In a recent review essay, “Morality and Virtue: An Assessment of Some Recent Work in Virtue Ethics“, David Copp and PEA Soup’s own David Sobel present what I take to be a common understanding of the relationship between right action and virtuous people:
It is facts about the alternatives a person must decide among, including such things as the impact the alternatives will have on people’s ability to meet their needs, that determine what a person ought to do. It is not facts about what a virtuous person would want her to do, or facts about the motives that the person would actually be acting from if she were to do the various alternatives. If someone is drowning, for example, and if you can save her at no risk and at negligible cost to yourself, you ought to save her because otherwise her life will be wasted. It is because a life would otherwise be wasted that a virtuous person would want you to save her (552).
Virtuous people are disposed to perform or approve of actions that are antecedently right (perhaps those that maximize happiness, or which are in accordance with some prior set of duties). Here I wish to argue in defence of an alternative embraced by many (but certainly not all) virtue theorists: that it is the approvals of the virtuous that determine which actions are right (and which states of affairs have various values). In particular, I aim to undercut the main intuitions which appear to support the position of Copp and Sobel with respect to the drowning case.
When we consider the drowning case, the potential loss of life leaps out at us as extremely important. It may seem clear that such a loss of life is a bad thing, and that as we consider the case, we are reacting to this antecedently disvaluable state of affairs. There is no apparent need to appeal to virtuous people and their reactions. But crucially, when we do our moral thinking or respond morally to situations we do not do so as blank slates. We possess various virtues, at least to some degree. We may see torturing puppies as wrong immediately, but this is because we are compassionate, just, and so forth. It is far from clear that when we basically decent humans respond to actions (real or in thought experiments) as wrong immediately, that we are responding to some prior, independent wrongness (or rightness) to the actions themselves, or value in the world. Rather, it is quite plausible to hold that our responses are the result of training and perhaps some natural tendencies to sympathy, etc. [See, for example, Kyle Swan’s recent post ‘Poverty of the Moral Stimulus‘.] And if this is correct, there is no difficulty in maintaining that the virtues are explanatorily basic, even in these cases. We strongly disapprove of the potential loss of life precisely because we are virtuous (to some degree); there is no need to hold that we are reacting to valuable states of affairs that are prior and independent of our reactions.
Still, many will hold that the potential loss of life is fundamentally what makes inaction wrong (and rescue efforts right), and an appeal to the attitudes of virtuous agents misses the point. But note: the virtue theorist can entirely agree that virtuous agents would approve of saving the person precisely because a life would otherwise be wasted. The distinctive claim of the virtue theorist (on the present construal) is that the moral rightness of such actions consists in the fact that virtuous agents would respond postively. The virtuous, given their particular psychologies, will respond negatively to actions because of such features as causing suffering, being a case of lying, and so on; similarly, they will approve of actions that end or prevent suffering, and so forth. We can distinguish between why the virtuous will have the responses that they have, and the moral status of actions (or states of affairs). The virtue theorist can at once say both that, for example, acts of puppy-harming are wrong because of the suffering caused (this will be what concerns the virtuous), and because the virtuous would respond negatively to the action.
Humans are creatures who have evolved over time with a range of complex social practices, shared or common preferences, and so on. It could be a mere descriptive fact about most humans that they care about the suffering involved in puppy-torturing – the mere result of social training and certain common human emotional dispositions. On the current proposal, what changes things – what turns a mere disapproval common among many humans – into moral wrongness is the fact that particular kinds of humans under particular kinds of conditions would share a disapproval of this sort. It is the fact that virtuous agents in particular would have particular kinds of negative reactions that makes such acts morally wrong – rather than merely displeasing to most or many humans. There is thus no need to downplay the importance of the suffering involved in puppy-torturings on the virtue theorists’ view; indeed this suffering will be the focus of the concerns of the virtuous; similar points would apply to the drowning case. The only claim is that these actions, states of affairs, and so forth take on a moral status insofar as virtuous individuals in particular will have certain attitudes towards them.
[Of course, a lot will hang on whether a viable account of the virtues (that doesn’t rely on appeals to right actions or valuable states of affairs) can be provided…]
Hey Jason. I’m still not seeing the view as plausible, so let me ask some flatfooted questions. Is this a fair way to characterize your response to Copp/Sobel: Copp and Sobel are right to say that when I am in the drowning situation, I respond to the badness of the drowning, not to any facts about virtuous people; but that’s just because *I’m virtuous*, and virtuous people respond in certain ways to drownings; so, it really is facts about virtuous people that explain why saving the person is right.
