For a very long time, many philosophers thought that knowledge is justified true belief. But Gettier pointed out how one can have justified true belief and yet lack knowledge.
For a very long time, many philosophers have thought that performing a virtuous action is performing the right action for the right reason. Here I wish to point out how considerations analogous to those that Gettier identified show how one could perform the right action for the right reason, and yet not perform a virtuous action.
To make this analogy clearer, I shall specify which elements in the practical case function similarly to the more familiar elements in the original epistemological case. First, just as beliefs can be true or false, acts can be right or wrong. Rightness is the success condition of acts, just as truth is the success condition of beliefs.
But it is possible to believe truly without justification—one might just be lucky. It is also possible to believe falsely but with justification—one might just be unlucky. The same holds for the practical case. It is possible to do what’s right, even though one does not so act for the right reason; Kant’s example of the prudent shopkeeper comes to mind. And, it is possible to do the wrong thing, but be doing it for the right reason; think of the doctor who blamelessly gives her patient the wrong medicine. So, right action and justified action can come apart. The vast literature on moral luck discusses cases such as these.
But in cases where luck is not a factor, it appears that one acts as well as one can (viz. one acts virtuously) if one does the right thing for the right reason. Is this really so?
In Gettier cases, Smith has a justified false belief that p, has a justified true belief that p implies q, and so with justification infers that q, which happens to be true. Smith thus has a justified true belief that q, and yet we are reluctant to say that Smith knows that q, since Smith’s belief that q is based upon a false belief.
Now suppose Jones is E-ing, and is doing so for the right reason, but as it happens, E-ing is not really the right thing to do. (Let’s suppose E-ing is not evil either; it is merely suboptimal.) Jones is thus unlucky. Jones is also taking the necessary means to E-ing; in other words, Jones is M-ing in order to be E-ing. And suppose Jones is under no illusion about how to attain E; M-ing really is the necessary means for E-ing. Since Jones is E-ing with justification, then Jones is also M-ing with justification. (Practical justification is transferred from ends to means much as theoretical justification is transferred from premises to conclusion.) But M-ing just so happens to be the right thing to do—not because it is the necessary means for E-ing, but because it is (also) the necessary means for F-ing, where F-ing <> E-ing. Thus Jones’s M-ing is the right action, and Jones is M-ing with justification; she is M-ing for the right reason. Yet we are reluctant to say that Jones’s M-ing is a virtuous action, because her M-ing is aimed at an end that isn’t right.
To focus our thoughts, let’s consider a specific example. Suppose Jones has a spare $100, and she aims to donate it in a way to benefit the disadvantaged the most. I assume that acting out of concern for others is a virtuous motivation, and acting from this motive is acting with justification. (Egoists can suitably alter my example to make it more plausible to them, I suppose.) Jones hears that UNICEF is the only provider of mosquito netting to those threatened by malaria, and so writes a check to UNICEF to support their mission. Let’s imagine that UNICEF is indeed the only provider of mosquito netting. But purchasing mosquito netting is not how Jones can most benefit the disadvantaged. So financially supporting UNICEF’s efforts to provide mosquito netting is not the right action; she could in fact most benefit the disadvantaged by providing those threatened by cholera with water purification tablets. This would be the right action. But it just so happens that UNICEF is also the only provider of water purification tablets to those threatened by cholera; let us suppose for the sake of the example that UNICEF is the sole charity for both projects. Thus writing a check to UNICEF is also the necessary means to the right action.
Now when out of concern for the poor, Jones writes the check to UNICEF, she does so with justification, even though supporting UNICEF’s mission to provide mosquito netting is not in fact the right thing to do with her money. But writing a check to UNICEF (her means) is itself a right action, since it is the necessary means for most benefiting the poor and disadvantaged. So in writing the check to UNICEF, Jones does the right thing for the right reason. And yet her act is not fully virtuous, since what justifies her action is not the same as what makes her action right, just as in the standard Gettier case, Smith’s justified true belief is not knowledge, since what justifies Smith’s belief has nothing to do with what makes Smith’s belief true.
So, one can do the right thing for the right reason, and yet not act virtuously.
Plausible? Original? Feedback!
Eric, cool idea. I’m not sure I’m getting how the UNICEF example is an example of the schema you want. In your schema, you’ve got an act that’s wrong, and an act that’s right; the act that’s right is a means to doing the act that’s wrong (but also a means to doing a different act that’s right). But in the UNICEF example, there’s just one act: sending $100 to UNICEF. You seem to be saying there’s another act Jones does: supporting the mission to provide mosquito netting. Is that really a different act? It seems like sending the money is the only thing Jones *does*.
