I was just reading the bit of Gibbard's Thinking How to Live where he talks extensively about the good and the desirable—specifically, about its being a conceptual truth that they are one in the same. I know that this is a common line people take, and I'm curious how many of you agree with it, because I'm not entirely convinced.
Suppose classical utilitarianism were true. It might be that it would be good if my mother were dead, because this would save some large number of people from dying themselves. I'm really not sure it would be fitting for me to desire that my mother were dead.
Now, I'm not saying that it's entirely clear to me that it would not be fitting for me to desire this. I'm really not sure. But that lack of certainty—the apparent openness of the question—is enough to make me doubt that the identity is conceptual; so I'm curious why others take it to be.
One further thing: I wonder whether there is a connection between accepting this identity and the Humean theory of motivation. The relevant thought goes something like this: If X is good, then we ought to promote X. If we ought to promote X, then it is fitting for us to be motivated to promote X. If it is fitting for us to be motivated to promote X, then surely it is fitting for us to desire to promote X. If it is fitting for us to desire to promote X, then surely it is fitting for us to desire X itself. (I would be tempted to resist the last two moves, not to mention the Humean line itself.)
If we take it that the conceptual identity is best glossed as between ‘fitting to desire’ and ‘good.’ A lot depends on how we understand ‘fitting.’ If ‘fitting’ means something like a putative instance a of an attitude A about O is fitting if it is such that it can be about O while still being a genuine instance of A. So, it is fitting to believe that 2+2=5; though it is false, ‘2+2=5’ can be object of an attitude that counts as a belief. But, plausibly, it not fitting to hope that ‘2+2=5’ (or ‘2+2=4’)–its not the kind of thing to have hopes about.
But precisely because fittingness in this sense is a form of normativity which is about how the attitude instance relates to attitudes of that kind, it can be fitting to desire all manner of things which are not good. Desiring the death of your mother is fitting on this understanding–it is not a moral or aesthetic endorsement of the attitude to say so. It can be fitting to desire something–it can be the sort of thing about which one can have a desire, but then the identity does not seem to be a conceptual truth or an identity at all. (Which is not to say that there can’t a be a good explanation of evaluative concepts or properties in terms of pro-attitudes.)
But we might reject the interpretation of desirable as fittingness, and understand it as meaning something like ‘endorseable by moral standards.’ Then we get a vicious circularity, I think. But we might take it to mean ‘endorseable affectively in the light of my own self-conception.’ Then it would be much more plausible to say that you can affectively disapprove of your mother’s death from a filial perspective while affectively endorsing it from a broader perspective that you can also occupy (say, a philanthropic or impartial perspective). So perhaps a conceptual identity of the good with desirable in the sense of affectively endorseable from a philanthropic or impartial point of view would make sense.
Hello David. Nice question.
Wouldn’t the example be stronger if classical utilitarianism were false? If classical utilitarianism were true, why would it not be apt to desire your mother’s death given both that this was best? Given that, on classical utilitarianism, you ought to kill your mother to save five, surely you also ought to desire that your mother is killed to save five. If you ought to desire this, why is it inapt to desire it?
Wouldn’t a better argument showing how it could be true that something is good but not desirable draw on the non-consequentialist idea that it is sometimes wrong to do what is impersonally good or best, and the relationship between what we ought not to do and what it is inapt to desire? Here is a quickly drafted sketch:
1) It would be impersonally good that V is killed saving five.
2) It would be wrong for me to kill V to save five.
3) If it is wrong to kill V to save five, it is also wrong (or inapt) for me to desire that V be killed saving five.
4) What it is wrong (or inapt) to desire is not desirable.
5) Therefore, V being killed saving five is impersonally good but not desirable.
6) Therefore, what is impersonally good is not what is desirable.
Obviously, there is a great deal in this argument that is controversial. 4), for example, needs to be qualified to make it plausible. But it might be suggested, drawing on your own suggestion, that the possibility of a valid argument like this demonstrates that the good and the desirable come apart conceptually.
You seem to be assuming that if X is desirable, then it would be fitting for you to prefer X to ~X. But why not instead think (and I’m following Jussi Suikkanen 2009 here) that X is desirable if and only if it would
be fitting for an impartial spectator to prefer X to ~X. Thus, the reason it seems like it may not be fitting for you to prefer X to ~X is that you’re not impartial: that is, you have a special relationship with the person who is dead in ~X, and this, perhaps, gives you a fittingness reason to prefer ~X to X.
The only conceptual truth plausibly in the neighborhood is that the *intrinsically* good is the desirable.
