(Also posted on Flickers of Freedom.)
I am a fan of attributability as a conception of responsibility. The trick, as we all know, is to get clear on just what that means. Even if you don't think attributability is a conception of responsibility, it is surely necessary for responsibility, so getting clear on what it consists in is in everyone's interest. One popular theory of attributability is a kind of evaluative judgment view: an action or attitude is properly attributable to me just in case it is ultimately dependent on my evaluative judgments. I'm wondering, though, about the following possible counterexample to this view.
Consider a kind of Freudian case. Suppose for years I've been spurning lovers after only a short time together, based on what I think are evaluative judgments of their minor faults. But after going into therapy, I come to the realization that I had been spurning them out of a deep fear of rejection, given that my beloved father abandoned our family when I was still young. These (spurning) actions seem genuinely attributable to me, but not in virtue of their being dependent on the "evaluative judgments" I was making at the time or the product of any real evaluative stance, then. How would the evaluative judgments view deal with such a case? Perhaps the actions just aren't therefore attributable to me for purposes of responsibility/aretaic predication? Or perhaps I was expressing an attitude that was judgment sensitive were I ideally rational? I'd love to hear any thoughts you all might have about this (and I'm grateful to David Sobel for mentioning a case like this in some earlier correspondence).
Here are two possible responses that spring to mind:
(1)
You might think that fear involves an evaluative attitude: it involves thinking of something as bad (and possible). Then even if it turns out that you acted from fear all along, it’s nonetheless true that you acted on an evaluative attitude.
(2)
Here are three things you might learn at therapy:
(a) That you didn’t spurn those lovers on the basis of their perceived faults, but instead out of a fear of rejection.
(b) That it was because of your fear of rejection that you perceived faults in those lovers.
(c) That it was because of your fear of rejection that you spurned those lovers on the basis of their perceived faults.
For the counterexample to work, you need (a) to be true. But I wonder if the more intuitive interpretation of this case is some combination of (b) and (c). After all, presumably their faults were running through your mind as you spurned them, and that alone suggests that (a) can’t be right.
Thanks, Alex. I don’t much feel the pull of (1). I can go as far as thinking of (appropriate) fear as seeing something as fearsome, but presumably (a) the Freudian case involves inappropriate fear and (b) the “seeing it as such” was missing at the time of action.
Good distinctions in (2). You’re right that (a) is the cleaner counterexample, and I’ll grant it’s not the most natural interpretation. But going with (b) or (c) may not be something advocates of the evaluative judgments view would be happy with either, because both still point to a more fundamental source of agency than “evaluator.”
Hi Dave. It seems as though the fear issues would qualify as an obscuring factor in her evaluative response. (In this case, it was simply that, prior to the therapy session, the obscuring factor had been opaque.) As an obscuring factor, it would cause her to (1) have a response to the suitors that is not fitting, (2) have a response not aptly reflective of her base sentiment, or (3) both. Regardless of what option may be the case, the presence of an obscuring factor does not mean an evaluative judgment is not occurring, it simply suggests that the evaluative judgment is perhaps an inappropriate reflection either of the subject’s “true” evaluation or of the object’s actual worthiness of that evaluative response.
Therefore, if according to the evaluative judgment view, attributability of an attitude or action requires that action or attitude to depend on an evaluative judgment, and if the fear is an obscuring factor, and if evaluative judgments occur even in the presence of obscuring factors, then it seems, in the above example, that the actions would still be attributable to the agent.
Hi Dave, Can you say more about the assumption that the actions “seem attributable” to the person in the sense of “attributable” that is at issue?
Imagine that Jane is one of the dumped. She shares his “before therapy” understanding of why she was dumped and on its basis judges that she is better off not being in a relationship with such a superficial person. This is the virtue judgment that is based on the surface attributive explanation.
Therapy reveals that the simple explanation is false. Now I imagine that if Jane comes to accept the new explanation, she will revise her assessment of Dan. She would certainly no longer think that his behavior reflects on his character in the same way, and she might well not hold him accountable for what he did in the same way.
If the attributivists you have in mind were claiming that attribution is a precondition for intentional action, I could see a serious problem, but that is presumably not the case. And I need to hear more about why in this case they shouldn’t hold that accountability is diminished or removed.
Lori, thanks, this is interesting. You’re right, fear could operate as an obscuring factor. But I think the power of the case is that once we discover the new information, the target of attribution shifts to the fear, not that the target of attribution remains the same (an evaluative judgment), just that it was clouded. Here’s another way to think about it: the relevant sort of (aretaic) predication changes: I’m no longer quite the jerk I might have been thought to be; rather, well, I have daddy issues.
This goes to Brad’s comment as well: I’m one of those who thinks accountability is just a different conception of responsibility than attributability, and I’m not even sure the extent to which the latter is necessary for the former. Indeed, I think this is actually a case in which Dan, as you’ve called him, is attributability-responsible but not necessarily accountability-responsible, given that he actually lacked a certain crucial kind of control over the actions that are nevertheless attributable to him. So by the notion of “attributability,” I do have in mind something that grounds (typically) aretaic predication. What I’m suggesting with the example is that Jane’s predication appropriately will change (perhaps along the line suggested above), so that she responds to the fact that the action was still an expression of his “self” (whatever that turns out to be), but she no longer views him as accountable (perhaps precisely because the action is no longer traceable to an evaluative judgment and so is outside his rational control).