Eric,
Two nits:
In Gettier cases, Smith has a justified false belief that p, has a justified true belief that p implies q, and so with justification infers that q, which happens to be true. Smith thus has a justified true belief that q, . . .
I think this is too specific because there are Gettier examples not like this. Maybe I’m wrong, but aren’t true barn beliefs in fake barn country supposed to be Gettier cases if they are merely lucky?
And, you write:
And yet her act is not fully virtuous, since what justifies her action is not the same as what makes her action right, just as in the standard Gettier case, Smith’s justified true belief is not knowledge, since what justifies Smith’s belief has nothing to do with what makes Smith’s belief true.
I’m not sure this is the best way to put it, because you can get Gettier cases where the justification has something to do with what makes it true (it looks like a barn because it is a barn — though in fake barn country) and I’m guessing we can construct counter-examples running the other way too, though I’m not coming up with one that convinces me off the top of my head.
Ben,
You raise a good point, and I’m sure my example can be greatly improved. The answer to your question depends upon the identity conditions for acts, a tricky topic.
But I think it’s pretty clear that there are examples where you are both M-ing and E-ing, where M-ing is not equivalent to E-ing. For example, right now I am typing. And I am typing in order to respond to you. And if my internet connection (and other conditions) holds, I am indeed responding to you. But if my connection should fail, then while I am typing, I am not responding to you. So (if my internet connection holds) I am doing two things, not one.
Mark,
I will have to think about that. I think I characterize Gettier’s original cases accurately, but there may be other sorts of cases in which Kp <> JTBp that I do not.
Very clever. I don’t get the intuition that Jones does not act virtously. He aims at benefitting the worst-off. He does this by usually reliable way even when the way in which the worst-off get helped is different than what he expected. My intuition is that he was virtuous.
I might even think that his actions were right, justified and virtuous when everything were the same but due to some freak accident UNICEF lost his check and no worst-off were in fact helped. My intuition is that rightness of action like justification is not tied to the actual consequences of the action but rather to expected consequences. So, if there is good reasons to believe that UNICEF helps the worst-off and they usually do and person donates money with that end, then the action is right, done for the rigth reasons and virtuous (even when the actual consequences come apart from the expected end of the action). If this is right and the right-makers are in the expected consequences, then the right-makers and justification do not come apart.
Eric, well done. Let me check and comment on your claims:
1) “Jones’s M-ing is the right action,” that is: “Jones’s M-ing is the best she can do, because by M-ing Jones Fs, and F-ing is the best thing to do”
2) “Jones is M-ing with justification; she is M-ing for the right reason.”, that is: “Jones Ms in order to be E-ing, and E-ing is good enough to justify Jones’ M-ing”.
3) “Yet we are reluctant to say that Jones’s M-ing is a virtuous action, because her M-ing is aimed at an end that isn’t right.” This is not clear. I’d rather say:
“The problem is that Jones by M-ing does the right thing (the best thing) (F-ing) for a reason that is good enough, but is not the best of reasons (in order to be E-ing rather than in order to be F-ing).” Prior to virtuousness, I think there is a question about reasons. Jones’ behaviour appears to be defective in some such respect: she does what there is most reason to do, but her reason, while good enough, is not what makes what she does the thing there’s most reason to do. Is this redescription correct?
Then there is the question of what we should say of such a predicament. This, I think, depends on further facts about Jones: a) Does she know that M-ing is necessary to F-ing? If she doesn’t, and there’s no way she should be expected to know that, then she’s just fine practically speaking, she’s just epistemically not perfect. b) Does she know about F-ing, yet somehow believes that E-ing is better than F-ing? If so, then she fails in her practical beliefs, and in this sense she is not perfectly (substantively) rational, or if you want, morally wise, and thus not (fully) virtuous, assuming that (full) virtue includes having true moral beliefs. In conclusion, I would describe the case more in detail to get the result you want.
Eric wrote: “But M-ing just so happens to be the right thing to do—not because it is the necessary means for E-ing, but because it is (also) the necessary means for F-ing, where F-ing <> E-ing. Thus Jones’s M-ing is the right action, and Jones is M-ing with justification; she is M-ing for the right reason. Yet we are reluctant to say that Jones’s M-ing is a virtuous action, because her M-ing is aimed at an end that isn’t right.”
I’m a bit puzzled: isn’t it true that Jones’s reason for M-ing is that M-ing is necessary for E-ing? Wouldn’t the right reason for M-ing be that M-ing is a necessary means for F-ing? If this is correct, then why should we think that in the case described Jones is M-ing for the right reason?
