Todd Long on Moderate Reasons-Responsiveness.

I’m reviewing the Campbell, O’Rourke and Shier volume. The review will be short, so I won’t be able to say much about it (it’s always hard to review collections in any case). But of course I’ll have lots to say about many of the papers. Rather than waste my brilliant insights, I’ll post them here, if I think they might interest anyone. I begin with Todd Long’s interesting discussion of Fischer and Ravizza’s (F & R’s) account of moral responsibility.

Long aims to force some kind of revision on the F&R account of responsibility. That account states that an agent is responsible if she acts upon a moderately reasons-responsive mechanism of her own, where a mechanism is an agent’s own if she has taken responsibility for it, and a mechanism is moderately reasons-responsive if regularly receptive to reasons, including moral reasons, and at least weakly reactive to reasons.

Long tests the sufficiency of the F&R account by means of some variants on Frankfurt-style cases. F &R claim that agents in Frankfurt-style cases are not morally responsible for their actions if the intervention actually occurs, because the condition outlined above is not satisfied: specifically, in these cases the agent does not act upon a mechanism of her own. F & R argue that in the alternative sequence in Frankfurt-style cases (the sequence in which there is an intervention), the agent acts upon a different mechanism than in the actual sequence (in which she acts upon her own). And in most such cases this seems right. But Long aims to construct cases in which the mechanism is the same in both sequences.

In one such case, the alternative sequence intervention consists in the intervener adding new inputs to the very same mechanism that is at work in the actual sequence; specifically, he adds reasons directly into the agent’s deliberative mechanism. In the other case, the intervener subtracts reasons from the deliberative mechanism. In both cases, the intervention causes the agent to perform a wrongful action that, in the absence of the intervention, she would not have chosen.

Now, should F &R say that in these same-mechanism cases in which the intervention is actual, the agents are morally responsible for their actions? Long believes that they are committed to saying yes – and that “yes” is in any case the right answer (in other words, Long does not believe that his cases are counterexamples to F & R). He points out that F & R allow that an agent can be morally responsible for an action without being blameworthy for it, and he claims that this what we should say here.

I can think of two kinds of case in which an agent is morally responsible for an action without being blameworthy for it. In the first, the agent is not blameworthy because she is praiseworthy. In the second, the agent is not blameworthy because the action is morally neutral.  Obviously, the kind of case Long has in mind, in which the action is wrong, does not fit either category. It seems to me that in this kind of case the separation of moral responsibility and blameworthiness is unmotivated here. What blocks the reactive attitudes, here, which doesn’t block moral responsibility? In any case, if Long’s move works, then it turns out that F & R’s account is a lot less interesting than we thought it was. We thought it allowed us to make interesting discriminations between cases, but it turns out that it doesn’t. All the action needs to shift to the (so far mysterious) account of aptness for the reactive attitudes instead.

But Long is right that the move Fischer makes in response to his cases is equally implausible.  Fischer argues that the agents in the Frankfurt-style cases are not responsible, for precisely the reasons that agents are generally not responsible in these cases: because they do not act on the same mechanism in the sequence in which there is an intervention. Long is right that making this move multiplies mechanisms beyond necessity, and that it leaves F & R without a properly general account of moral responsibility. We need to be able to hold mechanisms fixed across scenarios, or we shall not be able to make the needed comparisons to assess moral responsibility. In any case, it seems implausible to individuate mechanisms by their inputs.

I’m going to suggest a third possibility: build upon the epistemic conditions F & R build into their theory. I suggest adding the following condition to their undeveloped epistemic condition:

In order to be morally responsible for an action, an agent must either (a) make her choice in the light of a sufficient number of true beliefs about the situation or, (b) in cases in which (a) is not satisfied, be responsible for her ignorance.

This condition seems circular, because it has the unanalysed notion of responsibility in (b). But I don’t think this is fatal, because we can cash out (b) in terms of (a) and the rest of F &R’s account. So ‘responsibility’ in (b) can be analysed out.

In Long’s cases, the agent is not morally responsible, because she acts on the basis of false beliefs, or in the absence of sufficient true beliefs, and she is not responsible for her ignorance. Now compare Long*, in which the intervener acts so as to insert either sufficient true beliefs, or to remove a sufficient number of false beliefs, for the agent to make the right choice. In this case, the agent is morally responsible (praiseworthy), it seems to me, but Long will need to hold that thought she is morally responsible, she is not praiseworthy. It is a virtue of the suggested amendment that it treats these high tech interventions in exactly the same way as low-tech interventions in people’s beliefs: lying to them, say, or telling the truths.