I wonder whether “desire” should be read as “preference” here, or as something else..
Consider this argument against Gibbard’s view, in light of Eric’s and Doug’s points.
An impartial spectator can bring about world 1 or 2:
(1) You are somewhat virtuous and very beautiful
(2) You are virtuous but dreadfully ugly
Assume we can control for all other differences.
Assume the spectator is in a pickle because the worlds contain an “equal amount” of intrinsic good. Perhaps he flips a coin to choose which one to bring about.
This seems to be a case in which it is not fitting for the impartial spectator to prefer the existence of something (beauty) that is in fact intrinsically good.
One move here is to shift from fitting to prefer to fitting to take pleasure in the contemplation of, so perhaps that is a better way to defend the Gibbard position? Does Gibbard tell us what he means by ‘desirable’?
Here are some different things we might mean when we say that an attitude A is fitting for object O (this is not necessarily exhaustive):
When I posted, I was thinking in terms of (2) and (3). Of course, as Eric B. suggests, we could consider something closer to (1). Surely, though, if it turns out that there are things that might be good but that can’t be objects of desire at all, then we should worry about the identity of ‘good’ and ‘desirable’. But let’s set that aside.
My assumption was that when most people say that the good is the desirable, they mean either that good things are such that it is appropriate (i.e., not a mistake, as such) to desire them, or (stronger) that good things are such that one always has a reason to desire them. I have typically been inclined towards the former ((2)) reading of fittingness in general, but I know that many people think that fittingness involves reasons (cf., various Jacobson & D’Arms pieces). I didn’t address this distinction in my original post because I’m worried about the weaker version, and I assumed if my worries were legitimate, they would be at least as legitimate with respect to the stronger version.
Now, as Douglas points out, we might instead understand things as in (4-6), thinking in terms of an impartial spectator. (I haven’t read the relevant piece by Jussi, but I suspect I know the rough line—Christian Coons has something about this, too, from RoME this past year—so I assume that we’re talking about either (5) or (6), and thus I’ll stick with the weaker (5).) On this line, it would be appropriate for the impartial spectator, though perhaps not for me, to desire my mother’s death. (Victor, I think making this move would take care of your case, since we’re now talking about someone who would, by definition, be concerned with the impersonal good.)
This brings us to the issue of what we mean by ‘desire’. If all we mean is something like a minimal preference for or motivation towards the object, then the identity sounds at least initially plausible. It certainly seems that it would be appropriate for an impartial spectator to be motivated to pursue my mother’s death in the case described. (Actually, I would have thought this identity was uncontroversial, perhaps with Eric W.’s emendation, but Brad’s argument seems to pose a threat even to this reading. There might be a way to deal with this, though, perhaps by making ‘the good’ there disjunctive or something.)
Sometimes, though, when we talk about desire we are talking about a particular attitude, perhaps something like the ‘taking pleasure in the contemplation of’ Brad suggests. If we understand ‘fitting of desire’ in this way, I think it may well be a mistake to identity the good with the desirable. It is not at all clear to me that it would be appropriate even for the impartial spectator to feel positive about my mother’s death.
At this point, we might just note the ambiguity and move on. Where ‘desirable’ means ‘appropriate to promote’ or something like that, it may well be a conceptual truth that the good is the desirable (again assuming we can deal with Brad’s worry). Where ‘desirable’ means fitting of actual desire, the same is not true.
I’m not entirely happy with this, though. ‘Desirable’ is one among a number of fitting-attitude terms like ‘fearsome’, ‘amusing’, etc. My understanding is that such terms are generally taken to regard when it is appropriate to have an attitude that we are all familiar with—fear, amusement, etc. It thus makes sense that desirability would concern when it is appropriate to actually feel desire (whether for us or for the impartial spectator).
If that’s the natural way of thinking about desirability, then why the ambiguity, why the alternative sense on which it is identical to goodness? My inclination is to think that we should regard this not as an equally good alternative meaning, but as a mistake produced by the kind of Humean thinking I mentioned above. It is natural to think that if it is appropriate to be motivated towards something, and that motivation requires desire, then it would be appropriate to desire it. Natural but, I think, mistaken.
I take it that your example draws on agent-relative duties that you have towards your mother. Let us suppose that X desires P implies that X is motivated to P (or, better, would be motivated to P were it possible for X to bring P about). Just as it would be wrong for you to kill your mother to save five, it would be inapt for you to be motivated to do so. Due to special relationships we ought not to do, or to be motivated to do, what is best.
