The standard view of Kantianism has been that it firmly denies that the right is grounded in the good.
But recently it has become almost de riguer for Kantians to reject this priority of the right and embrace the good as the ground of obligation. For some (such as Guyer and Wood), the issue concerns mainly what Kant really thought. They have tried to find texts in which Kant asserts that the value of something, say humanity or the good will, somehow explains why it is necessary to avoid certain behavior. Others (such as Nagel and Herman), base their apostasy on views about practical reason, worrying that those who embrace Kantian principles have nowhere to go to explain why we should do what we ought unless there is some good that is realized or comes of it. But all in all, recent work among Kantians on this issue seems to me to amount to a rebellion against the standard picture. It is the latter that I’m mainly concerned about.
The catalyst igniting this rebellion has been Korsgaard. In early work, she embraced a view in which Kantianism required a special value, the good will or humanity, as a source of reasons to act. But she has since seemed to reject that earlier view. Now, especially in the Sources of Normativity, she embraces a less robustly ‘realist’ view of value. Practial reason ‘confers’ value on everything, including humanity and the good will.
I myself am, perhaps misguidedly, a reactionary on this. I think the standard Kantian picture, of some sort that Korsgaard likes of late, is so obviously right that I have difficulty seeing how this recent rebellion can be sustained. Against it is the idea that a necessary requirement — something you must do — is wholly inconsistent with the idea that the reason you must do it is that you will realize or produce some good. You can’t get a necessary requirement out of a value based theory. So if you think there are ethical absolutes — as do Kantians (I’m assuming) — then you can’t accept a value based theory of obligation.
That’s my knuckle-dragging Kantian view. But I’m not all that confident in this seemingly (to me) obvious view. What do you think? Should Kantians become teleologists?
Robert,
I think it’s worth remembering that at least part of what sparked the rebellion was Kantians’ attempt to answer the ’empty formalism’ objection. If the right is not grounded in the good, then (critics said) the mere idea of practical consistency is not enough to generate determinate obligations. Hence, Wood, Timmons, Korsgaard, etc. sought to read the CI in terms of value or good, i.e., rational autonomy or humanity. So the “right is grounded in the good,” as you say.
But there’s still room for saying that this grounding does not reduce the right to the good or render Kantianism unsettlingly teleological. The trick is to appeal to the good without making the right a telelogical function of the good. My take has been this: Practical reason imposes upon us a norm of consistency and thereby makes certain principles of action rationally (and hence morally ) compulsory. But it does so by incorporating into the standpoint of practical reason the notion of publicity, that we act on principles that all rational agents, who are themselves sources of value worthy of our respect, can endorse. These constraints acknowledge the good without yielding maximizing teleology. I’m not sure if this view makes the right prior to the good, the good prior to the right, or what. I guess I think of them as equally essential and complementary. And I’ve often felt like this is what Kant was driving at with the Kingdom of Ends. (This view is realist in that it holds that rational autonomy is good indepdendent of its having value conferred upon it by practical agents, so it’s not Korsgaard’s view, I gather.)
They have no choice if they think: (a) that there is a reason to do the things that they think they ought to do; and (b) that if there is a reason to desire something then that thing is (by definition) good (ie they’re buck-passers).
Oops. I meant (a) to read “that there is a reason *to desire* to do the things that they think they ought to do.” So, to repeat, they have no choice if they think: (a) that there is a reason to desire to do the things that they think they ought to do; and (b) that if there is a reason to desire something then that thing is (by definition) good (ie they’re buck-passers).
Hi Robert,
Great stuff! I think there’s commonly a bit of a conflation going on in these discussions. Specifically, the idea that rightness is value-based is often conflated with teleology (or consequentialism), the view that (in your words) “the reason you must do [some act] is that you will realize or produce some good.” Folks like Herman and Guyer seem to run these together. But in addition to the possibilities that (1) rightness is primitive and that (2) what makes an act right is that it produces some good, there is a third possible view (or maybe a fourth, counting the one Michael Cholbi suggests). According to this view, (3) the value of humanity underwrites rightness, but we are not to produce or otherwise realize it; rather, acts are right because they somehow express (or “honor,” but I take honor to be only one kind of expression) the already-existing value of humanity. For lack of a better term, I call this a “value-based deontology”: it is value-based (rightness stems from the value of humanity), but it is also non-teleological and arguably a paradigm deontological theory of rightness (contra, say, Herman).