Hey Dave. I don’t know… I see you point with regard to predication, but I’m still not convinced. A change in predication upon receipt of further information may change our response to someone, it may lessen the degree to which we think one is responsible, but I think it may be erroneous to say that, in this example, predication and the target of attribution switches from “Dave” to “fear”. The fear influenced–perhaps greatly–Dave’s evaluative judgment, but I’m not convinced that an evaluative judgement wasn’t occurring nonetheless.
Consider the following:
Say you post an example online in which you use yourself as the hypothetical subject. Say I respond to your example, and in doing so, refer to hypothetical you in the feminine. Say on previous occasion, you have witnessed I have a tendency to react passive-aggressively to others. Therefore, you have reason to believe my reference to you in the feminine was another incident of my passive-aggressive tendencies and you are miffed by my slight. You confront me about my slight and express that you are offended. I respond, “Oh, wow…I hadn’t even realized I did that! Whenever I hear the term ‘daddy issues’, I automatically associate it with girls who work in unsavory occupations. My using feminine pronouns in response to your example was wholly unintentional, and certainly never intended as a slight.”
In this case, there is indeed an action (using feminine pronouns in reference to hypothetical you), but there is no attitude or evaluative judgment with regard to you that corresponds to that action. Therefore, there would be no attributability in this case. (If anything is attributable to me, it is with regard to my associating “daddy issues” with “females with unsavory occupations”.) In the previous example, I’m hard-pressed to see how it isn’t the case that the action of dumping girls is not accompanied by both a corresponding attitude and evaluative judgment (even if that judgment is perhaps falsely rendered under the heavy influence of daddy issue-induced fear).
I would be tempted to point to weakness of will cases too. They seem to force the view to (implausibly in my view) say we are not responsible for our weak of will actions or (threatening to really water down the view) to broaden the connection between our evaluative point of view and what is attributable such that weak of will actions count as sufficiently flowing from our evaluative point of view.
Lori: I don’t want to say that the target of attribution is no longer me; indeed, that’s the main point, that it seems I’m still the appropriate target of predication in light of continuing attributability. But now it no longer seems as if I’m the appropriate target in virtue of the action’s depending (ultimately) on my evaluative judgments. Rather, it’s in virtue of depending on something else, something noncognitive/affective, something that may operate independently of my evaluating self but still part of the subset of psychic elements expressive of me qua practical agent. Or perhaps I’ve misunderstood you.
Sobel: Watson wrote a follow-up paper to “Free Agency” called “Skepticism about Weakness of Will” in part to try to figure out how weakness of will might fit into the evaluative judgments picture (without counting as compulsion, say). Part of the problem was that he was still working with a notion of “evaluative judgments” understood as “judgments of what’s best.” But since then talk of evaluative judgments has turned to talk of “judgments of value/worth” where this doesn’t have to consist in “best” judgments. Couldn’t something like this account for weakness of will, then, without watering the view down too much, i.e., our weak of will actions may still be attributable to us in virtue of their dependence on a judgment of worth, albeit not what’s best?
Interesting. Most cases of weakness of will would involve some temporary thing judged worth having, the only problem being long term effects or the short term benefits being outweighed. I would have to think about the possibility of still counting as willing something that you think is good in no respect.
Perhaps one cannot count as acting unless one thinks that which one is aiming at is in some respect good. If that is false, that seems like bad news for the view you are working with. If it is true, … well I need to think more about that option.
Hi Dave,
This is not against the thrust of your substantive point – I am sympathetic to your argument and related ones made by Sher, but on re-reading “two faces” by Watson, I am not clear that he intends there to build the evaluative view of responsible agency (of the sort you seem to have in mind) into the idea of attributive responsibility. In fact, I think there is evidence he is not building it in, and might be committed to doing otherwise.
His examples of attributive agency include, for example, someone who foolishly gets drunk and does not get enough sleep on the day before a personally important performance. His point is that in describing the person as doing something foolish, we are attributing the action to him. And this involves linking the action to some aspect of the person’s character. But presumably the foolish person need not be thinking/judging that it is good to get drunk and not get sleep, so it seems odd to think Watson is assuming that a strong evaluative view is the best way to understand character in this context. On the contrary, it is natural to assume that the foolish person might be negatively judged because he fails to reflect and form a judgment and plan about how to best spend the night before the performance.
I suspect that the virtue judgment (“he is foolish”) is best understood as a judgement about how the agent’s actions, and the characteristic features that explain it, fare relative to a virtuous standard, and that this need not assume a rationalistic “evaluative judgment” account of character.