But, following Doug’s line, if ‘X is good’ implies that an impartial spectator may appropriately be motivated to pursue X, it is good that your mother is killed. Hence, your example does not count against the relationship between what is good and what is desirable.
However, if we replace duties not to kill grounded in special relationships (such as the relationship between you and your mother) with general prohibitions on doing what is best (such as the wrongness of killing as a means to save five), wouldn’t the case be strengthened against this challenge? For the duty not to kill as a means to save five (if it is a valid duty) applies to everyone.
Non-consequentialists typically understand these duties as duties not to bring about the thing that is impersonally best. But as they are duties that apply to everyone, they are duties that apply to impartial spectators. Impartial spectators ought not to be motivated to kill one as a means to save five. Hence, we could have:
It is impersonally best that one person is killed as a means to save five. But no one (including an impartial spectator) should desire what is impersonally best, as no one should be motivated to kill one as a means to save five. Hence, what is impersonally best is not desirable.
Victor,
That sounds right to me. If it can be wrong to promote the good, then it is hard to see why it would always be appropriate to desire the good.
That being said, I didn’t mean to be relying on anything controversial about non-consequentialist thinking or agent-relative duties in voicing my concerns. My point was that it might not be appropriate for the impartial spectator to desire my mother’s death, even if killing her to save the five is the right thing to do.
Victor
there’s a reason impartial spectators are called ‘spectators’. They are not agents – they do not act. For that reason, they are not under duties, nor ought they be motivated not to kill. So, there are no worries of the kind you suggest here.
Brad,
same kind of applies to your case. It’s not really what the spectator prefers to bring about as he cannot bring about anything as a non-agent. Rather, the question is what kind of outcomes is it fitting for her to prefer generally. If it is fitting for her to prefer outcomes that include beauty other things being equal over many other outcomes that do not include beauty, then on this view beauty is desirable. It is one of the things that make it fitting for the impartial spectator to prefer outcomes over many other scenarios.
Jussi,
I can see the line you are taking here but I worry about your use of the always slippery ‘generally’ and ‘other things being equal’.
You end by glossing generally as “in many scenarios” but this seems too weak. Their are surely many scenarios in which the absence of beauty is preferable to its presence, and many in which the opposite is the case. This threatens to imply that beauty and its absence are both intrinsically valuable.
Why not just move to the view I suggested: x is intrinsically valuable in situation S iff it is fitting for ImpSpect to take pleasure in the contemplation of x in S?
This avoids the need to give a satisfying account of “in general”..and it allows that there may be some things that are intrinsically valuable in one context but not in another (it seems your view rules that out).
David
I find that hard to believe.
if killing your mother is the right thing for a person to do, how could it be inappropriate to desire to do it? If it is right to do it, it is not in appropriate to do it. If it is not inappropriate to do it, it is not inappropriate to desire to do it. And if it’s not inappropriate to desire to do it, it is difficult to believe that it is inappropriate to desire that it happens? It is, of course, inappropriate to desire it for its own sake. But it might be appropriate to desire it for its instrumental benefits.
Perhaps you might argue that it’s sometimes right to do something that it’s inappropriate to do, which is of course true. But this doesn’t seem a case of that.
Jussi,
I was relying on the idea that X desires P means X would be motivated to bring P about if he could. That implies that all desiring beings are agents.
if impartial spectators are not agents, do they have desires or preferences?
If so, I guess X desires P can’t mean X is motivated to bring P about. Similarly, X prefers P cannot mean that X would have a tendency to choose P. If this is not what ‘desire’ means, what does it mean to say that impartial spectators have desires or preferences?
Perhaps something like Brad’s suggestion – X desires P if X takes (or would take) pleasure in the contemplation of P? Let us suppose that this is what it means for X to desire P. I am unsure whether if D may not kill one as a means to save five, it is apt for D to take pleasure in the contemplation of the one being killed as a means to five being saved. Perhaps this is not wrong. If so, the thesis that what is good is what is desirable can meet the objection.
I am not sure, though, that this is what we typically mean by desire given that I can desire never to take pleasure in the contemplation of anything ever again without contradiction (say because I think I deserve this). I am attracted to the idea that only agents desire, and hence that thinks are desirable only to agents.
If impartial spectators cannot desire or prefer, I guess it makes little sense to ask what they would, or ought to, prefer or find desirable.
Hi David,
I share your doubt about it being “appropriate even for the impartial spectator to feel positive about [your] mother’s death.” But I do not see this as a mark against the contemplative pleasure view. As Eric suggests, the target here must be intrinsic value and your mother’s death is not intrinsically valuable. And that fits the plausible intuition that the impartial spectator would not be pleased by her death.