This might in fact be not so much an addition to the F & R account as a fleshing out. Their epistemic condition states that “in order to be praiseworthy or blameworthy, a person must know (or reasonably be expected to know) what he is doing”. Conditions (a) and (b) simply put some flesh on this claim.

Comments

[Jussi Suikkanen]

'specifically, he adds reasons directly into the agent’s deliberative mechanism. In the other case, the intervener subtracts reasons from the deliberative mechanism. In both cases, the intervention causes the agent to perform a wrongful action that, in the absence of the intervention, she would not have chosen.'

Question of clarification; What's meant by 'reasons' here?

I can think of two options.

A) Pairs of desires and beliefs in the sense most people speak of explanatory reasons. Then it is not clear that these are normative. I take it that the reasons-responsive mechanism F&R talk about is responsive to normative reasons. Now, if these added 'reasons' are just psychological forces with 'push' it is not clear that if they prompt the agent to act she acts from her reasons-responsive mechanism.

B) True normative reasons - things out in the world that favour the action. But then the manipulator would change the situation to such that the agent has new good reasons to act in a way that she now does. And she detects that there are new reasons and acts from her RRM. I see nothing in this option that would undermine responsibility, agents just finds herself in a new situation and acts freely.

So, either way, I am not sure I see the problem for F&R. But maybe there is some third sense of reasons behind Long's objection.

[David]

A quick question: what do you mean when you use the term "morally responsible", for instance in the phrase "an agent is morally responsible for his action"?

In asking this question I don't mean for you to give me necessary and sufficient conditions which must hold if the agent is to be held morally responsible for his action. Instead, I am after what you think the term means rather than the conditions needed in order to be (or be held) morally responsible.

Many thanks

[Paul Torek]

Neil, what I really like about your analysis is its recognition of the breadth and dynamism implicit in the term "reasons-responsiveness". Reasons-responsiveness certainly extends, as you point out, to the formation of the agent's beliefs. Arguably, it extends to the formation and evolution of desires, too, even desires for ends and not merely means. (Not that I could argue that here.) Because of the controversies over what reason requires in these domains, it can be tricky to construct clear-cut examples of agents whose reasons-responsiveness has, or has not, been impaired. But perhaps that's as it should be.

[Neil]

In the order of comments...

Jussi, Long seems to mean "beliefs" by "reasons". Long wants us to share his intuition that if A lies to B, and B then performs an action rationalized by the false beliefs he has acquired from A, B is responsible (though not necessarily blameworthy) for his action. He says that if this isn't the case, then we're a lot less responsible than we thought - advertizing creates beliefs in us, as does political commentary, which are analogous. So these cases do not fall into either of your categories. They are false beliefs, unlike your (b), but they are beliefs, unlike your (a).

David, if I knew the answer to your question, I'd publish it in J. Phil! We tend to work with an intuitive notion of moral responsibility. If we have a definition, it is this: aptness for the reactive attitudes. But I don't think that definition will do, because our intuitions about reactive attitudes are a mess. I can do no better than point to paradigm cases and say *that* is what I mean by moral responsibility - the relationship between that agent and her action.

Paul, I'm glad someone agrees with me about including beliefs within the scope of MMR. It's surprisingly controversial. The best argument *against* it, I think, comes from Frankfurt-style cases. If we adopt a real-self view, then we can easily explain the responsibility of agents in such cases.

[Jussi Suikkanen]

Neil,

thank you for clarification. If Long means beliefs by reasons, I think that this would also be a good issue to take up. This is because it seems that beliefs are not a) exlanatory reasons (One cannot explain action by siting mere beliefs because they seem motivationally inert - one needs also desires.) and b) not good normative reasons to act on, i.e., they are not what favours the action (well, this is a big debate but I go with Dancy here: belief that there is water in the glass is not a good reason to drink it but that there is water in the glass). Therefore it is questionable whether they are reasons in any sense at all.

Of course if Black adds beliefs to Jones's psychology Jones might already have had desires that in combination with the new desires prompts him to act. Then, I guess this combination of mental states might be the explanatory reason to be sited to explain the action. The question remains though that if the desire part of the reason was Jones's and derived from his values (what I guess one might call RRMs), then it might still be the case that the action was Jones's in a robust sense that can ground praiseworthiness. I think it might. Say Jones was on a beach and Black added a belief that there is a child drowning. Jones would have had a desire to save the drowning child and thus he would jump to save the kid only to find that there is none. We might want to say that Jones still acted in an admirable way.