As non-teleological, view (3) escapes your criticism of teleology: “a necessary requirement — something you must do — is wholly inconsistent with the idea that the reason you must do it is that you will realize or produce some good.” Although, as a separate matter, can’t these two be reconciled? If we agree with Kant that all action has an end, and if we think that all ends are values (whether in the broader or more narrow teleological sense), then necessary requirements aren’t inconsistent with value-based theories. Arguably, at any rate, necessary requirements are independent not of value, but of our inclinations to bring about objects of value (assuming no inclination is necessary). An obligation can’t be universal and necessarily binding if its bindingness stems from my inclination; but it might be universal and necessarily binding and at the same time be directed at an object with inclination-independent value. If that’s right, then value-based theories of rightness are consistent with necessary requirements.
Michael C.,
Your rendition sounds quite Rawlsian in a way: accept a thin conception of the good, use it plus some assumptions and a theory of practical reason to generate the thicker conception of good. Never liked that. Cheating, isn’t it?
Michael S.,
That sounds right, that Kantians accept this theory of reasons, a theory that requires them to find some good in conforming to obligations. But can’t they accept (a) and (b) without being teleologists? That is, while the things there is reason to desire to do are good, they are good just because there is reason to desire to do them. And so categorically binding reasons are not held hostage to the good. If there is a categorically binding reason to desire to do something, then it is, just because of this, good absolutely.
Josh G. (since I’m using initials!),
Right, I didn’t mean to bind the way in which there must be some good that comes of conforming to obligations to just those two standard ways. I’m not sure I completely grasp your alternative, but I will say this: it all comes down to that ‘basing’ relation, as it were–how the right can be ‘based in’ the good. The sense at issue I believe is the sense in which teleological views say the good is a — indeed the only — source of reasons. This is a view that Kant himself rejected (well, that is controversial now I suppose). He rejected it (or so I think) precisely because he thought it inconsistent with categorical ‘oughts’. And that is because an independently defined good could only be a contingent source of reasons (or perhaps a source of contingent reasons) to conform to obligations. Moral obligation is (say Kantians) necessity: something you must do. So no matter what the circumstance, there is a reason to do it. But if you’ve got some independently defined good, the only (non-mysterious) reason-giving bridge you could build from that to acting will be a contingent one–my act produces that good, or realizes that good, or ________s that good. The puzzle is, How could there be an independently defined good that could give you a reason to do something no matter what (in all possible moral worlds, let us say)?
Surely someone would read this and say, Well, I don’t know f.a. about moral necessity! Just give up them crazy categorical imperatives and learn to love contingency.
Is that true? It would be an interesting moral theorem, if it were. But why isn’t it a necessary moral requirement for utilitarians to maximize overall utility? There is an objection from M. Smith that an act maximizes utility only relative to a set of alternatives, while for Kantians, the necessary requirement not to perform certain actions is grounded in their intrinsic properties. In short, no action has the intrinsic property of being utility maximizing, and in that sense no action is necessarily required for utilitarians.
But I’m not persuaded. No action has the intrinsic property of being an instance of promise-breaking or being an instance of lying, either. I’d be lying if I said sincerely with the pretext to convey information truthfully that “it’s raining right now”. But the action does not have the intrinsic property of being an instance of lying. Whether it is an instance of lying depends on lots of other contingent facts apart from the action. So Kantian necessary requirements (e.g. not to lie) are no more genuine than Utilitarian necessary requirements (e.g. to maximize utility).
I’m going to dip my toe in gently here, because I have retained my amateur standing.
But I think the insight about our need to resolve paradox (like good versus right, or autonomy versus heteronomy) versus our ability to resolve paradox was Kant’s most fundamental contribution. And the higher level unresolvable paradox is that we sense (intuitively) the rightness of CI, yet almost any human attempt to declare a particular act as universally acceptable or universally reprehensible falls short. On one hand, Kant himself recognized the difficulty of trying to separate acting on one’s own pure moral necessities versus one’s inclination, or trying to nail down what it means to be free, but on the other hand, it seems to me, betrayed (alas, he was human!) his own crucial insight when he declared particular acts (as opposed to the principle) to be categorical.