If that’s what you’re saying, then doesn’t the view come down to saying, more or less, that the rightness of my action “consists of” the fact that I myself or others like me would do it?
If the virtuous person explains why it is right to save the drowning person by appealing ultimately to the badness of drowning, isn’t he just mistaken on your view? It’s the attitudes of the virtuous that make the act right. The badness of the drowning could only be a superficial part of the answer, despite your protests.
I also think Copp and Sobel might complain that you’re failing to distinguish justifying explanations from causal explanations. The causal explanation of the virtuous person’s attitude toward the drowning person will involve the sorts of things you mention about upbringing. The justifying explanation won’t. (The virtuous person will justify his behavior by saying ‘drowning is bad,’ not ‘I was brought up to react in certain ways towards drowning’, though he might appeal to the latter if he thought someone wanted a causal explanation.)
The virtue theorist can at once say both that, for example, acts of puppy-harming are wrong because of the suffering caused (this will be what concerns the virtuous), and because the virtuous would respond negatively to the action.
Perhaps Copp/Sobel, weilding Ocham’s Razor, will respond as follows. We have a simpler explanation: puppy-harming is wrong simply because it causes suffering. And, other things being equal, a simpler explanation is a better explanation. If we need to appeal to suffering in order to explain why virtuous people respond as they do, then why not leave them out of the story altogether and explain the wrongness directly by the suffering. What does introducing virtuous folk into the story add to its explanatory power? Suffering gives us all the explanation we need; so let’s just leave it at that.
Jason,
I’m wondering if this view is open to a particular brand of amoralist challenge. You say that an action is wrong because a certain group of people (the virtuous) would disapprove of it (or have other sorts of negative reactions to it). Suppose I do not have a negative reaction to it. Can’t I ask you to explain why I should? If all you can say is that a particualr group of people (or even most people for that matter) have this negative reaction, then I rationally can claim that I do not care about these other people and their attitudes, and so I have no reason to try to instill the proper attitudes in myself.
A related worry is that it seems that on this view amoralism amounts to nothing other than to a statistical oddity rather than anything more robust. But it seems to me that an amoralist is not just statisically odd, but is also morally odd–he does not appreciate what is valuable. On your view, he just does not have typical reactions (or reactions of a certain group).
I’m having a lot of trouble focusing on what is at issue here, but to the extent that I can, it seems to me that a virtue theorist had better say something like this: a virtuous agent responds, not to the badness of puppy-harming, and not to its wrongness, but to the suffering of the puppy that the puppy-harming would cause. Its wrongness, then, could consist of the fact that the virtuous agent would (not do it, protest others’ doing it, whatever). Its wrongness might or might not be a reason for other, imperfectly virtuous agents, to avoid puppy-harming: it doesn’t seem to me that virtue theory is committed either way, so let independent considerations decide the matter.
(On the other hand, I feel sure that I remember agreeing with Sobel when he explained the argument to me once, so maybe I haven’t managed to focus well.)
Campbell, it’s not so clear to me that the ‘simpler’ explanation is simpler. (This is why I always use a safety razor.)
Couldn’t a virtue theorist say that puppy-harmings are wrong because a virtuous agent would respond negatively to them, and that a virtuous agent would respond negatively to them because of the suffering involved? Then, so long as the becauses are transitive (which they here appear to be, anyway), it would also be true that the puppy-harmings are wrong because of the suffering involved. This looks no less simple than the alternative view (that the puppy-harmings are wrong because of the suffering and not because a virtuous person would respond negatively to them).
Gentlemen,
Thanks for your comments! [Jamie actually anticipates the sorts of responses I’d like to make – extra thanks!]
I’ll try to address each in turn, but I’ll start with a point that will be relevant to all. I actually prefer an approach which mixes virtue and ideal observer theories. I’d want to claim that, e.g., an action is right if and only if it would be deemed right (where this is a certain form of approval) by an omniscient, unimpaired, virtuous observer.
Ben – I’d want to say that actual humans could say such things as “I saved that person from drowning because drowning is bad”. I’d just cash the ‘badness’ out in terms the attitudes of virtuous ideal observers. So, to judge that drowning is bad is to judge that an omniscient, fully-informed virtuous observer would disapprove of it (in some particular way).