On the other hand, if beliefs, as reasons, are added to agents deliberative perspective then one might also have grounds to argue that therefore the agent's action is not based on her reason-responsive mechanisms or at least not totally on them. After all, formation of beliefs is one primary locution of being responsive to reasons. When beliefs are 'added' this mechanism is certainly overridden. Thus, passing the RRM would make the action less the agent's own, and we would have less reason to praise Jones in the previous example.

[David]

Thanks for your response. I agree that the definition is not perfect but not for the reason that the intuitive basis of the reactive attitudes is not robust. Instead, I'd suggest that aptness for the reactive attitudes is a consequence of moral repsonsibility, not a defintion. As a consequence of one being morally responsible for one's action, one's performance of the action is apt for praise or blame.

Of course, sometimes a consequence can be a satisfactory definition. I, for example, accept as a definition that water is H2O. And it follows as a consequence of a substance being water that it is H2O. But, things don't seem to me to be like this with the notion of moral responsibility. I want to say there is a break in the consequence-definition link. Having said this, I have no argument for this!!

Another quick question if you don't mind: can you recommend any readings I might look at that attempt to explicate or defend a definition of moral repsonsibility? As I said above, I'm most interested in definitions and not - as seems to me more usual in the literature - conditions that need to be satisfied in order for one to be held morally responsible.

Many thanks.

[Neil]

In case you miss it, David, I have asked your question in a new post. It's a question that I want answered, too.

[Todd R. Long]

Thanks, folks, for the comments about my paper. I’ll reply to some things Neil wrote, as well as to something that Jussi mentioned.

1. Neil says, “I can think of two kinds of case in which an agent is morally responsible for an action without being blameworthy for it. In the first, the agent is not blameworthy because she is praiseworthy. In the second, the agent is not blameworthy because the action is morally neutral. Obviously, the kind of case Long has in mind, in which the action is wrong, does not fit either category.”

I don’t think that it is obvious at all that my cases fail to fit either category. I will want to think about this some more, but for now it seems to me that the act Schmidt does (vote for Hitler to take power) is either morally neutral or praiseworthy. Neil says the action is *wrong* (I presume because of the fact that Hitler did a lot of horrible things once he came to power). But, I don’t think that fact settles whether or not someone’s voting for Hitler is morally wrong. In my test cases, it is very plausible to think that Schmidt votes just as the vast majority of morally good, rational people would in his circumstances. Indeed, the cases can be construed such that Schmidt votes just as an ideal human agent in his actual circumstances ought to vote if he’s responding to his actual reasons. Given his actual situation (assuming now that an “alternative-sequence” is actual), it would be crazy for him to vote against Hitler. The reason that I am unsure whether his action is praiseworthy or morally neutral is that I am unsure whether “praiseworthy” applies only to supererogatory acts. If it does, then Schmidt’s act strikes me as morally neutral, or perhaps better, as “morally appropriate”; and, if it does not, then I think Schmidt’s act is praiseworthy if our usual voting acts are praiseworthy when they are the result of careful deliberation concerning our reasons (along the lines of F & R’s theory).

2. Neil is correct to think that one way to make it come out that Schmidt is not morally responsible for voting for Hitler is to monkey with the *epistemic conditions* that F & R have. As he says, F & R’s epistemic theory is undeveloped. I think it is well worth thinking more about the epistemic conditions. However, Neil’s suggestion—that one’s reasons for acting must be true in order for one to be morally responsible for that act—strikes me as implausible. Whether by “sufficiently true” Neil means “mostly true” or something much less stringent, I am doubtful that anything plausible along those lines is going to show that Schmidt is not morally responsible. One reason for my doubt is that it seems plausible to me that our epistemic twins who are under the thumb of Descartes’ evil demon can be morally responsible for their acts (although most of their beliefs are false). Why not? Epistemic externalists have recently been doing the mental equivalent of death-defying circus gymnastics to try to account for the fact (easily accommodated by epistemic internalism), that our twin demon-worlders are epistemically rational in their beliefs whenever we are. Perhaps something analogous is the case with respect to moral responsibility (although moral responsibility is surely more complicated). In any case, why would the mere fact that one’s reasons are false rule out one’s being morally responsible for what one actually does in light of the actual reasons one has during deliberation? Even if one willfully steers clear of certain kinds of easily publicly available evidence or fails in some moral obligation to gain more evidence about an important matter before acting (and, hence, let us suppose, deliberates on the basis of mostly false beliefs), it doesn’t follow that that person is not morally responsible for her act. Indeed, she might well be blameworthy in such a case. So, I don’t see the relevance of the mere fact that one’s reasons are false. If one had good reason to believe that one’s beliefs are false when one is deliberating, then that would plausibly make a difference in whether one’s act is actually praiseworthy or blameworthy, but it would not rule out moral responsibility for what one does. My sense is that careful reflection about the epistemic factors relevant to moral responsibility will lend support to my view.

