't think it's what practice-based accounts have in mind when they talk about our practices of holding responsible. Here are some comments on that.

First, most such theorists would want to distinguish between 'holding responsible' and 'holding to be responsible'. So the former is a distinctive stance toward the agent, whereas the latter is just regarding them as either a responsible agent generally or responsible for the particular conduct in question. So it's consistent with my view to say we hold them to a standard by taking them to be responsible for x and then comparing x to some standard. (That's roughly what I think, in fact. But that appeals to an explanatorily prior fact of responsibility.)

Second, it is natural to slide between talk of holding responsible and talk of holding accountable. But to hold accountable suggests seeking an account or justification for one's action, it implies a negative valence (such that we expect a justification), and so it seems most naturally at home with blame. It is odd to say that we hold the praiseworthy accountable.

Third, and perhaps more importantly, in the way these views are often developed, blaming someone is constitutive of holding them responsible. So the blaming amounts to holding them to a standard. I think there's a natural picture here, where it's part of the blaming that we're invoking an expectation, obligation, or other standard that's gone unmet. I don't think the picture naturally extends to praise. In praising someone, it seems odd to me to suppose it is constitutive of that exercise that we're invoking an expectation, obligation, or other standard.

Finally, and relatedly, the measuring metaphor is only consistent with a certain picture of what the standards look like. Many of our moral obligations just aren't the sort of thing you can exceed. An obligation against intentional killing, for instance, isn't the sort of thing you can exceed. So the relevant standards for blame wouldn't be those for praise, so you couldn't fall or exceed the very same standard.

OK -- apologies for the lengthy reply!

Hi Matt and Chandra,

I am having trouble seeing how the umpire analogy (which you both seem to like) is supposed to work and also how it is supposed to demonstrate a difficulty for the practice-based approach. Maybe one or both of you can help me out.

Begin with the analogy. I take it that the ump is supposed to be like the blamer/praiser and that the pitcher (or batter?) the blamee/praisee. The ump's call on a particular pitch is supposed to be like a particular response to a particular action.

Is this right? I'll assume it is in order to consider the way in which this analogy is supposed to help us see a difficulty for the practice-based model. (But if it isn't, then please correct me and then also let me know how it bears on what I am about to go on to say.)

I take it that the prior-fact model, given this analogy, is supposed to appeal to the prior fact of there whether or not the pitch was in the strike zone in order to explain the correctness of a given call by the ump. But the practice-based model is not supposed to be able to appeal to this fact as a *prior* fact. That is, I take it that the charge against the practice-based model, in terms of the analogy, is that there is no fact of the matter whether the pitch was a ball or a strike prior to the ump's call. The practice-based account, supposedly, has to say that the pitch is a strike because the ump called it that way; as opposed to the prior fact model, which gets to say that it is a strike because it satisfied some fact prior to his call, and the appropriateness of his call depends on this prior fact--namely, that it was in the strike zone or not.

Now, I don't see that the analogy (assuming I've got it right) works for at least two reasons. First, as any baseball fan knows, lamentably, the prior fact of whether the pitch was in the strike zone does not matter at all once the ump has made a call. If he calls a strike, it is thereby a strike; and when we see that it was outside the strike zone by a mile we groan and take the bad call. The ump's call has the illocutionary force of a declaration. So the case does not make for a good analogy in support of the prior fact model. The prior fact theorist, I take it, should accept that there is a fact of the matter established by the ump's call (namely, that a given pitch a strike or a ball), but claims that there is no fact of the matter established by the blamer's/praiser's response. Blame/praise does not have declarative force--at least not given the prior fact model. So I don't think the analogy serves its intended purpose very well.

There is a second reason to think this. (And my apologies for going on so long.) The ump's call matters at all only because of the constitutive rules of the game of baseball. One is tempted to say that the practice of playing baseball is prior to all of the facts at issue: the role of ump, the strike zone, the role of pitcher, the fact of throwing a pitch, the ump's call, etc. So I just don't see that the prior fact of a strike zone is anything like it would need to be to cause trouble for the practice-based model. The fact of there being a strike zone depends on a practice! And the ump's call and role also depend on the same practice. The idea, I think, is supposed to be that the strike zone is analogous to the prior fact of quality of will. And this latter fact is supposed to be prior to our practice of holding responsible. But then the analogy breaks down, because the strike zone is not prior to the practice of playing baseball.