These ideal observers (and we humans, too), if asked for the basis of their disapproval, could point to the suffering involved in drowning, the loss of future happiness, etc. But there would be no need to appeal to some independent and prior ‘badness’.
I think we also get a justifying explanation here – it is precisely because such virtuous, unimpaired, informed observers would react in these ways that an action has its moral status.
Campbell – My thought is that, in itself, suffering is just a psychological (and perhaps physiological) state. We find suffering unpleasant, we try to avoid it, and many of us are inclined to help others avoid it, and so on. But these are all merely descriptive facts about us. I’d hold that we should care, morally, about suffering, because virtuous ideal observers would do so (or at least – they’d approve of us having such concerns). I don’t think a simple appeal to the presence of suffering itself would explain why puppy-torturing is morally wrong: “So what – the puppies are simply experiencing certain mental states that we often try to avoid. Why does that matter?”. The suffering matters morally because virtuous ideal observers would be moved by it.
Scott – I’m not sure I can adequately respond in a short comment; I think the answer to the problem you pose will have to come in the account of the virtues. Roughly, I’d want to think of the virtuous ideal observers as embodying those traits that we’d come to value, if we were to be unimpaired, and omniscient. [There are a lot more details here, but I think this captures my basic view]. As such, it’s not just that the amoralist is atypical. We can see her behaviour (and attitudes) as flawed insofar as they would be disapproved of by beings who possess traits that humans value (virtues) and are also fully-informed (indeed, omniscient). [At least in typical cases, we find decisions or behaviour based on false beliefs, or in ignorance of relevant facts to be flawed, not just atypical.]
Jamie,
The metric of simplicity I have in mind is just this: the fewer because-claims, the more simple. Or, more accurately: if the because-claims in one explanation E are a proper subset of the because-claims in another explanation E’, then E is more simple than E’. As you point out, the virtue explanation has three because-claims: (1) the virtuous agent would disapprove of puppy-harming because it causes suffering; (2) puppy-harming is wrong because the virtuous agent would disapprove of it; (3) puppy-harming is wrong because it causes suffering. Whereas, the alternative explanation has only (3).
Now, perhaps you didn’t mean to include (3) in the virtue explanation (I may have misunderstood you). But there is a sense in which it’s in there: it’s implied. And, at any rate, it seems that Jason wants it to be in there; see the quoted passage in my previous comment.
Jason, you say:
I’d want to say that actual humans could say such things as “I saved that person from drowning because drowning is bad”. I’d just cash the ‘badness’ out in terms the attitudes of virtuous ideal observers. So, to judge that drowning is bad is to judge that an omniscient, fully-informed virtuous observer would disapprove of it (in some particular way).
This makes it sound like, in your view, “X is bad” is synonomous with “X would be dissaproved of by a virtuous ideal observer”. But then the apparently competing explanations — “X is wrong because it’s bad” and “X is wrong because it would be dissaproved of by a virtuous ideal observer” — are not really in competition at all. They’re one and the same explanation, but expressed in different words.
(Though, I suspect I may be missing something here.)
Campbell,
I think I’ve phrased things poorly. I would not want to include (3) as an additional part of the explanation of the wrongness of puppy-harming. In the passage you’ve quoted, I intended only (1) and (2).
Perhaps this better captures what I was trying to say:
We can claim that puppy-harming is wrong because it causes suffering, but precisely insofar as virtuous ideal observers would respond negatively to such suffering, and moral wrongness is a matter of such negative responses on the part of such virtuous ideal observers.
Does this help?
Campbell,
Thanks. Again, I’ve put things poorly. I don’t want to suggest that “X is bad” is synonomous with “X would be disapproved of by a virtuous ideal observer” – at least not as analysis of the actual term as used by actual people. Rather, I’d admit that this would be a revisionary understanding of morality (and moral terms).
I’m not quite sure what you have in mind with “X is wrong because it’s bad”. Is X an action that is both wrong and bad? Is the badness a form of disvalue somehow intrinsic to the act itself?
Still, I would want to hold that we are to understand such things as rightness, wrongness, good states of affairs (of various kinds), and so on as being determined (and constituted) by various approvals and disapprovals of virtuous ideal observers.
It might help to distinguish between a theory’s fundamental right-making property or properties and its intermediate right-making property. So one could say “It is wrong to A because A causes suffering” and “Causing suffering is wrong because it is the object of the ideal virtuous observer’s disapproval.” In which case for the virtue ethicist the disapproval is what fundamentally makes A-ing wrong. I take it, Jason, that this is close to what you have in mind?
