Julia Driver

Julia Driver (Dartmouth), "Luck," with commentary by Hans Maes (The University of Kent).  Both the paper and the commentary can be found here.

Comments

[Jussi Suikkanen]

Julia,

I wonder if there is conceptual room for the kind of proposal you seem to defend without revising the meaning of wrongness. I'll try to explain.

You seem to think that the criteria for acts' wrongness are 'external' to the agent - for the consequentialist the criteria are that the best consequences were not actually brought about. This creates the luck problem. Strange things happen and suddenly you act wrongly in an everyday situation doing normal acts like everyone else. I take it that the solution is to separate blameworthiness from wrongness. While the latter is 'external', the former is 'internal' - agent's intentions count towards blame. For the consequentialist then, perhaps the agent is blameworthy if she wasn't intending to bring about the best consequences. Thus, the agent can act wrongly but not be blameworthy or be blameworthy but do the right thing. I think Michael Smith from the other thread has somewhat similar thoughts (in opposition to Frank Jackson's view).

Now, my worry is that to say to someone that 'she acted wrongly' comes awfully close to expressing blame. This seems to happen either by implication or the blame might even be part of meaning of 'wrong'. I guess I'm referring to an authority here. According to Scanlon, 'to call an action morally wrong is to say that it violates important standards of conduct and *is therefore open to serious criticism*'. There just seems to be something platitudious about that. And, if it's right, then there just is no room for an account according to which acts can be wrong and not to be blamed for. That would be a commitment to a contradiction - you are already blaming in the antecedent.

Maybe you have different semantic intuitions here. But, if 'wrongness' does not implicate blame but describes the consequentialist property that is independent of the assessment of agents, then I'd like to hear more about to what purpose that predicate is used in our moral community.

[George Rainbolt]

Julia,
I really enjoyed the paper (and the comment as well). A couple of questions.

1. Following up on Dr. Maes second comment, what precisely is fortune and precisely how is it different from luck?
For what it is worth, Websters.com says that “LUCKY stresses the agency of chance in bringing about a favorable result 'won because of a lucky bounce'. FORTUNATE suggests being rewarded beyond one's deserts 'fortunate in my investments'.”
You seem to suggest, p. 17, that “fortunate” indicates that a good outcome was inevitable and beyond one’s control. Is that the way you would go? My semantic intuitions are slightly more in line with Webster’s. But I am not sure that, in ordinary language, there is much of a distinction here.

2. I found your case against epistemic reductionism very convincing. But then on p. 18, you adopt an epistemic understanding of relevant initial conditions. Doesn’t this open you up to all those objections that your raised against epistemic reductionism? In particular, aren’t you going to need to pick out a level of idealization?

(A note re the online format. It makes one more reluctant to ask a question for fear that it will turn out to be foolish. I fear that this question may overlook something obvious. In a usual conferences, my words vanish into the air only recorded in the imperfect memories of those in attendance. Here a foolish question lives on in black and white. Perhaps this will increase the quality of interaction but it might inhibit it as well.)

[Jussi Suikkanen]

Sorry,

made a small mistake in describing your position. What is relevant for blameworthiness is not whether the intentions do not aim at maximizing good but rather whether they are such that they do not tend to maximize good. Usually these two I take it are coextensive. Anyway, the view less internalist than I put it. Still, wrongness and blaming seems to come apart in a way that's counterintuitive for me.

[Julia Driver]

Hi Jussi --

Thanks for your comments. I think your linguistic intuitions are probably shared by a large number of people who find objective consequentialism odd. However, the other side of the case is this -- we sometimes do quite naturally say things like "She did the wrong thing -- but I don't blame her (because her motives were good, or she was making an appropriate attempt...)". This indicates that there is *a* distinction to be made here. Generally, in the literature, the debate seems to center on what the *priviledged* sense of 'right' is to be -- the one using an external standard or the one using an internal standard. But the fact that the above utterance seems sensible to me indicates that there is certainly a distinction to be made.

[Julia Driver]

Hi George --

Thanks for your comments too --

The distinction between luck and fortune was made because the account I end of favoring is a version of the modal account, and there are some cases where a person may receive an intuitively undeserved benefit, even though this isn't due to a fluke or accident. Thus, that is 'fortune' as opposed to 'luck'. I'm not sure this is reallly an important distinction at all for understanding moral luck, though -- it is really just an off-shoot of the account of luck itself.

On your second comment -- yes, I'm rejecting epistemic reductionism. But the view I end up adopting is attempting to accommodate the intuitions that seem to motivate epistemic reductionism. Namely, that luck judgements at least in many cases make sense when we consider the attributions relative to the epistemic states of the attributor or the person to whom the luck is being attributed. In the moral case, it seems likely that the epistemic states that seem relevant to us are those of foresight and/or intention. I don't give an independent argument for this in this paper -- it is in the book, though -- and I remain neutral in this paper on level of idealization. That's something I have to think further about, but my gut reaction at this point is to say that in the case of foresight it has to be what the agent can reasonably foresee.