More generally:
I think that you might be right that desirable/preference accounts of the good are sometimes motivated by Humean thinking, and I think you are onto a strong worry about that views that are so motivated.
But I suspect that in other cases desirable/contemplative pleasure accounts of the good are motivated by the idea that the good is what would please God. Just some armchair speculation of course!
Hi Brad,
I agree – I did oversimplify. However, there’s so many different things going on here that I don’t even know where to begin with.
Firstly, I think I would want to hold onto the idea that intrinsic value is an intrinsic property of the object. Intrinsic (aka non-relational) properties then are supposed to be properties that the object keeps under all circumstances. And, so, this would go for value too. This motivates to look for whether the object is fitting to prefer generally, by and large, in most cases and so on. I know that this is vague. But, I do think that this is how good works overall. Following Broome, I do think that concrete better than comparisons are basic and somehow on the basis of these we construct a vague notion of value – better than most.
Now, you might think that intrinsic value need not be an intrinsic property and so the same object can in some contexts have intrinsic value and in others lack it. I do have one worry about this, which is to say that it is not clear whether on this view we can say anything about what has intrinsic value full stop. It seems that all we have on your view for example is intrisic value in situation S. But, then I would still want to know whether the object is good.
One thing we could use is default reasons or default fittingness. You might think (and in fact I do think this) that by default beauty is fitting to prefer. If someone prefers beauty in a context, we do not seek a special explanation for this – it seems like beauty is turned on to make preferring fitting on its own. Yet, in some contexts, you are right that it is not fitting to prefer beauty but rather its absence. I guess I would want to say that in these cases we need a special explanation for why this is the case. There must be some consideration that disables beauty’s ability to make preferring fitting (as Dancy would put it).
The resulting view would hold that an object is intrinsically valuable if it is *by default* fitting for the impartial spectator to prefer it. This would also allow me to say that in some contexts even this intrinsically valuable object might not have intrinsic value (or maybe final value is better here) as in these cases there is a disabler present that disables the object’s ability to make preferring fitting. I think this might be a nice way to cash out the ceteris paribus clauses and combine the view with something close to your suggestion.
With your suggestion, I don’t think pleasure will be able to do much work. Two ways to cash this out. You might think that pleasure is a phenomenally basic experience so that in all cases there is same positive feel to the relevant contemplation. I doubt such experience exists. The other way to think about pleasure is to think of it in terms of the experiences we want to have. Yet, this view seems to collapse close to the view we started with.
Also, I would want to think that spontenously enjoyable things can be intrinsically valuable. Sex might be intrinsically good. Yet, it doesn’t seem fitting to take pleasure in contemplation of sex as it might be fitting to prefer to have sex. The former – contemplating sex – just seems creepy. Maybe this is the wrong kind of reasons problem.
Jussi,
I just tried to read your relevant piece but the links through my school were broken, so I’ll have to wing this. (It sounds like your view is quite different from Christian’s; his idealization is an agent, not just a spectator.)
Anyway, your view seems to imply that it is possible that something is desirable, yet is the sort of thing it would be inappropriate for any and all agents to desire. First, that just sounds odd to me. Second, it doesn’t seem to fit with how we think about fitting-attitudes terms in general (amusing, etc.).
(Just to be clear, I do think that something could be amusing, yet everyone could have reason not to be amused by it. Just add a dash of Evil Demon. But this is not like that. The point here is that it seems like something might be desirable in your sense yet be the sort of thing that, due to its very nature, it is inappropriate to desire.)
Brad,
The focus on intrinsic value might fix the worry; I’m not totally sure. For instance, I’m not confident that the I.S. should take pleasure in the death of a villain, even if the death of a villain is intrinsically good. (I suppose a lot of this will come down to what we mean by an intrinsic good.)
You might be right about the God thing, though (here’s my own bit of armchair speculation) I do wonder whether this itself is an outgrowth of a Humean line of thought (though, obviously, not historically starting with Hume). After all, I can wonder the same things about God as I did the I.S. It is not entirely obvious to me that God takes pleasure in everything that’s good. Casting Satan into Hell might have been good, even intrinsically so, but God might well have mourned the necessity.
David,
the paper is open access at JESP so you should be able to access it from anywhere (http://www.jesp.org/articles/download/Consequentialism.pdf ).
