What Is Moral Responsibility?

In the comments  on my previous post, David asks the following question:

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.

That seems to me to a bloody good question. The standard answer, if there is one, is the one which I had previously given David: aptness for the reactive attitudes. But it has always seemed unsatisfactory to me. I think part of the problem is this: though the definition might 'fix the reference', it actually doesn't tell us what moral responsibility consists in (so far as I can see). Moral responsibility must be some kind of relationship between an agent and her acts; it doesn't require any observers, or even the persistence of the agent after the act (for what it's worth, I have the same problem with definitions of knowledge: the problem with JTB is not that it open to counterexamples, but it leaves me in the dark as to what knowledge actually is).

So I want to echo David's question, and ask, further, for clarification. Am I missing the point? Is there nothing to say about what MR consists in, beyond giving necessary and sufficient conditions for it? Does anyone - in the literature or here - have anything illuminating to say that will answer David's and my question?

Comments

[Zac Cogley]

I concur that this is a bloody good question, and not one that gets a lot of treatment in the literature. Fischer's 1999 "Recent Work on Moral Responsibility" (Ethics, Vol. 110, No. 1, Oct., 1999) mentions a few options other than fitness for the reactive attitudes. Here's my brief gloss on Fischer's brief gloss on those notions:

Marina Oshana claims that responsibility should be understood in terms of being accountable, where accountable is understood as it being proper for a person to give an account of what she did (or why she did it).

Gary Watson holds that there are two senses of responsibility. One is essentially the Strawsonian version, while the other is linked to the notion of an action "belonging to an agent," or the action "being the agent's own" in some sense. Watson calls this non-Strawsonian concept of responsibility the aretaic face of responsibility.

Fischer also wonders briefly whether there is an important conceptual relationship between responsibility and autonomy, and suggests that responsibility is necessary for autonomy, but not vice versa.

I don't know whether or not this exhausts the possible options for explicating the concept of responsibility, but it's a start, at least.

In any case, this thread gives me a chance to ask a related and important question: What is a perspicuous way of explaining the distinction between the CONCEPT of moral responsibility (which would seem to call for one of the above analyses) and the CONDITIONS OF APPLICATION of the concept (which would seem to call for necessary and sufficient conditions to satisfy the concept)? To use David and Neil's lingo, the question of the "definition," then, is a question about the concept. The terms in caps are Fischer's way of making the distinction, but this way of drawing it is not intuitive (in my experience when trying to appeal to it) and I have been at a constant loss to make this distinction clear to other philosophers, even while gesturing at other related examples from other fields of ethics. (The concept of personhood and its conditions of application is my favorite example.)

The problem seems to be, at minimum, that any terms I try to use to draw the distinction immediately have other associations for the people I'm talking to. Some philosophers immediately think of necessary and sufficient conditions whenever you talk about a concept, and this makes them think of the conditions of application. I've found that even Neil and David's use of "definition" is ambiguous between the two projects for most people.

But regardless of the terminology that is used, is there a clear way of making out this distinction, or a way of explaining the different explanatory/conceptual projects involved?

[John Fischer]

In various places, I have suggested that we should distinguish between the concept of moral responsibility and its conditions of application. Of course, Quine (and others) have been skeptical about such a distinction, but I have simply assumed that Quine can be answered.

I suppose a definition seeks to explicate a concept. So the definition of "moral responsibility" seeks to elaborate our concept of moral responsibility. It tells us the essence of moral responsibility, or what moral responsibility is. One might use the apparatus of "necessary and sufficient conditions" to give the definition of moral responsibility.

Now we might also be interested in saying when exactly, or under what conditions, the property of moral responsibility is instantiated. One might also use the apparatus of "necessary and sufficient conditions" to help to answer this question.

Consider, as an analogy, Rawls' theory of justice. He distinguishes the concept of justice from particular "conceptions of justice". The concept of justice is shared by the proponents of the different conceptions of justice; the concept of justice is the definition of "justice". But the definition of justice is something like an "appropriate or fair distribution of relevant resources, treating like cases alike." This does not in itself distinguish between "liberalism," "utilitarianism," "Marxism," and other particular conceptions of justice. These are ways of specifying when the concept obtains, or the conditions of application of the more abstract notion which is accepted by the various parties to the dispute.

So with moral responsibility. I have pointed out, in various places (including the first chapter of RESPONSIBILITY and CONTROL (with Mark Ravizza), that there are various plausible candidates for the defintition or specification of the concept of moral responsibility, including the "ledger" view, the "Oshana view," and the Strawsonian view. Presumably, Jay Wallace's book is in part an attempt to argue for the Strawsonian view about the concept of responsibility (or definition of "moral responsibility").

Of course, if one presses the Quinean/Davidsonian skeptical line, one can question the distinciton between concepts and their conditions of application, or definitions and mere claims. After all, the distinction between concepts and their conditions of application does depend on some sort of analytic/synthetic distinction.