If so, we should focus on the merits of the fundamental principle of morality, especially since this is where consequentialism and deontology are traditionally (and, I’d argue, most helpfully) distinguished, and so where virtue ethics must say something distinctive if it’s to provide an alternative.
If that’s the case, though, this is where it seems to go wrong (for me, and Copp/Sobel). For, it seems more intuitively correct to say that what fundamentally makes torturing puppies wrong has nothing to do with who does or does not approve of it, but, rather, with how it treats the puppies. Or so I’d argue. Furthermore, if it were just a matter of who’d approve of it, then (as I argued in that post awhile back on whether consequentialism and deontology were the only viable theories of ethics) at the fundamental level, there is no right-making property other than that observer’s approval. It’s simply up to the whims of that observer. But then it becomes morally arbitrary, it seems, that torturing puppies is wrong. (Or, on the other side of the Euthyphro dilemma, if the ideal virtuous observer actually has moral reasons for saying that torturing puppies is wrong, then it’s not arbitrary; but, the content of those reasons would be what makes torturing puppies wrong [as seems intuitively correct to me], not the fact of approval.)
Josh,
[Just a quick reponse for now – I’m off to watch the Red Sox and then the debate…]
Yes – what you say seems to capture what I’m proposing. On the other hand, I’d resist the ‘fundamental’ versus ‘intermediate’ terminology.
Perhaps compare a moral realist who embraces non-natural supervenient moral properties. And so the property of badness supervenes on the suffering of puppies. Would you object to such a view that “What matters is what happens to the puppies; not the presence of some non-natural property”?
I’d want to say that such a realist could claim that what is fundamental here is the suffering of the puppies; this is precisely why there is badness present (and why acts of puppy-harming would be wrong). Similarly, on my view, I think I can say that harming puppies is fundamentally wrong (in an obvious, important sense) precisely because of what happens to the puppies (their suffering, etc.). What happens to the puppies will be the object of the concerns of the virtuous ideal observers. [The proposal does sort of blur normative/metaethical lines.]
I won’t try to tackle the Euthyphro worry right now – though I agree that it needs to be addressed.
Campbell,
Well, the entailed because does not strike me, intuitively, as decreasing simplicity. In any case, the extra, entailed because does not seem to me to count in any way against the theory.
Jason, I’m not sure I understand your response to Ben’s objection. He asks:
If the virtuous person explains why it is right to save the drowning person by appealing ultimately to the badness of drowning, isn’t he just mistaken on your view?
To this you reply:
I’d want to say that actual humans could say such things as “I saved that person from drowning because drowning is bad”. I’d just cash the ‘badness’ out in terms the attitudes of virtuous ideal observers. So, to judge that drowning is bad is to judge that an omniscient, fully-informed virtuous observer would disapprove of it (in some particular way).
At first, I thought your response was to deny that the virtuous person is mistaken on your view. I understood you as claiming that, because “drowning is bad” is synonymous with “drowning would be disapproved of by a virtuous person”, the virtuous person speaks truly, according to your view, when she says “saving the drowning person is right because drowning is bad”. But it seems now that this is not your response. So where have I gone wrong here?
A quick note on Josh’s “intermediate” and “fundamental” right-making properties. Doesn’t the virtue explanation go as follows?
x causes is suffering > the virtuous disapprove of x > x is wrong,
where p -> q means “p explains q” or “q because of p”. This suggests that virtuous disapproval is the intermediary, and that suffering is fundamental. In any case, that’s the way I was thinking of it. And I was picturing the alternative explanation like this:
x causes is suffering > x is wrong.
Now, this certainly makes the latter explanation look simpler.