But perhaps I am misunderstanding things. I could probably use some help.

(For what it's worth, I actually think that the analogy goes some way to making good sense of how the practice-based theorist can appeal to some fact that is, first, independent of the reaction (strike zone, quality of will) but, second, practice-depednent (game of baseball, practice of holding responsible.)

I was just going to make exactly the same argument as Nathan. Having read Matt's response above, I still think it holds.

Matt's first argument doesn't seem to respond at all, it's just a general argument that morality exists prior to responsibility judgments.

"Second, it is natural to slide between talk of holding responsible and talk of holding accountable. But to hold accountable suggests seeking an account or justification for one's action, it implies a negative valence..."

I dispute that. Nobel lectures. Tales of heroism. In the Confucian tradition, being a superior man means often being asked how you do it. The idea that we are only/more often asked to "account for" bad behaviour just doesn't seem to me to be obviously empirically true.

"blaming that we're invoking an expectation, obligation, or other standard that's gone unmet. I don't think the picture naturally extends to praise. In praising someone, it seems odd to me to suppose it is constitutive of that exercise that we're invoking an expectation, obligation, or other standard."

I can't see why you think this is odd at all. Some of our linguistic forms of praise explicitly state or refer to the standards: "above and beyond the call of duty," "went the extra mile," "be the bigger person"...

"An obligation against intentional killing, for instance, isn't the sort of thing you can exceed."

Again, I can think of many examples off the top. Peter Singer's argument for giving more to charity. Intervening in killings. Conscientious objection. Being a vegetarian. Any of these *could* be viewed as an excessive discharging of the duty not to kill.

In general, I think praiseworthiness is very interesting, though, so I'm glad you're working on it.

Thanks, Ben.

I won't speak for Chandra -- what I liked about the analogy was just the idea of a judge trying to "track" features of the world. I didn't want to hang anything about my argument on the analogy's substantive content, however. So I don’t claim that our blaming and praising are like the umpire’s calls in baseball, except for the tracking feature of the analogy.

Part of why is the reason you note. Umpires make the call and thereby make (in some important sense) the pitch a strike or not. (A funny scene from The Naked Gun depends on this element.) Lots of sports have this feature, but what makes sport so frustrating at times is that umps are presumably trying to get things right. (Indeed, as modern broadcasts constantly remind us, we do check an umpire's work against the actual facts of the each pitch.) Moreover, the rules of baseball specify a strike not as one called as such by the umpire, but by detailing an area through which the ball must travel. So, presumably, the umpire is trying to adjudge whether the ball in fact did that. In this way, umpires are like judges trying to figure out what the law says. When they make such a judgment, that is what the law says (from the law's perspective), but they can still make mistakes.

Of course, why should we care about the trajectory of a ball? You're right to note that we probably don't except for its being part of the game of baseball. And our practices of blaming and praising aren’t like that. Again, you’re right, though I still think the practice depends on some prior facts. In this case, they are the prior facts of the path of the ball. It’s true that there is no strike zone prior to rules of baseball, but there are still the facts that constitute that zone regardless. The practice has to specify which facts to pay attention to, the ones that will be salient for the practice. It “promotes” those facts. So, to push the analogy, which I still don’t intend to put too much stock in, I don’t think we should think that to be out on strikes is just to be called out on strikes. The practical upshot of being called out is, of course, decided by the umpire’s call. But people argue with umpires all the time, and it doesn’t seem to me they’re making a conceptual mistake.

But, just to be clear, I don’t think the challenge I’m trying to make out for practice-based views hangs on or is even particularly strengthened by the umpire analogy. I just liked a particular feature the example brought to mind.

Hi Matt,

Thanks for the reply.

You say: "Of course, why should we care about the trajectory of a ball? You're right to note that we probably don't except for its being part of the game of baseball. And our practices of blaming and praising aren’t like that."

But the practice-based theorist thinks that they are exactly like that in a relevant respect, doesn't she? We don't care about the quality of the agent's will, for instance, independently of our practice of holding responsible (where this is supposed to encompass both praising and blaming). This was my second point in my previous comment. The baseball analogy seems to me to highlight how many things that might appear to be practice-independent facts, or facts that are prior to our practices, are not so--at least, not facts that we care about independently of those practices. There is no strike zone independently of the practice of playing baseball. Now, as you put it in the previous comment, there are facts such as the trajectory of a ball independently of the game. But we don't *care" about this fact independently of the game, do we? Similarly, we don't *care* about quality of will independently of our responsibility practices, do we? At least, this is the point I want to press on behalf of the practice-based theorist.