[Jussi Suikkanen]

I wonder if the sensibility of that utterance could still be explained without going as far as the externalist picture.

So, you might say that a part of the meaning of wrongness is that others are warranted to blame the wrongdoer as Scanlon says. Blame itself, not blameworthiness, could then be attached to wrongness claims by conventional implicatures that can be cancelled with utterances like the one you give. And, given that you are only warranted to blame others when they have acted wrongly doesn't imply that you *have to* do so. You might have some better reasons not to do so. By default you might do so, hence the implicature.

[Julia Driver]

Hi Jussi -- That's an interesting suggestion, and I'm not sure that what you are suggesting would be incompatible with my view. It would be okay to hold that when we say that someone did something wrong, there is a presumption of blame unless the genuine conditions of blame are not satisfied. This is completely compatible with my view. So, there is generally a connection between having done something wrong and being blameworthy, because often when people do something wrong the conditions of blame are satisfied (whatever they are). However, you seem to hold the view that "...you are only warranted to blame others when they have acted wrongly..."; but I have doubts about this -- suppose that someone, out of some kind of vile motive, -- perhaps sadistic pleasure in the suffering of another -- kills someone who, let's imagine, ought to be killed -- someone who is a powerful genocidal maniac, let's suppose, who simply can't be stopped any other way. Again, the sadist may have done the right thing, but for the wrong reasons. We may be very reluctant indeed to praise such a person, and, indeed, we might even blame him for acting out of such a motivation.

But I also don't want to committ myself to the view that the picture I am suggesting is the *only* way to account for the distinction -- or, at least, the intuitive plausibility of utterances like "She did the wrong thing, but I don't think she is blameworthy." But the objective consequentialist does have a nice way to handle it. It is an extra consideration for the view.

[Jussi Suikkanen]

Julia,

Sorry didn't make myself clear there. I think you misuderstood the 'only' in my sentence. My mistake. The purpose was to say that you are only *warranted* to blame someone when they have acted wrongly instead of *required* to blame. Thus saying to someone that they have acted wrongly means only to say that they are blameworthy and is not to express blame directly via linguistic meaning. The blame itself may only be implicated. I agree that there are other grounds for blame too.

I guess my question for the objective consequentialists, which doesn't tie blameworthiness to wrongness in any way, is that if there is such a sense of 'wrongness' is it one that is ever used in ordinary language? So far I have seen no support for this. Can you use the utterance 'that was wrong' appropriately to someone and expect that they don't feel blamed? Wouldn't you feel blamed if someone accused you of wrongdoing? If the answers to these questions are 'no' and 'yes', then the externalist is rather defining a wrongness* than illuminating the wrongness we talk about.

[Julia Driver]

Hi Jussi --

Thanks for the clarification of your comment. Again, I'm finding your points very interesting.

It's not that the objective consequentialist must provide an account that doesn't tie blameworthiness to wrongness in any way. It's just that the oc notes that there is a distinction between doing something wrong and being blameworthy. The oc can hold that usually when one acts wrongly one is blameworthy, for example. That's because, usually, when one acts wrongly one has intended to do something bad; or one has been poorly motivated; etc. But not always. I think you are attributing a much stronger claim to this view.

That having been said, what do you think of utterances like: "It turned out to be the wrong thing to do, but of course you had no way of knowing that." In understanding how a person might react to such an utterance, directed her way, it would probably be useful to distinguish between remorse and regret. It would be very natural to feel bad -- and in fact that is probably useful if it makes one more attentive in the future -- but regret, not remorse is appropriate. If there were blame present, then remorse would be appropriate.

[Antti Kauppinen]

A minor point: Jussi, you must be thinking of conversational implicature, not conventional, since Grice defines the latter as implicature that can't be cancelled (though it can be detached and isn't a part of the truth-conditions). I guess the most plausible suggestion would be that we're dealing with a generalized conversational implicature, like the temporal (and causal) meaning of 'and' in the sentence "What happened during the hunting trip was that Cheney shot Whittington in the face and Whittington cried in pain". The implicature is that Cheney shot first and Whittington cried then (as a result). Unlike an implication or a conventional implicature ("he's Russian but honest"), this can be cancelled without contradiction or linguistic inappropriateness: "-- I mean, first Whittington cried in pain, and then Cheney shot him in the face to shut him up".

[Jussi Suikkanen]

Antti,

thanks for the help. I always get these things confused without the book.