It doesn’t follow from my view that a valuable thing because of its nature would inappropriate to desire for normal agents. Far from it. It is fitting for the impartial spectator to desire the very same commonplace things as it for the rest of us. It will be fitting for her to desire there to be knowledge, pleasure, friendship, achievements, biodiversity, and so on and so on just like for any of us.
However, when we consider what is fitting to prefer for us as situated agents, there are certain things that make a difference that should not matter with respect to intrisic value. These are our relations and our agential involvement. The impartial spectator is so defined that she has none of these. It is for that reason that we can account for agent-neutral intrinsic value by thinking about what it is fitting to prefer for this person who is abstracted away from personal relations and agency. So, this is very much in the tradition of fitting-attitudes accounts. I don’t worry about oddness that much.
Brad,
sorry – I tried to respond to you earlier but for some reason the comment did not get through.
I grant that I was being sloppy, but there are many moving parts that we need to be careful about.
Firstly, I am thinking about intrinsic value as an intrinsic property of the objects. For this reason, if an object has it, then it has to have it in all circumstances. This leads to the ceteris paribus clauses. I accept that ‘in many cases’ is not the right way to formulate the view.
Here’s a better one inspired by Dancy. It seems plausible to think that it is fitting to prefer beauty by default. Beauty is such that it is automatically turned on to make fitting it appropriate. We expect this if we have no further information – we look for no further explanation. In contrast, when it is fitting to prefer the absence of beauty over beauty, we assume that there is some special explanation for this. There is some relevant consideration that is a disabler, that disables beauty’s ability to count in favour of preferring. The suggestion then is that it is by default fitting for the impartial spectator to prefer things that have intrinsic value.
So, on this view, we get a theory of intrisic value full stop, and not just a view of intrisic value in a situation like on your view. We can also generate a theory of intrisic value (or better final value) by thinking about what it is appropriate for the agent to desire in the given situation taking the disablers and enablers into account. I think dealing with the cases like you suggested first we need to think in a more fine-grained way about what the objects of preferences are in that situation.
Let me just also offer few worries about your proposal. Firstly, we need a story of pleasure in it. You might go for a phenomenal account on which all pleasures have the same positive feel to them. I am sceptical that this is true. Instead, you might go externalist and think that pleasures are experiences we want to have. This collapses your view close to the original suggestion.
Finally, I don’t like the contemplation part. Some good things are better enjoyed spontanously. Being intimate with someone is a good example. This might be intrisically good. Preferring it too sounds fitting but contemplating being intimate with someone and experiencing pleasure as a result seems creepy to me.
Thanks, Jussi. I will have to download your article and read it when I have a chance. Thanks for taking the time to type up the thoughtful response!
I hope I’m not missing anything here; I’m having trouble keeping the order straight, as some responses came in while I was working on my last comment…
Victor,
I don’t see why this should be the case. A desire is not just a motivation; it involves other things, like perhaps the pleasure that Brad alludes to (though I find Jussi’s worries about that particular way of cashing it out compelling). It seems to me perfectly possible that there are things I should be motivated to do yet that I should not desire. Of course, you can use desire in a more limited way if you wish, just to mean a motivation. But I think that’s misleading. For one thing, it makes the idea that motivation requires desires vacuous.
Jussi,
Thanks for the link to the article. I was looking for the wrong one!
My point was only that your view seems to allow for the possibility of some X, such that X is desirable but it would be inappropriate for normal (or even idealized!) agents to desire X. The death of a villain might be intrinsically good, yet I think it might be inappropriate for any agent to desire it. Yet, if we understand the desirable as what the I.S. would prefer, the death of a villain would indeed be desirable.
Just to be clear: I am not objecting to the idea that we can understand intrinsic goodness in terms of what it would be fitting for the I.S. to prefer. I’m just objecting to the idea that we should use this to characterize what it is for something to be desirable, for the reasons already indicated.
David,
Suppose either two children or one child is going to die, and all else is equal; otherwise we all kick the bucket. It’s desirable that on child dies instead of two. Now suppose you learn that the one child is yours, and the other two children are strangers. It is appropriate to desire that the two children die instead of your child.
This isn’t exactly the case you’re denying is possible, but it’s pretty close. Maybe there are worries about the ceteris paribus clause and whether we can exempt your relation to the children in it. Maybe. But this type of case probably happens often.
The answer involves distinguishing intrinsic value from value for an agent, I think. There is a sense in which it is appropriate to *both* desire two children to die rather than one, and one child to die rather than two. These are just different senses of ‘appropriate’, or involve different types of constraints on appropriateness.