A second piece of evidence (in addition to the other examples he uses): in the appendix Watson indicates we should be catholic about the aspects of the “self” that make up character. He mentions, “purposes, ends, choices, concerns, cares, attachments, commitments”. This list includes some items that are on the non-cognitive end of the spectrum.
It is true that he wants the character that grounds attributability to involve “will” in the sense that it is something that less reflective animals (he mentions dogs) lack, but I don’t think this needs to involve the adoption of the strongly rationalistic view.
Hi David
suppose that
1) some non-evaluative element of my psychology is causally necessary for some action to be performed.
This fact is consistent with
2) some evaluation is also causally necessary for the action to be performed.
2) may be true either because evaluation may be caused by the non-evaluative judgement, and such causing may be necessary to produce the action, (part of Lori’s point) or because independently caused evaluative judgement is also causally necessary to produce the action.
In your example, 1) seems as though it might be true, but it is not clear that 2) is false. Defenders of the evaluation view can accommodate your case by showing that 2) is true in your case (which i take it is what Lori and Brad might be getting at in part).
In common with Brad’s reconstruction of Watson, though, I think that a person can be responsible for an action where the presence some vice that is not connected with evaluation of the action explains why the action is performed. This might also be true in your case. For example, suppose that had you been more reflective, you would not have responded to your daddy issues. You are responsible for the dumping.
But I take it that what your post is intended to explore is not really related to these things. I take it that you want to explore the following question:
It is generally true that if D judges that v is valuable and this judgement causes D to v, D is responsible for ving. Is it also true that if D has non-evaluative features of his psychology (daddy issues, for example) that cause D to v, D is responsible for ving in virtue of the fact that these features of D’s psychology play the same, or a similar, role in responsibility judgements to evaluative judgements.
Is that right?
Thanks, Brad: Helpful as always. I agree that the strong evaluative view expressed in “Free Agency” is not necessarily attached to the attributability stuff in “Two Faces.” Indeed, I’m trying to preserve many of the insights of “Two Faces” precisely by prizing attributability apart from evaluative judgments. Perhaps, then, the view I’m responding to should be called “Watsonian.”
Nevertheless, at most we can say that the account of attributability in “Two Faces” is ambiguously put. There are the passages you cite, but there is also this central one: “If what I do flows from values and ends, there is a stronger sense in which my activities are inescapably my own: I am committed to them. As declarations of my adopted ends, they express what I’m about, my identity as an agent. They can be evaluated in distinctive ways…because they themselves are exercises of my evaluative capacities” (270-271 of AA). And this sort of talk lends itself to reinterpreting the examples you’ve given as falling under the evaluative rubric: the drinking was foolish insofar as it depended on a bad evaluative judgment; and one could rather easily construe the items on the “self” list you cite in evaluative terms.
Victor, you’ve pretty much nailed the issues in your final paragraph, although at this point I merely want to establish attributability of, not necessarily responsibility for, some actions or attitudes that are not dependent on evaluative judgments. Whether attributability of this sort is constitutive of one conception of responsibility is a further question (albeit one to which I heartily believe the answer is Yes!).
Thanks to everyone for your (as usual) excellent comments and advice. I will be in and out of contact for the next few days (it’s Jazz Fest time in New Orleans — Springsteen on Sunday!), but I hope some will continue the conversation in my absence, and I’ll catch up soon (if the conversation hasn’t yet grown stale).
Dave,
I wonder if the evaluative judgments view of attributability might respond to your counter-example by appealing to a dual-process model of evaluative judgments. The idea would be that, though the evaluative judgments which render certain actions attributable to me are arrived at via a rational process of deliberation, at least some of the time, there’s no reason to think this is always the case. We have (I think) good evidence that at least some of our evaluative judgments occur without the aid of our rational deliberative neural architecture.
In the Daddy Issues case, then, the spurning action is dependent on an evaluative judgment of a social intuitionist sort, but not on a process of rational deliberation that precedes or reenforces it. Thus, it is attributable in just the way other actions are attributable–it is dependent on one’s evaluative judgment. What we can’t say is that one arrived at the relevant evaluative judgment because she deliberated about it.
Presumably, we can sometimes say this about other of our evaluative judgments, so the case puts pressure on the evaluative judgment view without amounting to a full-fledged counter-example.
I’m not sure I agree with this sort of spin, but I wonder if it’s an option that is at least available to the evaluative judgment position, as a way of accommodating the case.
Hi, Eli, sorry for the delay. The thread may be dead. At any rate, I think this spin puts pressure on the coherence of the evaluative judgments view, which is supposed to be one from which reasons flow, and if the judgments are produced via some social intuitionist model, it’s hard to see these as reasons (a) we should pay attention to, and (b) that flow from or are expressive of the self. It’s also not clear to me what the relevant evaluative judgment is in the Daddy Issues case. I want to say that, while the judgment that those spurned had various sorts of faults, the cause of my spurning them was actually my fear of abandonment (or some such), which isn’t (in the case as given) expressed via any evaluative judgment. That is to say, the so-called judgment masked the real cause, which was the fear, which was nevertheless expressive of the self.