And yet here we are, two hundred plus years later, still trying to reconcile that paradox.
Mike,
Even Hare, perhaps the most Kantian of consequentialists, still held that it was contingent fact that slavery, torture, genocide — the most heinous acts — were wrong. That’s because it’s just a contingent fact for consequentialists that any given thing optmizes. Things might have been quite different. Kantians are often thought to hold that things couldn’t have been quite different, that there are some things you may not do no matter what–this is the sort of necessity we’re talking about. I know many (e.g., Michael S) find this view preposterous, but there are other ways of understanding the necessity invoked by Kantians that are less preposterous. For instance, some (e.g., me) hold that what is necessary is that you not choose against your own deepest convictions about what to do, and that you never reason about what to do in such a way that you’re regarding persons as having ‘weights’ to be put on a scale. Here, the idea isn’t that some types of actions are necessarily wrong, but that you necesssarily not reason to what to do in certain ways.
Robert,
I’m not sure I see the diagreement. It’s a contingent fact, on utilitarian grounds, that slavery, torture, genocide — the most heinous acts — are wrong. I agree entirely. What I said was that, on utilitarian grounds, ‘maximizing utility’ is necessarily required (perhaps it would be better to say that maximizing utility is required in every world in which utilitarianism is true, but suppose we skip the nicer distinctions) I can’t see how that could be mistaken. Then again, there is the charge, noted above, that no action satisfies the maximizing standard in every world in which it occurs (i.e., no actions are intrinsically maximizing actions). Yes, agreed. But you add,
For instance, some (e.g., me) hold that what is necessary is that you not choose against your own deepest convictions about what to do, and that you never reason about what to do in such a way that you’re regarding persons as having ‘weights’ to be put on a scale.
But if you are necessarily required not to choose against your convictions, then, since your deepest convictions vary from world to world (e.g., in some worlds, I’m sure, you’re by conviction a pro-lifer and in other worlds you’re by conviction a pro-choicer), the actions satisfying that standard will vary from world to world. So there are no actions that satisfy your standard in every world in which they occur. But then, your standard expresses a necessary requirement only if the utilitarian standard expresses one. That’s the conclusion I was after.
Mike,
I guess consequentialists used to say that it was a necessary truth that we ought to maximize the good. But that was because they thought of this as analytic. But few hold that view anymore. Perhaps some think it is some sort of metaphysical truth that connects ‘ought’ to ‘maximize’. But it’s probably safer to think it’s contingent, if true at all. It’s at least conceivable that we ought not do that.
Right, convictions vary. It isn’t the content that necessary. It’s the procedure.
Robert,
I always understood Kant, in rejecting “value-based” views (in the second Critique, which I take it was the passage you were referring to), to be rejecting the view that moral obligations can be generated heteronomously. But there he also seems to then conflate the (teleological) view that the right action is right because it brings about the good with the (heteronomous) view that it is right because it brings about the object of desire.
But only the second is incompatible with categorical obligations. I see now that by “categorical” and “necessary” you mean the view that obligations cannot fluctuate with circumstance. Even if we use “categorical” in this way, some value-based views seem able to generate categorical imperatives. They just have to say, as Mike suggests, that there is some right-basing value (e.g., happiness) that holds in all circumstances. Wasn’t something close to this essentially Kant’s own argument when he put forth the claim that humanity has “absolute value”? That is, that since all action has an end, categorical imperatives in fact require some value.
Of course, as you point out, there are lots of understandings of “categorical.” I usually follow M. Singer’s reading of the Groundwork that “categorical” means “independently of inclination.” (It’s not that our obligations hold no matter what, but no matter what we are inclined to do.) This, on Kant’s parsing, gets us universality and necessity. If this is the privileged understanding of “categorical”, then we can again base rightness on value, so long as that value is determined independently of inclination or desire. That, then, would still be consistent with the principle that morality cannot be heteronomous. And, put in your terms, it would also be consistent with the idea that rightness not be based on anything contingent (maybe I’m going broken-record here, but inclination/desire would be a contingent basis for rightness; value need not be a contingent basis for rightness, if as Kant suggested of humanity, its value does not depend on us having an inclination/desire for it).