Ben,

I’m a little unclear now if you mean your comments to be helping the practice-based theorist meet the symmetry challenge, or just as independent support for practice-based views. If it’s the latter, I’m happy to admit there could be independent lines of support.

But if it’s the former, I may need some help in filling in the details. So here’s my basic idea: when it comes to blaming and praising agents across a variety of normative domains, both moral and non-moral, the same features undermine blame and praise symmetrically. We should want an account of this symmetry. A natural and attractive conjecture would be that these features affect symmetrically because of something all the instances share. One proposal for what they all share is an independent responsibility relation, which grounds the relevant evaluations, both positive and negative. If the features undermine this relation, then we would expect the symmetry we find. I then suggested that practice-based models would have trouble explaining the symmetry, since they explicitly deny that there is any such prior responsibility fact.

If I read your suggestion as a reply to this challenge, then you might be saying something like the following: It only looks like there’s a prior fact. In reality, that fact is only significant because of our practices. We wouldn’t care about that prior relation if we weren’t in for the business of praising and blaming. So what appears an independent fact is really given its significance by our practices, so the practice model is actually needed in order to do the relevant explaining. Is that what you mean?

If that is what you mean, I’m a little unsure of how to take it. It starts to look like the claim is ‘we wouldn’t (shouldn’t?) care about the responsibility relation but for all the practices that care about that relation’. And I think that’s right – but trivially so. The independence of the responsibility relation isn’t supposed to suggest that it is important shorn of all connection with the relevant practices at that level. I doubt that anything in ethics is important in that way.

Hi Ben, Thanks for pressing me on the umpire analogy. I really should have said “strike zone” rather than “strike”. That is, we could provide an account of the strike zone in baseball in terms of the conditions under which certain responses from the umpire are apt or fitting. In any case, I don’t want to push this analogy too hard. There are many other analogies I could make that might be better. For example, one could provide an account of Being Square in terms of the aptness or fittingness of certain “geometric responses”, such as the judgment “That’s square”. In general, for any interesting property in the world P, we can provide a response-based account of the property in terms of the aptness or fittingness of certain responses R; the trivial default example is that the relevant members of R consist of judgments of the form “That’s a P”. In the vast majority of these cases, we correctly reject a response-based account of the P’s because we recognize that the relevant R’s are tracking the P facts; it is not necessary to talk about the R’s at all in characterizing the P’s.

So in order for a response-based view to be interesting, it needs to show the following is true:

(D) The responses R don’t simply track the P facts. Rather, there is some deeper conceptual or metaphysical dependence relationship between R and P.

For example, Jacobson and D’Arms try to establish that (D) is true for a certain class of sentimental values (e.g., aesthetic values) which are supposed to depend on certain emotional responses. I don’t think they succeed, but I respect their approach, especially because they are clear about trying to establish (D) in their domain of interest.

From my understanding of the literature on the practice-based view (which admittedly may be incomplete), I don’t think defenders of these views really try to do this at all. In fact, Strawson seems to undermine (D) from the very start when he himself offers an account of the relevant P that is the basis for responsibility in terms of “quality of will”, where this P gets characterized without adverting to the reactive attitudes at all; he in effect says the reactive attitudes are apt or fitting when they correctly track quality of will.

Ben notes that we *care* about quality of will because of the responsibility practices they are part of. But *caring about P because of R* is not a suitable conceptual or metaphysical dependence relation to justify a response-based approach. People care about things for all sorts of reasons (satanists care about pentagons because of the role of this shape in satanic practices). That doesn’t mean we should adopt a response-based approach in characterizing these P’s.

I have already gone on *way* too long. But let me add that John (Fischer) makes a point that parallels what I am saying here in chapter 8 of Deep Control. He comes at it a different way by talking about certain biconditionals. I want to echo something John says in that chapter: It is really hard to understand what people committed to the practice-based view are actually saying. So if I have completely misunderstood the practice-based view, I would love it if someone would correct me.