Otherwise one can build into the appropriate response the distance of the desiring agent from the state of affairs they are desiring. The closer the agent is to his desire’s object, the more appropriate is his desire (or the greater degree of desire is more appropriate, if you think these come apart), all else equal.
See Graham Oddie’s: “Value, Reality, and Desire” 2005 for a defense this type of approach.
David
thanks for the response. Let’s suppose that it is permissible for me to kill my mother to save the five.
I desire to save five and killing my mother is the only way to do so. It is appropriate for me to desire the saving of the five, but inappropriate for me to desire the killing of my mother (at most, I am willing to do so, and I agree that desire involves more than just motivation – I don’t desire to go to the dentist, but I am willing to do so to get good teeth).
But given that it is not wrong for me to kill my mother to save five, it is appropriate for me to desire to save five, despite the fact that killing my mother is necessary to achieve the goal.
If my mother is killed by someone else, and this was necessary to save five, this is good overall. But it is also desirable overall – desirable that the five were saved despite the killing of my mother. It is desirable overall compared with the alternative – that my mother was not killed, but the five were not saved.
So whilst you are right to say that it is not desirable that my mother is killed, the saving of the five is desirable despite the killing of my mother that inevitably goes with it. It is more desirable (and also better) than the alternative under consideration – that the five are not saved, but my mother is not killed. And isn’t that all the person who defends the view that goodness = desirability needs? After all, it is good that the five were saved despite the killing of my mother. All that needs to be shown is that it is also desirable that the five were saved despite the killing of my mother.
)))((((((
(·)…(·)
….U….
[____]
Christian,
I think this is orthogonal to the issue I’m raising. My point is that it might not be appropriate for me to desire any child’s death at all, regardless of what child’s death it would be appropriate for me to pursue (if any), given the circumstances.
Victor,
Consider:
(1)-(3) concern the same state of affairs. It is a good state of affairs (we are supposing). Because goodness is extensional, (1)-(3) are all good. But because desire is intensional, it seems possible that it is appropriate for me to desire (1) and (2), but not (3). In other words, I take it that if a state of affairs is good, it is good under all descriptions. But it may be appropriate to desire it only under certain descriptions (or perhaps even under none at all!). I think this is a (though not the only, or even necessarily the primary) reason to worry about identifying goodness with desirability.
Thanks for the clarification David. I’m a complete novice, but is goodness extensional despite the fact that neither ‘intrinsically good’ nor ‘instrumentally good’ seem extensional? Consider:
1) It is intrinsically good that in circumstances C, five lives are saved
Is true but
2) It is intrinsically good that in circumstances C, my mother dies
Is false
1) It is instrumentally good that in circumstances C, my mother dies
Is true but
2) It is instrumentally good that in circumstances C, five lives are saved
Is false.
Also, suppose that saving five kills my mother as a side-effect. The killing of my mother is not the means to save the five. If so, my mother being killed is neither instrumentally good nor instrinsically good. In this case:
‘In circumstances C, it is good that my mother dies’ seems false.
Compare the standard trolley case where the trolley is turned from five to one. In these circumstances:
1) It is good that the five are saved.
2) It is good that the trolley is turned.
3) It is not good that the one is killed.
Victor,
It’s the blind leading the blind here; this isn’t my usual area, either. But I’ll give it a go.
I take it that when I say that a state of affairs, under some description, is intrinsically good, what I mean is just that the state of affairs is good and that the description appeals to that state’s intrinsically good-making features (or some of them, anyway). Thus,
really means
Similarly, when we say that
this really means
Since intrinsic/instrumental seems to modify the feature being described, not the state of affairs itself, I don’t think this threatens the idea that goodness is extensional. Whether the state of affairs is good doesn’t depend on whether the features described are intrinsically or instrumentally good (or on anything else about how the state is described, for that matter).
(You might just see this as a more evidence that we should be equating desirability with intrinsic goodness, but as I’ve already said, I’m not convinced even of that, though for reasons that have nothing to do with this intensionality stuff.)
As to the side-effect stuff: I think I would just deny (3). Certainly, the state of affairs ‘one is killed’ is not good, all else being equal. But all else is not equal; once I recognize (3) as a redescription of what happened in a trolley problem, I realize that the one being killed is good. (Or I would, if I shared that intuition about the trolley problem.) This is precisely why I included ‘In circumstances C’ in my description of the mother case, because without it, it sounds like I’m saying, sans context, that the state of affairs ‘my mother dies’ is good!
David, sorry, scratch the previous misguided comment. I see.