(Side Note: Maybe I’ve misunderstood your point. You say “The sense at issue I believe is the sense in which teleological views say the good is a — indeed the only — source of reasons.” But on another prominent construal of “teleology” (one that is arguably the construal of folks like Guyer and that I thought we were talking about with such claims as whether “the good [is] the ground of obligation”), teleology claims not that the good is a source of reasons, but that it is the basis for obligations or rightness. I only mean to be talking about the right-basing sense of “teleology.” If instead we’re talking about the reason-basing sense of “teleology”, then — without some further premises linking obligations to reasons — I take it all back!)
Robert,
Sure it’s conceivable that the utilitarian principle is false. Who would deny it? But it is also conceivable that your procedure is mistaken. Surely, there is no contradiction in the supposition that your procedure is mistaken; the procedure is not knowable a priori. So we have again a parallel between your view and theirs. But this doesn’t have much to do with whether the utilitarian principle is metaphysically necessary. For all anyone knows, of course, it’s a posteriori necessary. So if the claim is that your view is either a priori or a posteriori necessary and the utilitarian view is not, then I see no epistemic advantage to your position over theirs. Am I supposed to be seeing a difference?
Josh,
I don’t know what ‘is the basis for’ or ‘grounds’ mean unless they mean ‘provides a reason or rationale for’. At any rate, that is what these Kantians mean by it (see, for instance, how Barbara Herman treats it). Any other thing they may have in mind I’m not all that interested in.
What Kant thought? Hard to say in the end. But he certainly thought that absolute requirements are possible only if we are autonomous, and autonomy is inconsistent with being governed by an independently existing good. As you say, he also thought that such an independently existing good would have to give us desire based reasons. But his main line of argument is that they are not reasons that could support a moral necessity.
Mike,
Yes, you should see a difference. Utilitarians I take it hold that the principle connecting ought with maximizing is true, but contingently so — and, by the way, that’s a good thing in their eyes. Kantians take the required procedure not only to be correct but necessarily so. Either or both might be wrong of course. But what they would be wrong about would have different modal statuses.
Robert, sorry, what I wrote must have been ambiguous. I meant to contrast “provides a reason for” not with “grounds,” but with “provides an obligation.” So to say that a view is value-based might mean either that value is why the agent has a reason to perform act A, or that value is why the agent has an obligation to perform A. Of course, some think that every obligation provides a reason to act, but that’s a further step. All I wanted to point to is that my comments are about the second kind of value-based view, not the first. For what it’s worth, I think the folks you’re targeting are often ambiguous on this question: sometimes they’re talking about reasons, other times about obligations (e.g., when they’re talking about the formula of humanity, which is a principle of obligation; or, although I don’t have my book in front of me, I have a foggy recollection that Guyer explicitly talks about rightness, rather than reasons).
Josh,
I see the ambiguity, but I see a difference in the two “Why’s” — the former being explanatory, the latter being justificatory. In any case, I’m pretty sure Guyer is thinking this way at least in his responses to Korsgaard’s later work, as well as the others at points, though I agree it is not always clear everyone is talking to each other.
I could easily be mistaken, but it seems the train of thought is this: We start (ala Glaucon, etc.) asking what reason there is to do what we ought. We then are tempted to say that there must be some good — initially, for us — in doing it. That then is supposed to make it rational to do our duty. We’ve now quickly arrived at teleology. Then you get intuitionists such as Pritchard defending a Kantian-style view and so claiming the whole question is vexed. It makes no sense to ask for a reason to be moral. So we don’t have to be teleologists. But now Kantians seem saddled with a bizarre view: You can be obligated but can’t provide a reason to act? That then brings us to the contemporary scene, in which Kantians are backpedalling, trying to say, oh yes there is some good in acting morally, just not the consequentialist kind. That all just seems to accept, whole hog, the view that the good is the sole source of reasons. Or at least that’s what I’ve been urging.