Besides Ravizza and my book, we also seek to lay out the possibilities in our introductory essay to our Cornell University Press anthology, Perspectives on Moral Responsibility. Also, in addition to Wallace, one might find helpful the entries on "free will" and "moral responsibility" by O'Connor and Eshleman in the Stanford Encyclopedia of Philosophy?

Have I helped, or just made things worse?

[John Fischer]

To explain a bit further. One might distinguish, ala Locke, a nominal and real essence of something. The definition of the word, or the concept, is associated with the nominal essence. This may be different from the real essence.

In my way of thinking about moral responsibility, the nominal essence is something like the ledger idea, Oshana's "explanation view," or the Strawsonian aptness of reactive attitudes view. On the other hand, candidates for the real essence are Ravizza and my "guidance control" view, the libertarian's regulative control view, and so forth.

[John Fischer]

Another way of putting the point. Suppose one thinks that what we mean by "moral responsibility," or, alternatively, what we have in mind when invoke the notion of moral responsibility, is "being an apt candidate for the reactive attittudes." Still, even if this is correct (and I take not stand on this issue), one would be interested in knowing which agents, on which occasions, have this property. So, for example, we would want to know whether the agent must be free to do otherwise, in order to have the property, or whether causal determinism is compatible with the agent's having the property, and so forth. So the question about what the meaning of the term "moral responsibility" or what the concept is, is different from the question of who can properly be said to have the predicate, "morally responsible", and on what occasions, or who has the property associated with the concept in question. They are importantly separate questions.

[Neil]

Thanks for the responses, John. I'd forgotten about the ledger view (time to reread *Responsibility and Control*). It's possible that what I, and David, are after is itself confused. There seems to be something of a paradox of analysis lurking nearby. I'm not satisfied with a reply like "MR is accountability", because it seems uninformative: a synonym is offered, not an informative definition. But when I'm given necessary and sufficient condtions, I feel informed - but I'm not sure what I'm being informed about!

That said, and at the risk of further revealing my confusion, here's a way of sharpening the question in the way that should have particular relevance to you. In my previous post, I took issue with Todd Long's suggestion that you and Ravizza should say that the agents in his Frankfurt-style cases are responsible, and responsible for something wrong, but not blameworthy. The examples you give in *Responsibility and Control* of responsibility without blameworthiness are (from memory) cases in which the action is not morally significant. But Long's cases are not like this. If he's right that there is room for a conceptual distinction between "responsibility for a wrong action" and "blameworthiness", then just what is the conceptual difference?

[David]

I'd just like to thank everyone for their comments, and Neil for kindly making this a topic of discussion. It's been very helpful to me, as a new student in this area.

One thing strikes me: suppose you and I differ on the necessary and sufficient conditions that must obtain in order for one to be held morally responsible for one's action. (I, for instance, do not believe that one need have performed the action freely in order to be held responsible, but you do). One would want to say, I think, that we have in mind the same concept - "moral responsibility" - which we believe has different conditions of appliation. But, there's a temptation to say we have in mind a different concept if we were to identify the nature of the concept itself with the conditions of application of that concept. However, I'd want to hold onto the thought that we do have in mind the same concept.

What I'm trying to say, I think, is that the difference between the nature of the concept itself and the conditions of its application is a robust distinction - and one that needs to be recognized in this sort of work. But, it's hard to get a clear handle on the difference without begging questions.

Many thanks

[Rob]

David,

I'm doubtful as to how well the archives of this blog support your contention that the difference between the nature of the concept of moral responsibility and the conditions of its application is a robust one. Rather, I think quite the opposite impression suggests itself: they are inextricably inter-related, with difference notions of the conditions of application manifesting differences in definitions of the concept, and vice versa.

[David]

Yes, perhaps you are correct, Rob. I suppose what I mean is that I can't shake the feeling that there is a difference, but there doesn't seem a way to spell it out. I should have said that " is robust" predicates my intuition of a distinction and not the explication of a distinction.

This discussion has greatly helped me as I move to the dissertation phase of my degree. Thanks again to all.

[John Fischer]

Rob, I believe, is suggesting the Quinean point.
As to Neil's question, I agree it is difficult to spell out the distinction between being morally responsible (on my sort of view) and say being blameworthy. At least I'm in good company, as Harry Frankfurt has the same kind of view. On my view, being morally responsible is being an apt target for the reactive attitudes (at least insofar as I'm willing to provisionally adopt the Strawsonian approach to the concept of moral responsbilility). So, to use a metaphor, an individual who is morally responsible is "in the ballpark" for the reactive attitudes; it would not be a "category mistake" of some sort to apply the reactive attitudes to such an individual (on the basis of the relevant thing--action, omission, consequence, and so forth). But whether the agent is actually blameworthy requires something additional--but difficult to spell out in general. I should admit that I don't have a general theory of blameworthiness (or praiseworthiness). I can identify relevant factors, but I don't have a general theory. In Nomy Arpaly's work, she tries to do exactly this--to provide a general theory of blameworthiness (as opposed to moral responsibility). So, in principle, one could meld her approach, or something similar, onto mine--mine tells you when someone is in the ballpark for (say) blame, and hers tells you when to blame such a person, or perhaps when such a person is blameworthy (arugably, the latter two notions are related but different).