1) It is intrinsically good that [in circumstances C five lives are saved]
and
2) It is intrinsically good that [in circumstances C my mother dies]
are both true – it is the complete state of affairs that is good, not the complete state of affairs under some description.
So the question is whether
3) It is desirable that [in circumstances C my mother dies]
is false, even though
4) It is desirable that [in circumstances C five lives are saved]
is true.
I’m still not sure why ‘good’ and ‘desirable’ should come apart in this way, though. It is apt to desire the complete state of affairs under the most complete description given that the good outweighs the bad, we might think. If so, why is it not apt to desire the state of affairs under any other description? Isn’t it the complete state of affairs that we aptly desire, not the state of affairs under a particular description?
Perhaps the reason why it seems jarring to say that the killing of your mother is desirable is the same reason why it seems jarring to say that the killing of your mother is good – it implies that the your mother’s death is the property in virtue of which the complete state of affairs is desirable.
Suppose that ‘taking pleasure in’ is an aspect of desire. It is not apt to take pleasure in your mother’s death, but it is apt to take pleasure in the complete state of affairs where your mother dies but five are saved. If so, it is apt to take pleasure in this state of affairs under any description – it is the state of affairs that we take pleasure in, not the description of it.
As I say, I’m sorry if my comments are all over the place – this is far outside my comfort zone.
Victor,
No worries! We’re outside mine, too, but it’s fun to think through this.
I think the sticking point is here:
I’m just not convinced this is true. I think it might well be inappropriate to take pleasure in any state of affairs in which my mother dies. It’s like your example of going to the dentist—makes sense to pursue, but perhaps not to desire.
David
Now that I understand your position (at last) I think it’s fascinating. Does this capture it?
1) Take some state of affairs which is good because its good properties are sufficient to outweigh its bad properties. 2) It may nevertheless be inappropriate to be in certain affective states with respect to this state of affairs in virtue of its bad properties, such as to take pleasure in it. 3) Desiring a state of affairs involves being in an affective state of that kind. 4) A state of affairs is desirable only if it is appropriate to be in such an affective state. 5) Hence, a state of affairs can be good without being desirable. 6) Hence what is good is not what is desirable.
There are various ways in which this might be true.
A) Some people might have agent-relative reasons not to take pleasure in the relevant state of affairs (which might explain your case involving your mother). If so, those identifying good with desirable can claim that good is desirable for the impartial. But you think that the idea might be valid for impartial spectators too.
B) It is inappropriate to take pleasure in the relevant state of affairs as a whole, but appropriate to take pleasure in its good properties considered independently (I can take pleasure in the saving of the five, considered alone, but not take pleasure in the whole state of affairs where the five are saved and one is killed)
C) It is inappropriate to take pleasure in the good properties of the state of affairs in virtue of the fact that they are properties of a state of affairs, somehow appropriately circumscribed, with the relevant bad properties (I ought not to take pleasure in the saving of the five, considered alone, given the relationship between the saving to the killing).
I wonder which of B and C (or both) you had in mind, or which you think is more attractive?
Here is another thought. All of this still leaves open the following, which might satisfy those who think that there is identity between goodness and desirability:
A good property of a state of affairs is a desirable property where a desirable property is a property that it is always apt to desire (if C is never true) or where a desirable property is a property which it is conditionally apt to desire (if C) is sometimes true.
I haven’t read everything in this thread carefully, but (a) it seems to me that Brad hit an important point in suggesting that preference rather than desire is the right attitude to focus on, and (b) in that case we can consider the analogy (or identity) with better than rather than good. (And I’m with John Broome in thinking this is a good idea for independent reasons.)
So, instead of ‘desirable’ try ‘preferable’, and then see if ‘x is preferable to y’, in the fitting attitudes sense, matches up nicely with ‘x is better than y’. I boldly conjecture that this will work… better.
Victor,
Yes! That’s very much what I was thinking. Sorry if my own expressions of it lacked clarity.
As to (B) vs. (C): I think that I’m more inclined towards (B), though there’s a sense in which I want to accommodate both. This is just the stuff about intensionality. I think that a desire for ‘saving five lives’ is different from a desire for ‘saving five lives in circumstances C’, even if both desires are somehow “in reference to” C. If C is such that saving the five requires my mother’s death (and I know this) then it might be appropriate for me to have the former desire, but not the latter. Or so I’m tempted to think, anyway.