Fair enough, Robert: sounds like we’ve been talking past one another, with you discussing value-based theories of rationality and me discussing value-based theories of rightness (although I take those to be explanatory too). I guess it’s an interesting side-question as to how the folks you’re targeting should be read. I agree that Korsgaard focuses on reasons, but I’ve always thought that the best interpretation of people like Herman and Guyer is that they are engaging the debate (with, e.g., consequentialists) over what makes right acts right, rather than what makes it rational to do what’s right. More food for thought, I guess. (Of course, now all of what I said about value-based theories of obligation is non-responsive to your question of whether Kantians should be teleologists with respect to reasons!)
Robert
I was surprised by this:
In which worlds does the maximising principle not hold, according to utilitarians?
Even if utilitarians deny that the proposition “You ought to maximise the good” is necessary, there are other propositions in the near vicinity which they would, I think, accept as necessary: e.g. “Actually, you ought to maximise the good”, or “In worlds like this, you ought to maximise the good”. Right?
Campbell,
Right, I’ll gladly take that back. But I was just surmising that the principle should be contingent, so that it would not turn out to be analytic — that would be a bad outcome, so it should be a welcome addition to a utilitarian. As I said, some do hold it to be, nevertheless, necessary in some other sense. For instance, some claim that it is just bizarre to think it could be wrong to do what is best. That may make the maximizing view seem therefore necessary. But, aside from that intuition, the necessity claim doesn’t sound appealing so to me, but perhaps it does to others. Maybe we should think it is a metaphysical fact about the property of rightness that it is identical with maximizing the good.
In any case, this isn’t the necessity that is at issue between those who think ‘oughts’ are moral necessities and those who do not. The issue is that the fact that some action maximizes is a contingent matter. The essential properties of any action could be the same and yet it not have maximized. So that very same action could have been the wrong thing to do. Kantians deny this, at least in some sense, and so they reject the maximizing principle –and any other view which would imply that there are no moral necessities.
Also, I think I now see why Mike A. was saying there might be one moral necessity for utilitarians: maximize the good. I was confusing that with the theoretical claim that what we ought to do is to maximize. I can’t recall who at the moment, but I seem to remember someone holding this view. It’s an interesting view because it looks as if there is then one ‘ought’ that is not grounded in maximizing the good, namely, that you ought to maximize the good. But I don’t know what else to say about it.
I think Jackson and Smith have a good line on this. As I recall, they say that a moral theory implies an “absolute prohibition” iff there’s some kind of act, characterised solely in terms of intrinsic properties of acts, such that the theory implies that acts of that kind are always wrong.
The “intrinsic properties” clause is added for the following reason. Consider this kind of act: acts that fail to maximise total utility. Utilitarianism implies that acts of this kind are always wrong. (Moreover, it implies that an act of this kind is wrong even if it’s required to prevent a greater number of acts of this kind.) Yet whether an act fails to maximise total utility doesn’t depend solely on the act’s intrinsic properties. So this is not an absolute prohibition, on the Jackson-Smith definition.
Robert,
You say,
the fact that some action maximizes [the good] is a
contingent matter.
Yes, but what the consequentialist claims, if I understand Mike and others rightly, is not that some particular act is necessarily related to maximizing the good. Rather, it is that maximizing the good is necessarily related to what one ought to do.
You might say that this is still a far cry from showing how some particular “act” (in the token sense?) is necessary to do. For example, it doesn’t show that making false promises is necessary. But surely the consequentialist would respond that what you call “making false promises” could just as easily fall under another description, either “maximing the good” or “not maximizing the good.” Thus we have a consequentialist account that shows certain actions to be necessary or not.
You (and Kant) seem to want to say that some descriptions of the same act are somehow more appropriate for certain moral purposes than others. “Maximizing the good” may be a weaselly way of describing what could be more carefully characterized as “making a false promise.” But I’m not sure why the latter description should somehow be more appropriate than the former, for purposes of determining moral necessity.
Perhaps we could tell some story about the way a description is fixed by the “maxim of one’s will,” but I’m not certain.