Thanks for pressing these excellent questions! I wish I had excellent answers!
And, David, good luck on your dissertation!

[Rob]

So, I wonder, does anyone sharply divorce being blameworthy from being morally responsible, such that the former can be the case without the latter (perhaps ever) being the case? Are there proponents of a kind of ethical outlook which accommodates the reactive attitudes, but without accepting the reality of moral responsibility or conflating it with blameworthiness? Is it even a coherent agenda?

[Jussi Suikkanen]

Rob,

that actually sounds like a coherent possibility. Say you were a strict consequentialist and a hard determinist. In that case, you might first say that people's attitudes of blame and praise had beneficial consequences. Therefore these attitudes were appropriate. You could then go on to say that people were worthy of blame and praise in the case of blaming and praising them was conducive to general welfare. A strict consequentialist would say that there cannot be any other grounds, such as desert, for assessing whether blame and praise is justified that the consequentialist criteria.

Yet, it would be coherent to also say that no-one is ever morally responsible for what they do. This could be argued in the traditional way that our acts causally flow out of our characters, and under the conditions of determinism there are no self-forming actions for which there already weren't sufficient causal backgrounds. Thus, we could not have done or willed otherwise, and therefore we cannot be morally responsible.

All of this sounds coherent, and I am sure there have been some people who thought like this (there seems to be a philosopher for every coherent view and sometimes even incoherent). Only problem perhaps would be that if it turned out that *saying* that people were morally responsible had beneficial consequences then, in order to be consistant, the sharp divorce might collapse.

So, it looks like a coherent view to take. But, that does not make it a good view. The two views I used in the example at least are such that I would like to argue against them.

[Mark Smeltzer]

Rob,

I think the view I've presented honors the possibility of such cases. Suppose agent A is evil and objectively blameworthy, yet we have no grounds to hold A morally responsible for any of his evil actions due to certain features of A's environment (generally speaking, these will be known cases of direct and very strong indirect manipulation of the agent [c.f. Pereboom's "four cases" argument]).

On my view it would be true that A is objectively blameworthy, but not because he is morally responsible for his actions -- in fact, on my view A is blameworthy *in spite of the fact* that he is not morally responsible for his actions. Moreover, despite the fact that A is objectively blameworthy in this case, we cannot (presumably) know this, and thus A cannot be held morally responsible for his evil actions.

The question of whether A *is* morally responsible for particular events would be beyond answer, but would depend on whether A's evil character "made the difference" in determining the particular event-outcome(s) in question (a question that we wouldn't ever be able to answer conclusively). In cases of indirect manipulation (even very strong indirect manipulation) it is possible for an agent's character to perform this function despite the challenges presented by the agent's environment. In cases of direct manipulation, the agent is not responsible by definition.

In the end, I am skeptical about the frequency of such cases in everyday life (though moderate indirect environmental manipulation does seem to present a legitimate threat to responsibility), but it always is nice to have a theory in one's pocket that accounts for the problems that can be imagined as well as realized.

[Paul Torek]

Just to complicate the discussion, I'd like to simultaneously suspect that Quine can be answered, but that the answer will still leave philosophical enthusiasts of conceptual analysis pretty humbled. Relatedly, I recommend that we not choose between the definitions mentioned above, but begin by assuming that the "meaning" of the terms, and the "narrow content" of our thoughts, is a highly disjunctive affair. That is, something would count as "moral responsibility" if it captures enough of these defienda. And, as a corollary that doesn't quite follow (but has decent confirmation from experience), that we are better off looking for a Lockean "real essence" early, with minimal discussion of the nominal essence.

First discover hydrogen, oxygen, and their favorite compound. Then define water. Hope that enough of the conceptual penumbra surrounding "water" are captured, and further conceptual debate becomes unnecessary.

[John Fischer]

Paul: that's pretty much what I do. As you perhaps have noted (or will note), in my work I spend only a bit of time laying out a disjunction of possibilities for the nominal essence or "concept" (or whatever). I spend most of my energy on the "real essence". My view: the real essence, or conditions of application of the concept, of moral responsibility = guidance control (and some epistemic condition). Guidance control = moderate reasons-responsiveness plus mechanism ownership. I am officially neutral on the "conceptual" or "definitional" issue, and am willing to accept some sort of "disjunctive" account (roughly speaking).