As to your final suggestion: That might well be right. But it sounds a bit like just saying that it’s appropriate to desire good things, unless they have undesirable features. This will only be interesting if it’s not true of everything. But I think it might be; I’m inclined to think that desirability is fairly permissive. It’s okay for me to desire anything I like, so long as it doesn’t have undesirable features. (It’s an interesting question whether this is true for other attitudes. Are things amusing “by default” and only cease to be so if they have unamusing features?)
Jamie,
I think that whether that works or not will depend on what we understand a preference to be. If preferring something is, say, just being disposed to pursue it over other options, then that sounds fine to me. But if preference involves any sort of positive affective state, then I think this will have the same sorts of problems I was trying to raise for desirability.
David
Great. I feel the pull of the judgement that an impartial spectator could typically take pleasure in [five being saved], but ought, at least sometimes, not to take pleasure in [one being killed and five being saved] (Case 1).
What explains the judgement?
Here is one possibility: Just as there is something wrong with directing our pleasure towards evil, there is something wrong with directing our pleasure towards a state of affairs with evil in it. This explanation seems tempting: your mother’s death is a tragedy, and it is wrong to take pleasure in a state of affairs with a tragedy in it. I think you might be tempted by it.
I doubt that this is the right explanation. Compare taking pleasure in [one person, Betty, suffering a tragedy at t1 and five great joys at t2-6]. I don’t feel the pull in this case. It is not inapt to take pleasure in a person’s life as a whole where that life has tragedy in it. In your earlier comment, you thought suggested it inappropriate to take pleasure in any state of affairs with Betty’s death in it. Consider the whole of Betty’s life, with great joys in it, but including Betty’s death. I think it not inappropriate to take pleasure in her life as a whole.
Another possibility, which I think the previous example suggests, is that the judgement in Case 1 has something to do with the separateness of persons. We consider the state of affairs from the point of view of each person independently, and we prioritize the worst off our affective reactions.
Victor,
Ugh, I just wrote a long response and the site deleted it. Hopefully I can recreate…
Nice point. I agree with you that what you suggested can’t be the right explanation (though I admit I found it appealing). I’m just spit-balling here, but suppose I think:
Let me explain. In the Betty case you offer, my intuitions shift depending on whether the tragedy in question is a necessary means to the joy. Suppose that it is. In that case, the state of affairs ‘Betty suffers at t1 so as to experience great joy at t2-6’ seems desirable. After all, I take it that this state of affairs was a net gain in well-being for her.
Suppose, though, that the suffering was unnecessary. In that case, it seems to me that the state of affairs in question is actually constituted by two independent states of affairs: ‘Betty suffers at t1’ and ‘Betty experiences joy at t2-6’. And I think that the undesirability of the former makes the whole state of affairs undesirable. After all, how can it be appropriate for me to desire that Betty suffer and then experience joy if I know it is possible for her to experience the joy without the suffering? Similar points go for her life as a whole.
One further thing. You write: “It is not inapt to take pleasure in a person’s life as a whole where that life has tragedy in it.” I agree, but I wonder if this is because we can take pleasure retrospectively as well. Certainly, I think I can take pleasure in the life Betty had, if I know that it was a life full of joy, with a few setbacks. After all, it was a pretty great life! But (similar to before) I think it would be inappropriate for me to desire that she have such a life if I believed it possible for her life to be at least as good without those setbacks.
David
Good. I’m not sure whether the fact that S has bad properties that it could have lacked always renders S undesirable. ‘S is desirable’ might imply ‘S has many good properties and few bad properties and hence is to be desired over many available alternatives’. This could be true even if there is an available alternative to S that has all of S’s good properties but lacks some bad properties.
Let’s suppose that you are right.
You are still inclined to contrast the interpersonal and the interpersonal though.
Given that we can desire a state of affairs where one person is worse off in some respects if they are better of in other more significant respects, what explains the inappropriateness of desiring a state of affairs where one is worse off but others are better off?
Victor,
That might be right, though I might want to distinguish between desiring that ‘S obtains rather than T’ and desiring that ‘S obtains’. And I might think that the former is desirable, but not the latter. I also suspect this is getting into subtleties our language isn’t used to dealing with.
You’re right that I’m inclined to contrast the inter- and intrapersonal. I suppose I’m disinclined to desire tragedies and I think that a situation in which someone ends up worse off is a tragedy, at least in one sense, even if it was, in another sense, good because of how it helped others. On the other hand, I’m less inclined to think that it’s a tragedy at all if I undergo some negative experience for the sake of a (more) positive one (though I suppose that there’s some cosmic sense in which I might bemoan the need to take the bad with the good).