One might think that in the case described, Jones has a reason for M-ing—namely, the fact that M-ing is a necessary means for F-ing—but this is not the reason for which she acted. One might also think that Jones has a reason to believe that M-ing is the right thing to do, because we are assuming that Jones justifiably (though falsely) believes that E-ing is the right thing to do, and also justifiably (and truly) believes that M-ing is a necessary means for E-ing. Yet, to repeat my question, why should we think that Jones is M-ing for the right reason in this case?
Eric wrote: “Now suppose Jones is E-ing, and is doing so for the right reason, but as it happens, E-ing is not really the right thing to do”
Maybe this is where I get confused: Is it possible that Jones is E-ing for the right reason if E-ing is not the right thing to do? One might think that E-ing for the right reason is E-ing in light of the feature that makes E-ing the right thing to do. But in this case, by stipulation, there is no fact that makes E-ing the right thing to do, and consequently, E-ing for the right reason seems impossible.
Eric,
I think this argument is bad:
Compare:
But the conclusion is false. I am in fact alone in my office.
I know this is slightly off-topic but here’s another counter-example to the proposed analysis of virtuous action in terms of right action done for right reasons. Doesn’t really have anything to do with Gettier cases but at least I get the right intuition.
Smith is a mob-hitman. When he gets a job from the Godfather, without a further thought he carries it out to the perfection. He doesn’t care about his victims at all. One day he is ordered to kill Jones. As he is going to get him, he suddenly gets a flash-back from his childhood – a happy memory about being with his father. When he comes to Jones’s house, he sees Jones playing with his children. He realises how much they would be hurt if he killed Jones. For this reason, he decides not to kill Jones. Instead, he helps the family to escape the angry mob on a personal cost and by putting his own life under threat. He succeeds in this. When he goes on to kill his next victims, he has no problems whatsoever in excecuting them. Smith is relieved for this.
My intuition is that here is a case where Smith does the right act for the right reasons but I wouldn’t call the act virtuous. It was a jolly decent thing to do from a vile person.
Campbell,
Hmmm. I’m not sure that the argument you offer is indeed analogous to mine. But let me approach this in a slightly different way.
Now I am typing; of this, I am fairly certain. But my wifi crashes from time to time, so I wonder whether I am ALSO responding to you. Responding to you thus seems to be a different act than typing, even though I am responding to you (if I am) by typing.
I see that my argument here might not be sound. I might wonder whether water is H20, but this doesn’t show that water is different from H20. I will think about this some more. Thanks.
But back to the original point, I think I don’t need to worry about Doug’s reply in cases where M-ing is a necessary means for E-ing, but having Med is not sufficient for having Eed. For then, we clearly have two acts.
Example:
“What are you doing?”
“I am making bread. See, now I am mixing the flour and yeast.”
“That’s all you have to do to make bread?”
“No, I also need to put the dough in the oven.”
So one can be mixing the flour and the yeast, and one can be making bread, and that there are two acts here.
Uri,
You identify what I too think is the problem with my argument. To say that Jones acts for the right reason is ambiguous. It may mean that Jones acts from the right motive (e.g. charity, rather than a desire for social approval). Or it might mean that Jones has the right end (e.g. financing the delivery of water purification tablets). So, in one way Jones acts for the right reason, and in the other way she does not.
Even so, the two aren’t always easy to distinguish. Jones should finance the delivery of water purification tablets because that’s the most effective way to benefit the disadvantaged–it is a means to this further end, an end that seem to be partially constitutive of having charitable motivations. But this is an end that Jones has; she just wrongly thinks that she can attain this end by financing the deliver of mosquito netting. So Jones does have the right ultimate end, as well as the right motive (charity, kindness).
Eric,
Here’s another kind of example, I wonder if it is a practical Gettier case in the sense you intended?
In the new collection “McDowell and his critics”, in McDowell’s response to Dancy, he discusses the following pair of examples, which sound Gettierish to me.
In both cases a person, call him McD, believes that there are carpenter ants in the garage roof because an expert tells him there are. In the first case the expert is lying (the expert thinks that the piles of sawdust originate from a do–it–yourself– project), in order to direct extra business to his cousin the exterminator. But the expert is wrong, the piles are in fact made by carpenter ants.
In the second case the expert knows there are carpenter ants, and McD acquires this knowledge from the expert. In both cases, McD calls the exterminator, and the reason for acting in both cases is that there are carpenter ants in his garage roof.
So, the right action for the right reason, but some relevant difference along the way. [I think McDowell’s point is that in the first case, the fact itself “did not exert rational influence” on McD’s will, in the second it did.]
Jussi,
could your example suggest that there is a requirement that a certain (moral) reason RELIABLY motivates one’s actions in order for one’s right actions done for the right reason to count as virtuous? This would give some sort of parallel with the epistemology issues. Or maybe something more is needed?