Julia,

Well, the connection will be somewhat contingent in any case. After all wrongness and blameworthiness will be attributed by assessing different things (consequences of acts/intentions) with different criteria (their actual value/tendency to bring about good). It may be that this in actual world overlap. But this we could only know a posteriori especially in each individual case. I just wonder if we actually know a priori that if someone has done something wrong we are warranted to blame her.

About the utterance. I have hard time finding a natural example of that case in which moral wrongness is discussed. If I'm in a position to say that some act turned out to be morally wrong, then I assume that my moral-epistemic position is such that the other agent could have quite easily have reached it. I don't take myself to be much of a moral expert who would have some moral knowledge others could not have pretty easily. And, I'm quite sceptic whether such moral experts exist.

[Julia Driver]

Jussi --

I wasn't thinking that the plausibility of the utterance "It turned out to be wrong, but of course you couldn't have known it" rested on assuming any moral expertise. I think that when people make such utterances they are using the objective sense of 'wrong.' This isn't an argument for holding this sense to be privileged -- just evidence that such a distinction between the objective and subjective sense can be made, and the objective sense does not in such cases speak to the blameworthy issue. Here what is implied is that the person is not blameworthy at all.

By the way, on the topic of moral expertise -- do you think there are people who are better at making moral judgments -- that is, they tend to be more reliably correct than others on moral issues? There might be, but we might also have very good reason for treating any particular claim to moral expertise with scepticism.

[Jussi Suikkanen]

I guess the situations where I can imagine people using that expression would be where the expression is used in an elliptical way. So, what is not said but given by the context is that the agent for instance had an end she wanted to pursue and the act was wrong one for that end, or something like that. But when I think of acts that are morally wrong - killing, stealing, deceiving, and so on, it's hard to think that for someone these turned out to be wrong but she couldn't have known.

Here's a test case which I think comes from Thompson. You go home and turn the lights on from a switch as you do. By some deviant causal chain that blows up the gas-owen of the nextdoor neighbour who dies immediately. Now, an oc theorist would probably need to say that in this case what you did turned out to be wrong but you couldn't have known it (of course you wouldn't be blamed). But my intuition just is that we couldn't say this - there is no sense of moral wrong that would warrant saying this. Turning lights on just cannot turn out to be morally wrong.

That's a good question. I think it depends. Certainly there are people who are more reliable correct in some moral issues. There are people who are more kind for instance - who are so sensitive that they can tell what kindness requires in the situation better and more reliably than others. I'm more sceptical about expertise about thin notions like wrongness.

[Julia Driver]

I can see what the worry is -- however, there are a wide variety of ways one could go about formulating oc. One way that I think would get around this worry is to opt for doing it in terms of objective probabilities -- i.e. the right action is the action that max. the good (or, is most likely to, objectively). This isn't the only way to do it, though, another strategy would be to have a condition that the act cannot be of the sort that systematically fails to promote the good, etc. Off-hand, I favor the first. However, for my purposes in the paper, either way is fine since I am just pushing the externalist line more generally.

[Nenad Miscevic]

Julia
I am glad you brought out the problem with modal definition when it comes to necessary states of affairs.
Here is a proposal about epistemic luck. Consider a simple situation of miscalculation (or a faulty attempt at proving a theorem) which yield the correct result by fluke: two mistakes cancel each other and a happy end ensues. Imagine now that calculations (or the proofs) are very difficult, and that the calculator (or the theorem prover), call her Jane, is a good mathematician, who normally has reason to trust her capacities. Jane arrives at correct result R, which is further corroborated in application. In fact, it takes a genius to detect two subtle mistakes that have cancelled each other. Then, Jane is a priori justified in believing the result R. She thus has a true justified belief, which is, by most people’s lights not a piece of knowledge. Since the result is arrived at in the armchair, and R is a necessary proposition, we have an example of armchair luck.
The possibility of luck in a priori domain shows that the modal definition as offered in the literature is inadequate, since it relies on the possibility of it being the case that not-p. When p is necessary, such a definition should be supplemented by one pointing to variation in belief, not in the fact believed. For instance, veritic armchair luck, i.e. that it “is a matter of luck that the agent's belief is true” can be captured thus:
The agent's belief is true in the actual world, but in a wide class of nearby possible worlds in which the relevant initial conditions are almost the same as in the actual world—and this will mean, in the basic case, that the agent at the very least forms her belief in the sufficiently similar way as in the actual world—the agent has a false belief.

A further issue is the one that you would describe more as fortune than as luck arises when the agent’s cognitive structure might have differed in a minimal way from the actual one, and beliefs.