3. The one thing I did not understand about your post, Neil, is your claim concerning “Long*, in which the intervener acts so as to insert either sufficient true beliefs, or to remove a sufficient number of false beliefs, for the agent to make the right choice.” You add: “In this case, the agent is morally responsible (praiseworthy), it seems to me, but Long will need to hold that though she is morally responsible, she is not praiseworthy.”

I take it that by “the right choice” you mean “votes against Hitler”. Why do you think that I’m committed to thinking that the agent is not praiseworthy in that case? I take it that, supposing the agent gains, via the intervenor, very good reasons (whether true or false) to vote *against* Hitler such that, at the time of the agent’s deliberation the agent has comparatively little reason (whether true or false) to vote for Hitler, then the agent would plausibly be praiseworthy (assuming that praiseworthiness doesn’t depend on supererogation) for voting against Hitler. I don’t understand why you think I am committed to saying otherwise.

4. Neil says of his epistemic condition idea: “It is a virtue of the suggested amendment that it treats these high tech interventions in exactly the same way as low-tech interventions in people’s beliefs: lying to them, say, or telling the truths.”

It is, indeed, a virtue of Neil’s view, and it is a virtue of mine as well. If I, through reputable means, learn that a child molester has moved next door, and I then enact a family policy according to which my daughter will no longer be allowed to play in our yard without adult supervision, I take it that I am morally responsible for enacting that policy, and perhaps, praiseworthy for having done so. And, this is so even if the report is false; that is, I am morally responsible for my act even if the person who has moved next door is innocent (that is, even if my reason for acting — my belief that a child molester has moved next door — is false). Does this count as a low-tech intervention in people’s beliefs? [I didn't used to believe that there was a molester next door, but via the intervention of the new evidence, I now *do* believe there is a molester next door.] It seems to me that nothing need change about the case by the introduction of a powerful counterfactual intervener like my Block. Concerning F & R’s epistemic condition that a person must know (or reasonably be expected to know) what he is doing, it seems clear to me that in each of these cases (low-tech and high-tech), I could know what I was doing. I need not be befuddled as I deliberate in the low-tech case, and, if the counterfactual intervener is powerful enough, I need not be befuddled at all as I deliberate in the high-tech case, either. Indeed, from my perspective, the two cases could be introspectively indistinguishable.

5. Regarding Jussi’s points: With respect to reasons-responsiveness, my comments above show that I certainly do mean for *beliefs* to be included in an agent’s reasons for acting. It is also clear enough from what Jussi said that an agent’s *desires* play a role in any plausible view of reasons-responsiveness. Feel free to plug in whatever desires you need to make my cases come out as I say they do. Although I did not discuss this point in my paper, I can’t think of any reason why my cases can’t be construed along the lines of Jussi’s beach example. Just as Jones-on-the-beach acts admirably and is morally responsible for jumping into the water to save a child, even though he will be puzzled when he finds no child there, so Schmidt (to extend my examples) acts admirably and is morally responsible for voting for Hitler, even though he will be puzzled when he discovers how bad Hitler really is.

[Neil]

1. My bad, Todd. What I was objecting to was what I thought I saw in your paper: an attempt to dissociate “responsibility for a morally valenced act” from “praise(blame)worthiness”. That’s an implausible move. But if what you meant was that someone can be responsible for an act, but not blameworthy, because the act is morally praiseworthy or neutral, then I agree.

2. My discussion is marred by an equivocation – of which many people besides myself are guilty – on “responsibility”. I think by “responsibility” in the context I meant “blameworthiness”. This yields an account which mixes internalist and externalist elements: it is a necessary condition for blameworthiness that the agent takes herself to be choosing the bad (if you like) but not sufficient (consider Ginet’s case of the agent who believes, falsely, that she can prevent a tornado destroying a town, and chooses not to. Ginet thinks she is not blameworthy, and I concur. Hence the externalist elements). I have no opinion about whether Long’s agents-in-a-vat are praiseworthy for choosing the good. But they are not blameworthy for choosing the bad.

[efau]

Can State be a moral subject? And, assuming that it can, is moral responsibility transfered from the state, in democracy or republic, to citizens?