Francesco,
I think that must be one requirement for virtuous actions. It’s not just that you do the action once for the right reasons but usually overlook those morally significant reasons.
Another thing is that doing the right action for right reasons doesn’t seem to always be enough morally speaking for virtuousness of the action. Some right actions like not killing a person usually aren’t even in the business of being virtous actions (special circumstances apart). In many cases, right actions are something very low-key and we have many options that all would be right.
As I am typing this, this typing hopefully is an action that is the right to do, as would be writing emails to various people, going to eat something, getting home to chill, seeing friends, and so on. All of these actions can also be done for good and right reasons. But, they wouldn’t count as virtuous actions. Virtuous actions need to be morally good in some sense that goes beyond rightness.
I think these considerations already ground the required counter-examples to the proposed analysis without the Gettier considerations that do not seem to create the required intuitions.
So, the view we would get so far is that a virtuous action is an action that is both right and morally good and done for the right reasons by an agent who has a stable sensitivity to that kind of reasons. We then want a counterexample against this view. Probably there are such. We can even assume that Eric’s case satisfies these conditions. Even then, the case where UNICEF ends up aiding the worst-off most in a slightly different way than in which the agent intended doesn’t seem to make me want to say that the action was not virtuous.
I thought that I just had abnormal intuitions about this. However, I’ve told the story to various people and no-one I met so far has had the intuition that Jones’s action was not virtuous. If you run the original Gettier cases one people, you get the right intuitions even from the ‘folk’ (less so in Asia but that’s another story).
Jussi,
Why think that a virtuous act must be done by someone with stably virtuous dispositions? In your hitman case, I think his act is virtuous. The hitman isn’t virtuous, but even vicious people can occasionally do virtuous things, I’d say. I think Tom Hurka argues essentially the same thing in a recent paper in Analysis. I remember thinking he was right.
I think Campbell’s right that there’s a flaw in the argument. As he points out, it’s fallacious. The general form seems to be this: the x(Fx) might not have been G. Therefore x is not the y(Gy).
A fallacious argument doesn’t mean a false conclusion: perhaps there are independent reasons to do with the individuation of acts that would make these different acts. However, if I’m right in thinking that the Davidsonian view is still the orthodox view here, then the suggestion runs contrary to orthodoxy. Both act descriptions in this case pick out the same basic action or physical motions (the typing), and hence refer to just one action — just as the flicking of the switch and the alerting of the burglar in Davidson’s example are the same action.
It seems you want the virtue of an act to be non-accidental, and you want the justification for an act to be tied in some way to doing the right things. Although your account isn’t Kantian, I think Herman’s “no accident” test might be informative if you are fleshing out a reliability requirement. She contends that a dutiful action has moral worth iff its performance was no accident. She further couples this with a ‘guarantee’ requirement, such that an act has moral worth only if another agent acting from a morally similar reason will perform a morally permissible act.
Great idea, Eric. I agree that your example isn’t that clear. May I suggest a new one?
As I understand the Gettier examples, the point is that justification is severed from the appropriate link with the truth of the belief. The belief in the examples is justified but it’s just lucky that it’s true.
So, in the moral case, we want an act that is done for a good moral reason, but where it’s just lucky that it turns out to be the right thing to do. Try this:
Alfred demonstrates laudable beneficence. Every time he sees someone in need, he writes them a $100 check. However, Alfred doesn’t realize that he’s always failing to do the best thing when he does this; Alfred is a cancer researcher and, if he had kept his money, he would have been able to self-fund some important research which would enable a cheap, efficient cure for cancer to be discovered twenty years before its time. Every time Alfred writes a check, he postpones the potential cure for cancer, and thereby is doing the wrong thing.
One day, though, Alfred writes a $100 check to Bill, out of his customary beneficence. Bill had loaned Alfred $100 without interest back when Alfred was in grad school, then both of them completely forgot about the debt, which was never repaid. Alfred’s action was the right thing to do, because Alfred had an outstanding obligation to Bill (consequentialists can change the example to say that Bill will use some of the money for something even more consequentially positive than finding the cancer cure).
Alfred’s beneficent motive in writing the check to Bill is preventing the cure for cancer being found. So Alfred’s action, although done for a good moral reason, and the right thing to do, wasn’t virtuous.
I’m not sure how strong my intuitions are on this, but it’s the closest parallel to the Gettier cases I can think of! Maybe, as Jussi suggested, it’s hard to separate the notion of the right act from the notion of having good moral reason, or all-things-considered moral reason, to do it. This would distinguish it from the case of truth and reasons for belief in a way that would prevent Gettier cases being applicable.