[Jussi Suikkanen]

Julia,

I wonder if the fix is so simple. If some kind of determinism is true at least on the non-quantum level, then the objective probability of the actual consequences coming about are even before the act 1. If that's true, then objective probability consequentialism collapses into actual consequence consequentialism, and the same objections apply. In order to avoid this, you'd probably have to make the probabilities less objective and more epistemic (as in your dices case). This is the road to expected utility consequentialism and internalism. Your latter fix sounds like rule-consequentialism, where wrongness is failure to do types of acts that promote good. In that view wrongness has (or can have) a more conceptual connection to blameworthiness.

[Julia Driver]

Hi Nenad -- thanks for the example -- wasn't the physicist George Gamow famous for mistakes that would be 'corrected' by other mistakes? This case does make me have second thoughts about making use of the distinction between luck and fortune, which Pritchard had in his modal account. It may be best to not even concede the fortune point. So, for example, when someone claims that he is lucky to have the parents he has he is just making a mistake, since it could not have been otherwise. Also, since it could not have been otherwise, when he thinks of himself as fortunate and experience whatever relief comes with that reflection, he is mistaken. It could not have been otherwise. Hans also had doubts about a distinction between luck and fortune in his comments. But hidden necessities show that we are not lucky even when we think we are, and the same may apply to good fortune.

But we also have to distinguish such cases from the following kind of case, that needs simply to be paraphrased. Someone at North Carolina State -- I think it was David Austin -- came up with this example when I presented the paper there: suppose that lunch is $4 and I go to the cafeteria and put my hand in one pocket -- and pull out $2; I then put my hand in another pocket, and pull out another $2; and think to myself "I am so lucky that 2+2=4!" even though 2+2=4 is necessary. But these kinds of cases are best thought of as expressing luck that one just happened to have the right amount of money, for example, or that what was in your pocket just happened to correspond to what they were charging at the cafeteria.

But the whole issue of luck and necessity is extremely interesting -- no doubt finer distinctions might need to be made in considering different sorts of necessity, too.

[Julia Driver]

Jussi --

OC has come to mean different things in the consequentialist literature. I guess that partly has to do with the fact that the objective/subjective distinction gets bandied about in a lot of different ways. I think the classic statement of OC in the contemporary literature is in Railton's "Alienation, Consequentialism,and the Demands of Morality," where he writes: "[i]Objective Consequentialism[i] is the view that the criterion of the rightness of an act or course of action is whether it in fact would most promote the good of those acts available to the agent." This trades on the now pretty familiar distinction between decision-procedure and criterion of rightness. OC is then often taken to simply be providing a criterion for evaluation. Now, I don't think that anyone disputes that there is an objective sense of 'right' and 'wrong'. I take it that the dispute centers on the issue of of which is the priviledged sense -- that is, which sense -- objective or subjective -- is taken to be the main concern in ethics. Frank Jackson, for example, argues that it is the subjective sense since it is that sense that speaks to what is going on in the agent's mind at the time of action, because the 'ought' that corresponds to the subjective sense of 'right' is, in his words, "the ought most immediately relevanat to action" ["Decision-Theoretic Consequentialism and the Nearest and Dearest Objection"]
While I agree with Jackson on many issues, on this one I opt to go the route that Railton promoted. What is important to my account is that the 'right' action is the one that promotes the good. Again, there are a huge number of ways that this can be developed. But the crucial point is simply that what makes an action right or wrong is not the agent's state of mind [i]considered completely independently of the consequences generated systematically by that type of state of mind[i]. In my book, though not in the "Luck" paper, I explore different ways to spell this out. One way, which is the one that I favor, is to argue that our moral evaluations are quite nuanced -- that when we look at what people do we are concerned to evaluate that act, but also the person performing the act, or the person's motives or intentions. The correct way to evaluate all of these is along consequentialist lines. Sometimes -- due to differing objects of evaluation -- we get kind of ambivalent evaluations. For example, someone may fail to act rightly, but in ways that reflect well on her character and/or motivational structure. It may be that in some circumstances, for example, evil dictators ought to be killed to save the lives of many innocents. It may also be true that a compassionate person who has a strong aversion to killing in general might be able to bring himself to do it. You may not accept this particular example, maybe it just looks like squeamishness instead of a healthy aversion to killing, but there are lots of other examples along these lines -- In any case, one way to explain the ambivalence is along OC lines -- the failure to act is wrong, but we don't blame the person because in general we approve of the compassionate motivational structure.
The other way to go is to say that the right action is the action that maximizes expected utility (or something along those lines) and that the reason why this is the case is that when agents act so as to maximize expected utility they are more likely to actually promote the good. This is a kind of indirect OC. This, I take it, would be another form of OC that would get around your worry about deviant causal chains leading to bad outcomes. I think this is an extremely interesting alternative that I am considering more in the book -- possibly I might even come to adopt it! But I don't think this explains the ambivalence cases as well. There are advantages to keeping certain modes of evaluation distinct. SC, or an OC that looks a lot like an SC, loses that advantage, at least in my opinion.