My thought is that, in itself, suffering is just a psychological (and perhaps physiological) state. We find suffering unpleasant, we try to avoid it, and many of us are inclined to help others avoid it, and so on. But these are all merely descriptive facts about us. I’d hold that we should care, morally, about suffering, because virtuous ideal observers would do so (or at least – they’d approve of us having such concerns). I don’t think a simple appeal to the presence of suffering itself would explain why puppy-torturing is morally wrong: “So what – the puppies are simply experiencing certain mental states that we often try to avoid. Why does that matter?”. The suffering matters morally because virtuous ideal observers would be moved by it. (Jason, Oct. 8, 11:59 am)
This argument begs the question: of course “in itself” suffering is just a psychological state that we try to avoid; but why think that the obtaining (or possible obtaining) of a psychological state cannot give rise to reasons for action? But here’s the more pressing worry: if the argument works at all, then why doesn’t it work in the mouth of the ideal virtuous observer? That is, when faced with an instance of suffering, why doesn’t the ideal virtuous observer say precisely what Jason has said: ‘Suffering is just a psychological state, so why should it matter?’ If the ideal virtuous observer takes it to matter, then there must be something wrong with the suggested argument (that descriptive facts can’t have normative implications); if he doesn’t take it to matter, then, by hypothesis, the rest of us don’t have reason to take it to matter (because it’s only a moral issue if an ideal virtuous observer takes it to be one.)
Not in any respect that counts as an advantage. Suppose someone thought that although atoms and the books they compose exist, individual pages of books do not really exist (quantification over them can be paraphrased away, blah blah blah). Would this odd view count as simpler than one that countenanced the pages as well? Maybe simpler in some uninteresting respect.
Campbell,
I do want to say what you suggest. We can truly say that “saving a drowning person is right because drowning is bad”, so long as we cash out the “right” and “bad” in terms of the approvals and disapprovals of virtuous ideal observers. [I was trying to emphasize that this would not involve appealing to facts about badness or rightness that are independent of the attitudes of virtuous ideal observers.] Does this get at your question?
Troy,
We (‘mere humans’) may naturally be moved by suffering, but we can still be unsure about whether we ought to be so moved [I perhaps put the concern too strongly in my original post. When I said, roughly, that we might think “pain is unpleasant for others, but so what?”, I had in mind the worry that perhaps with more information, we would think we ought to abandon this concern.]
The virtuous ideal observers will also be inclined to be concerned about suffering insofar as they are virtuous (and possess such traits as compassion and benevolence). The difference between the cases is that the virtuous ideal observers would also be omniscient. They would know all of the descriptive facts about suffering, its impact, and so on. The thought is that even in light of full-information, the virtuous ideal observers would still care about suffering. [Put another way – we should care about suffering because fully-informed virtuous observers would maintain and embrace this concern upon reflection.]
Jason,
Maybe you could say a bit more about rejecting the distinction between fundamental vs. derivative right-making properties. The reason I find it a significant distinction is that I read the dispute between rival (normative) theories of ethics as happening at the fundamental, rather than the derivative or case-judgment level (since, e.g., both utilitarians and deontologists can generate derivative rules to help the needy and both can say that you ought not torture the infant). So my inclination is to say that unless a theory makes the authority of the virtuous observer/agent the fundamental right-making property, that theory is not (purely, anyway) virtue ethical. It sounded, though, like you may want to go pluralist, a la Ross, and have multiple right-making properties, such that the approval of the virtuous observer is one fundamental right-making property and the prevention of suffering is is another.
Josh,
I’d be willing to accept the distinction you draw; it’s the terminology that I think is a bit misleading. So I do want to say that there is just the one “fundamental” right-making property in your sense (a particular form of approval of virtuous ideal observers), though, of course, there are many actions or features of actions of which these ideal observers could approve (so there is a pluralism in that sense – they might approve of acts of truth-telling, preventing suffering, and so on).
The reason I don’t like the terminology is that I consider the suffering of puppies (or “what happens to the puppies themselves”) to be “fundamental” in another, equally important sense. It is precisely what happens to the puppies (or drowning people, etc.) that evokes the approvals or disapprovals of the ideal observers.
Campbell has a comment above which I think captures the way in which I’m seeing the puppy-suffering as fundamental:
x causes is suffering > the virtuous disapprove of x > x is wrong,
where p -> q means “p explains q” or “q because of p”. This suggests that virtuous disapproval is the intermediary, and that suffering is fundamental.
Of course, there would still remain the differences in right-making properties between utilitarians, deontologists, and the virtue/ideal observer theorist that you mention. I’m not sure what to say here – I suspect we might end up just with clashing intuitions. Perhaps I’ll be able to come up with more later.
Jason, my worry is a little vague. But I think it relates to a remark you made earlier, as an aside:
The proposal does sort of blur normative/metaethical lines.