Campbell, I note this at [December 17, 2006 at 06:13 AM] above. I don’t find the distinction in S&J helpful. I agree that no particular action has the intrinsic property of being utility maxmizing. But I say,
Of course I agree that a full orbed description of action A such as, A = “S’s knowingly expressing a false proposition with the pretext to convey information to P at t” has the intrinsic property of being an instance of lying. But it is also true that the action A = “S’s maximizing overall utility at t” has the intrinsic property of being utility maximizing. So Kantians and utilitarians are in the same boat on the S&J criterion.
Mike
Sorry, somehow I missed your earlier comment.
Perhaps my glossing J&S’s proposal in terms of intrinsic properties was misleading, for the reasons you give. It might be better to say that the kind of act in question must be characterised without reference to the alternatives. To say that an act maximises overall utility is to say something about the alternatives: that none of them would result in greater overall utility. Whereas to say that an act breaks a promise is not to say anything about the alternatives.
This is an interesting twist. I guess I don’t think saying that an action maximizes overall utility does say anything about alternatives. The utilitarian says, in every possible world, regardless of the alternatives available, the action(s) that maximizes utility is required. The Kantian says, in every possible world, regardless of the alternatives available, the action(s) that is consistently universalizable is required.
In neither case is there a world in which the alternatives are relevant to what you ought to do.
Mike
I can do no better than quote directly from Jackson and Smith (“Absolutist Moral Theories and Uncertainty”, J Phil 2006):
Do you really mean to deny that “an action’s failing to maximise utility is a relation the action has to available alternatives”?
If I may, I’d like to return to an earlier point in the thread. I had said to Robert, at the beginning of the thread, that Kantians have no choice but to be teleologists if they think: (a) that there is a reason to desire to do the things that they think they ought to do; and (b) that if there is a reason to desire something then that thing is (by definition) good (ie they’re buck-passers).
Robert said that this sounds right: Kantians accept this theory of reasons, a theory that requires them to find some good in conforming to obligations. But, he asked, can’t they accept (a) and (b) without being teleologists? That is, while the things there is reason to desire to do are good, can’t they hold that they are good just because there is reason to desire to do them? If so, then categorically binding reasons are not held hostage to the good. Rather, if there is a categorically binding reason to desire to do something, then it is, just because of this, good absolutely.
But I think it might be a good idea if we distinguish between two kinds of reasons in evaluating what Robert said. On the one hand, there are reasons for action. On the other hand, there are reasons for desiring. The important question, in evaluating what Robert said, is how these two kinds of reason relate to each other.
Just to keep things simple, let’s suppose that the only reason for desiring something is provided by the nature of the exercise of our rational capacities: because the exercise of our rational capacities has the nature that it has, there is a reason for everyone to desire that they exercise their own rational capacities and that they do not thereby interfere with others’ exercise of their rational capacities (at least in so far as those others are doing likewise).
As Robert implicitly admits, on this supposition there are two good things—the exercise of one’s own rational capacities and one’s not interfering with others’ exercise of their rational capacities–where the categorically binding reasons that there are to desire these two good things are not themselves hostage to the good. Rather, according to the story just told, the good of these two good things is constituted by the existence of such categorically binding reasons to desire them.
But though that’s true, note that our reasons for action, as distinct from our reasons for desiring, would be hostage to the good. For the only things that we would have reason to do, according to such a story, are those things, among our options, that produce some good (ie those things, among our options, that produce some outcome that there is categorically binding reason to desire). So I think that the view I described at the outset is more telelogical about reasons for action than Robert admits. In a nutshell, reasons for action are hostage to what there is reason to desire (= what is good).
Moreover, I think that the view may also more ultimately reduce reasons for desiring to facts about the good (in a non-buck-passing sense). After all, a good question to ask, at this point, is what that feature is in virtue of which certain considerations do, while other considerations don’t, provide reasons for desiring. The reason for asking this question is because the standard story about reasons for (say) belief–something about truth conduciveness–simply cannot be told about such reasons, as these are reasons for desiring, not reasons for believing. So what exactly does it mean to say that a certain consideration provides a reason for desiring?