Finally, I am not really sure if we are on topic any more—we seem to be debating the practice-based view itself rather than Matt’s very intriguing argument against it (I like your extension to non-moral praise quite a bit Matt). So maybe this is a topic for another day or else maybe contact me directly to continue the conversation. Sorry if I am hijacking the thread. Just can't resist throwing in my $.02 on this topic that I find really interesting.

Matt,

I was originally just trying to find out how the umpire analogy was supposed to work in support of the prior fact model--but I did suggest that I thought it ended up working better as support for the practice-based model. I'm still confused.

I guess I don't really have a grip on what the prior fact is supposed to be prior to. Opposing the prior fact model to the practice-based model suggests that the fact is supposed to be prior to the practices. And I take it that the relevant sense of priority is explanatory. So I take it that the prior fact view you want to motivate claims that there are facts--you cite a responsibility relation--that are explanatorily prior to our practices of holding responsible.

But--and this is the point I was after--the umpire analogy suggests that while there may be facts prior to particular instances of the practice, these facts are not explanatorily prior to the practice itself. The practice explains why we pick these facts out as relevant at all and also why they are supposed to be tracked in a given instance of engaging in the practice.

Now, I'm not sure what to say about the symmetry worry because I am unsure how to understand the dialectic. The worry seems to be that the practice-based model cannot explain why there are certain facts that undermine both praising and blaming responses. And the proposal seems to be that the way to handle this data is to appeal to the explanatory priority of these facts to the practices (partly) constituted by these responses.

Perhaps my suggestion could be put like this: there is a distinction to be made between (i) priority to the practices and (ii) priority to particular instances of them. I don't yet see (and your final comment in your previous reply suggests that you might agree) that there are good reasons to think that there are responsibility facts prior to our practices of holding responsible (taken to include both praising and blaming), even if there are good reasons to think that there are responsibility acts prior to particular instances of this practice.

Am I understanding your view correctly? If not, please help me out. Also, do you find it plausible to distinguish between a practice and a given instance of it, such that a fact could be explanatorily prior to the latter while being explained by the former? If not, why not?

Chandra,

I just saw your comment after I posted my previous one. Does the distinction I mention give you what you are looking for? To put it in your terms, does the claim that there are certain P facts because there are certain practices, partly constituted by certain R responses, which are supposed to track the P facts, provide something like the kind of conceptual of metaphysical dependence relation you are after? The idea is that the facts tracked by the responses depend on the practices, which are partly constituted by the responses.

@Ben: To your question: “Does the claim that there are certain P facts because there are certain practices, partly constituted by certain R responses, which are supposed to track the P facts, provide something like the kind of conceptual of metaphysical dependence relation you are after?”

I say the answer is no. In fact, I can’t really understand what it means in the first place to say that there are certain P facts *because* of certain practices.

Let’s say Strawson is right that the relevant responsibility facts, i.e., the P facts, tracked by the practices are facts about qualities of will. In particular let us say a person acts out of ill will if the motive on which she acts is one that displays a lack of concern for the welfare of others. What does it mean to say these P facts—these facts about a person’s lack of concern—explanatorily depend on our practices? Suppose Mary A’s and in doing so, she acts out of ill will. Her acting out of ill will is a fact about the motive that moves her. How can this fact—a fact her motives—depend explanatorily, conceptually, or metaphysically on our practices or reactive emotions. Surely, an intrinsic duplicate of Mary in a world without the reactive attitudes could still display ill will in her actions if the motive that moved her displayed a lack of concern for the welfare of others.

So again, I deny that moral responsibility depends in any deep conceptual or metaphysical way on reactive emotions or our practices. We can specify the relevant responsibility-conferring facts—facts about qualities of will—independently of any mention of the practices, and so these facts can’t be dependent on those practices. Am I missing something here? I feel like I am.

Hi Chandra,

This is very helpful. I may be beginning to see where our disagreement lies.

You say: "let us say a person acts out of ill will if the motive on which she acts is one that displays a lack of concern for the welfare of others. What does it mean to say these P facts—these facts about a person’s lack of concern—explanatorily depend on our practices? Suppose Mary A’s and in doing so, she acts out of ill will. Her acting out of ill will is a fact about the motive that moves her. How can this fact—a fact [about] her motives—depend explanatorily, conceptually, or metaphysically on our practices or reactive emotions. Surely, an intrinsic duplicate of Mary in a world without the reactive attitudes could still display ill will in her actions if the motive that moved her displayed a lack of concern for the welfare of others."