There seems not to be any great consensus on the precise boundaries of these two branches of ethical inquiry, but I tend to think of them in the following way. The task of explaining why certain actions are right, wrong, permissible, obligatory, etc. is the domain of normative ethics. While the task of explaining what it is for an action to be right, wrong, permissible, obligatory, etc. — i.e., the task of explaining what we mean when we say that an action has one of these properties — is the domain of metaethics.
Now, it seems to me that if your proposal succeeds at one of these tasks, then it must fail at the other. At times, it seems that you have in mind a metaethical view: e.g., when you talk of cashing out moral terms like “right” and “bad” in terms of the approvals or disapprovals of virtuous ideal observers. But then I find it hard to see how the view could explain why torturing puppies is wrong (to use your example). Consider an analogy. Suppose I ask why Hugh is a bachelor, and you answer “becuase he’s an unmarried male”. I don’t think you’ve answered my question. You’ve told me what it is to be a bachelor — what we mean by the term “bachelor”, what our concept of bachelor is — but I’m not any closer to undestanding why Hugh is a bachelor. I could know perfectly well that a bachelor is an unmarried male, and yet still intelligibly wonder why Hugh is bachelor — i.e., why Hugh is an unmarried male. In order to help me understand that, you need to give an answer like “becuase he enjoys the single life”. By analogy, it seems that I could know perfectly well what it means to say that torturing puppies is wrong — e.g., I might know, as you propose, that this means that troturing puppies would meet with virtuous disapproval — and yet still intelligibly wonder why torturing puppies is wrong. That is to say, I might wonder why a virtuous ideal observer would dissaprove of torturing puppies. In order to resolve the latter, we need an explanation like “torturing puppies is wrong because it causes suffering”.
If all of this is roughly correct, then the two explanantions on the table — i.e., the virtue explanation and the suffering explanation — appear to be compatible, since they aim to explain different things. The virtue explanation belongs to metaethics and purports to explain what we mean by saying that torturing puppies is wrong. Whereas the suffering explanation belongs to normative ethics and purports to explain why torturing puppies is wrong.
Jason,
You say:
We (‘mere humans’) may naturally be moved by suffering, but we can still be unsure about whether we ought to be so moved.
and
The thought is that even in light of full-information, the virtuous ideal observers would still care about suffering.
But how do I know that the virtuous ideal observers would still care about suffering in light of full information? I can only be confident about this if I am confident that I, in fact, ought to be moved by suffering. Of course, one could say that this is precisely the role that virtuous ideal observers play: we can’t be sure how we should be moved without them, so the way to find out how I should be moved is to find out how a virtuous ideal observer would be moved, and to do that we need to find a virtuous ideal observer. Only: how do I know that any particular person I find is a virtuous ideal observer? Perhaps I could test him by, for instance, seeing if he is moved by the suffering of a puppy . . . Ultimately, my worry is that I am at least as confident in any direct claim about how we ought to be moved as I am in any claim about how a virtuous ideal observer would be moved. So I’m not sure what talk of the virtuous ideal observer adds.
Sorry for the delayed responses – I’m still a bit hesitant on some of this, but…
Campbell,
Your characterization of the normative/metaethics divide does seem to capture the intuitive difference between the domains.
I think I’ll stick to my guns here, and hold to the proposal in both domains.
[One minor point – the proposal is somewhat revisionary, so I would not claim that I’m providing a strict analysis of ‘rightness’, etc.]
You note that if we ask why Hugh is a bachelor, we’d like an explanation along the lines of “He likes the single life”, or “He lives in the middle of nowhere and meets no one”, etc.
I think my proposal can give a parallel explanation in the case of wrongness (and thus could work as a normative ethical theory). Why is puppy-harming wrong? Because fully-informed virtous observers, given their psychologies, will be moved by the suffering of the puppies, and will be repulsed/upset/etc. by such suffering. They will thus come to disapprove of acts of puppy-harming (and notice how there is a sense here in which we can say that the suffering of the puppies is why acts of puppy-harming are wrong – this is what will provoke the disapproval of the virtuous ideal observers).
[And this disapproval will then constitute the wrongness of acts of puppy-harming; the metaethical claim.]
Troy,
I agree there is a bit of an epistemic gap here, but I don’t think it is quite as severe as you suggest.
First, I think we can know that fully-informed, virtuous ideal observers will disapprove of suffering. We ourselves (normal humans) have adequate knowledge, and are sufficiently good ourselves to make judgements in most ordinary cases. So I think we can make use of the direct judgements that you suggest; to the extent that we respond with approval or disapproval to a given case (upon reflection), we have strong, if defeasible evidence that virtuous ideal observers would have similar responses.