One answer–and perhaps this was Kant’s own–is that that in virtue of which the considerations that are reasons for desiring count as reasons for desiring is that there is a transition from contemplation of those considerations to desiring in the perfect psychology, where the concept of perfection is a distinct conception of goodness to the buck-passing conception alluded to earlier. (The idea is not, at least according to the suggestion on offer, that what makes a psychology perfect is that there is a reason to desire it, for if it were then we would have no explanation at all of that in virtue of which some consideration is a reason for desiring: we would have gone in a circle.)
However, if this is right, then note that the concept that is doing all of the explanatory work is the concept of perfection as applied to psychologies. Reasons for desiring are themselves hostage to the good. They are hostage to the good because they are defined as those considerations, contemplation of which in a perfect psychology, leads to desiring. (This sounds, to me at any rate, a lot like the sort of role that Kant wanted the good will to play, where the good will, as distinct from other goods, is good unconditionally.)
Michael,
Thanks for making all of that explicit, what is probably implicit among Kantian teleologists. I see the challenge, and don’t know exactly what I want to say. But off the cuff I’m inclined to say something along these lines:
You have the explanations all bottoming out in the concept of perfection, itself a conception of the good. This is very much in the spirit of Barbara Herman’s views. And you may say that explanations have to end somewhere, and, thus, so why not here.
But there is a reason to resist having the explanation bottom out here. For is it really circular, or at least circular in a way that offers no explanation, to explain why something is a reason for desiring in terms of the fact that the psychology that a perfect psychology would want to possess does or would in fact desire it? I don’t think that it is. That a psychology operating in the way in which a perfect psychology operates also, in addition, wants to operate in that way says something informative: It says that it endorses its own operations, as it were. Imagine a psychology that operates in a particular manner, but reject or is averse to psychologies operating in that manner. That is a psychology at war with itself. So it seems to me that there is a need to say something more, that the good will, or perfect psychology, is not the last word, since a necessary bit to make it perfect is that it (as it were) endorses itself.
Campbell,
I’m familiar with the text. I of course agree that you did not misrepresent what they said. But I never suggested you did.
What I mean to say is that I do not see the alleged distinction between actions that fall under kind K = maximizing overall utility and actions that fall under kind K’= speaking non-lies. I claim that those KINDS of actions are necessarily required.
The passage you quote claims that whether a particular action A maximizes utility depends on its relational properties (or, relations, if you like). In particular, on its relation to other alternatives in a given situation. My claim was that whether that’s true depends on the description of A. Suppose A = ‘S maximizes overall utility’. In that case, is it true that whether A maximizes utility depends on it’s alternatives? No, it doesn’t, since it is necessary that action A maximizes utility. That is, ‘S’s maximizing utility’ entails ‘utility is maximized’. A in fact has the intrinsic (non-relational) property of maximizing utility.
So, again, I’m not disputing that you’ve got the position stated correctly. Rather I’m denying that the alleged distinction in the position.
Robert asks: “[I]s it really circular, or at least circular in a way that offers no explanation, to explain why something is a reason for desiring in terms of the fact that the psychology that a perfect psychology would want to possess does or would in fact desire it? I don’t think that it is.” I’m not quite sure that I understand what the suggestion is, so let me try to figure it out by working through the example I gave in my post.
We’re assuming, for the purposes of argument, that the nature of the exercise of our rational capacities does indeed provide us all with reasons to desire that we exercise those very capacities. What we want an explanation of is why the nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities. Why does it provide us with a reason to desire that, as opposed to a reason to desire that we never exercise our rational capacities, as opposed to a reason to do anything else?
Robert’s suggested explanation seems to be that someone with a perfect psychology would want to possess a psychology with that very nature: that is, a psychology in which its rational capacities are exercised. The trouble is that though that’s true, given the assumption we’re making, it doesn’t sound like it could be an explanation of what we wanted explained. It couldn’t be an explanation because it just a statement of the assumption that we were making in asking for an explanation.