A couple of points in reply. First, even if Mary's acting out of ill will is a fact about the motive that moves her, it is a *normative* fact. It is not enough to establish a claim about quality of will to cite the agent's effective motive. Quality of will is a function of the reasons for which one acted given the reasons one had for/against so acting. Mary might act on the same motive in different circumstances, and these different actions, both issuing from the same motive, might display distinct qualities of will. And part of the explanation of the difference may appeal to our practices. We take acting on one's desire to stop for a cup of coffee to be perfectly benign when there is time, but it is rude when one is late for an appointment. It seems very difficult to explain the difference in quality of will displayed by Mary when she stops for coffee in these two scenarios without appealing to our practices of scheduling, honoring appointments, treating people with respect, etc. (Now, one might make the case that these practices don't essentially involve the reactive attitudes, and I might agree. But they are still practices, and so the fact of Mary's quality of will, it seems to me, depends on practices--some of which might involve reactive attitudes (it may be appropriate to resent her when she walks into the important meeting late and sipping a Frappucino).)

Second, and relatedly, there are familiar views about normativity on which such facts depend on our practices. So one thing that may be at play here are divergent metaethical views about the nature of the normative. I am partial to some version of constructivism; perhaps you prefer some version of realism. Maybe this difference explains some of the distance between our views on the matter.

In any case, I hope these comments are useful. I am finding this discussion helpful in sorting out some different ways of thinking about these matters.

Hi Ben, Your response is indeed helpful and really interesting. I won’t try to weigh in on constructivism in metaethics (you are right, I am a realist -- of the Railtonian kind). But leaving aside the merits of your proposal, it is clearly not the tack taken by Strawson or most defenders of the practice-based view. Metaethical constructivism plays no role in their argument. Rather, they seem to be arguing quite directly that being responsible is in some deep sense conceptually or metaphysically dependent on reactive emotion responses specifically.

Thanks to Ben and Chandra for keeping the discussion alive and interesting. My teaching yesterday wiped me out so it was all I could do to follow along. But now that I’m refreshed, I can resume fighting my own battles...

I think recent disagreement has covered some interesting ground, and I’m not sure what to make of all of it. I thought I would take a small step back here, though, and try to assess where we’re at. (Apologies at the outset for the length!)

First, I want to try to make clear the distinction between practice-based models and the prior-fact model in terms of explanatory priority, which was the focus of my initial post. So, consider this quote from Watson:

“It is not that we hold people responsible because they are responsible; rather, the idea…that we are responsible is to be understood by the practice [of holding responsible]” (‘Limits of Evil’).

That sums up the distinction pretty nicely, I think. He is characterizing a practice-based view, which explicitly denies any prior fact of responsibility. Similarly, for someone like Wallace, there really is no fact of “A is responsible for x”; such a claim would be shorthand for “our blaming practices show blaming A for x to be fair”.

In contrast, the prior-fact model says the explanatory priority lies with an independent responsibility relation. Satisfying some criteria makes one responsible for x, and this is a fact our practices can go to work on. So, there is a fact “A is responsible for x”, no matter what our practices look like. So long as there are creatures with our powers of agency, in my view, there would be responsibility facts.

Second, Ben might be right that one or the other model may sit more or less comfortably with certain metaethical commitments. If, for example, one thinks there are no practice-independent moral facts, then one won’t be attracted to a prior-fact model. So, if one’s other commitments incline one to draw the ambit of our moral practices very widely, and see their significance as providing a foundation for ethical theory, then the prior-fact model will be at odds with how one conceives the ethical enterprise.

But what I want to stress is that the force of the symmetry argument, as I see it, is that it not only points to symmetry in our moral assessments, but in our non-moral ones too. Even if one sees our ethical enterprise as being entirely practice-dependent, and even if our non-moral normative enterprises are also practice-dependent, it’s hard for me to see that these practices are going to look much alike. And if we need to explain the symmetry operating across these domains out of those practices, the more different they are, the more the difficult such an explanation will be. In contrast, appealing to a prior-fact can explain an indefinite range of practices, since it allows for any practice to pick up on that relevant fact. So the more diverse the practices, the better things look for the prior-fact view. Or so I claim.

The comments to this entry are closed.

Books about Agency


3QD Prize 2014: Marcus Arvan