In more complex cases, where we lack crucial information, or where our vices distort our reactions to cases, then I agree we might not be able to judge as ideal observers would (or anticipate their reactions). But then this seems to me to reflect limitations to human moral knowledge, rather than a problem with the account of morality.
I think this is most clear on a broadly reliabilist account knowledge; what matters is that we can reliably anticipate the reactions of ideal observers; but notice that this does not require that we ourselves attain the standpoint of an ideal observer (anymore than accurately predicting the reactions and behaviours of a bat requires that we ourselves possess echolocation).
You might be worrying that we can’t be certain that we’ve captured the judgements of ideal observers. But then most epistemologists have rejected calls for certainty and instead embrace various forms of fallibilism.
Troy, your response makes me think that I’m missing something here, but I’m not sure what it is. So, let me have another go at stating my worry …
You say that want to “hold to the proposal in both domains”. But I’m worried that this is not possible. Consider the following two claims:
(M) To say that X is wrong is to say that X would be disapproved of by a virtuous ideal observer.
(N) If X is wrong, then its being wrong is explained (at least in part) by its being such that a virtuous ideal observer would disapprove of it.
As I understand things, to adopt your proposal as a position in metaethics would be to affirm M, whereas to adopt it as a position in normative ethics would be to affirm N. Hence, to adopt it both in metaethics and normative ethics would be to affirm both M and N. However — and here comes my contentious claim, I think — one cannot affirm both, because they’re inconsistent; if wrongness just is virtuous disapproval, then the latter cannot explain the former (just as Hugh’s being an unmarried male cannot explain his being a bachelor). To see this, assume that M is true. Then, to say that, e.g.,
is just to say that
And clearly that doesn’t explain anything.
Now, I feel like I’m probably skating over a number of subtle but important distinctions — synonymy vs. extensional equivalence, concept identity vs. property identity, yadda, yadda, yadda. In which case I hope someone will set me straight. But, on the face of it, this seems like a decent worry.
Campbell,
I wonder if this might help. I’m not sure I’d want to embrace (N) exactly, but rather
(N*)If X is wrong, then its being wrong is explained (at least in part) by its being such that it evokes or produces the disapproval of virtuous ideal observers.
[Like you, I’m worrying that there might be significant distinctions that I’m ignoring, etc. Still…]
Jason, can you tell me the difference between these views? I don’t get it.
(N) If X is wrong, then its being wrong is explained (at least in part) by its being such that a virtuous ideal observer would disapprove of it.
(N*)If X is wrong, then its being wrong is explained (at least in part) by its being such that it evokes or produces the disapproval of virtuous ideal observers.
What extra work is “evokes or produces” doing in N*?
First, thanks very much for inviting me to me a member of this blog. I like the idea of such a place very much. The pace of the discussion on this thing astounds me.
Second, thanks for thinking about stuff I (partially) wrote. That is fun for me.
Third, I have tried to respond to Jason’s concerns but each time what I write turns to nonsense before my eyes before I can press “post”, but I will keep at it. Sorry for the contentless post, but I did not want to delay any longer in offering the above thanks.
Ben (and Campbell),
Yeah – I see how the distinction might appear to be one without a difference… It might just be a matter of emphasis.
What I’m trying to get at, I think, is the following. In the metaethical story, the wrongness is constituted by the disapproval of virtuous ideal observers; there is no need to mention the properties of the action itself.
But in the normative story, to focus only on the disapproval of virtuous ideal observers misleading. To explain the wrongness of an action here (at the normative level) requires looking to the properties of the action (causing suffering, etc.) that evoke the disapprovals, given the psychology of the virtuous ideal observers.
So – (N*) at least shifts our attention to the act itself (and its properties), rather than just to the disapprovals of the virtuous ideal observers.
I need to say more here [e.g., is there really a substantive difference between N and N*?]. But I wanted to at least get something up for now.
PS – What of the following view? Take someone who holds that the property of rightness just is the property of maximizing happiness (a metaethical claim, along the lines that Moore would want to reject). Of course this person would also presumably be a utilitarian at the normative level. The position to me seems at least consistent, and not empty at the normative level. And if so, is there a relevant difference between this view, and the virtuous ideal observer proposal?
David,
I look forward to your comments (in a nervous, “Oh no, what’s he going to say??” fashion…)