The assumption, to repeat, is that the nature of the exercise of our rational capacities provides us all with reasons to desire that we exercise those capacities. True enough, given this assumption it follows that someone who is perfectly rational would therefore have and exercise the capacity to grasp this reason–that is, they would grasp the fact that the exercise of our rational capacities does indeed have the nature that is has–and they would then proceed to desire the thing that this reason provides us all with reasons to desire: namely, the exercise of those very capacities. But that just leaves us in the same situation we were in to begin with. Why does the nature of the exercise of our rational capacities provide us all with a reason to desire the exercise of those capacities, as opposed to a reason to desire that we never exercise those capacities, as opposed to a reason to do anything else?
Robert might be suggesting that the property of being perfect is just the property of being desired by those with a perfect psychology, where what makes the psychology of a creature with a perfect psychology perfect is the very fact that it desires its own psychology. But if this is the suggestion then we’re in familiar territory, territory that Rae Langton explores in her forthcoming Philosophical Review paper ” Objective and Unconditioned Value.” For the claim would then be much the same–and hence would be just as implausible–as the claim that the good is what is loved by the gods, where the gods are in turn good, but good precisely because they love themselves.
Michael admits that we would indeed want an explanation of why (if it does) the nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities. But he offers none. How then does one get reasons out of this nature?
I have offered a partial, perhaps only necessary element of an, explanation of why. But Michael’s rendition is not quite my proposal. I do not say that being a perfect psychology just is being one desired by a perfect psychology. Rather, the approval, endorsement or desires of a given psychology can be the source of value only if it also desires, approves of, or endorses itself as such a source. That, at any rate, offers less mystery, if it doesn’t eliminate it. It does say something informative, and is not a mere triviality.
By the way, the analogy with Euthyphro should not be construed as Michael does, viz., as the claim that the good is what is loved by the gods, where the gods are in turn good, but good precisely because they love themselves. It is that the good is what is loved by the gods, and the gods love that the good is what is loved by themeselves. They approve of the fact that what is good is what they approve of. That’s what makes it a good standard of value. They don’t just approve of themselves.
I think this is relevant to the discussion between Campbell and Mike A, so I’m going to throw it out there. There is an important and very-little-discussed distinction between two kinds of consequentialist position.
According to one, there is a single act-type which everyone is at all times obliged to perform: maximize the good. According to the other, there is no such obligation. But whenever anyone ought to do something, that is explained by the fact that her doing it will result in the best outcome.
On the first view, there is an act type that is obligatory no matter what its alternatives. It is the one Mike A. mentions. Of course, it is contingent just _how_ to do this thing that we are always obliged to do, and for any way of doing it, whether it is a way of doing it depends on _its_ alternatives. But whether you ought to maximize the good does not turn on what _its_ alternatives are. But on the second view, there is no such act type.
The first view is a very natural way for deontologists to think about consequentialism – it’s just the view that there is just one obligation, to bring about the most good – but probably most consequentialist views are of the second type, and I suspect that Campbell likes the second view better.
Let me see if I can get the hang of the second-person address convention in a blog! Robert, you say:
I do not say that being a perfect psychology just is being one desired by a perfect psychology. Rather, the approval, endorsement or desires of a given psychology can be the source of value only if it also desires, approves of, or endorses itself as such a source.
Michael admits that we would indeed want an explanation of why (if it does) the nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities. But he offers none.
The proposal I offered was in the earlier post. The nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities because in a psychology that is perfect there is a transition from the belief that the exercise of rational capacities has that nature to a desire to exercise those capacities. The crucial feature of this proposal is thus that it makes use of an independent concept of perfection, a concept quite different from the buck-passing conception, and defines reasons by using that independent concept of perfection.
The proposal is thus that there are two kinds of goodness. There is conditional goodness: this is the kind of goodness possessed by things in virtue of the fact that there is a reason to desire those things. And there is also unconditional goodness: this is the kind of goodness possessed by those psychologies that we need to appeal to when we explain what it is for a consideration to be a reason to desire something. We thus explain conditional goodness by appeal to unconditional goodness. (This is all supposed to sound very reminiscent of Kant.)
You continue:
Moreover you say that this says “something informative.”
As I understand it, you therefore start by going along with my proposal. a consideration is a reason for desiring something, you seem to be suggesting, because in a perfect psychology there is a transition from the belief that that consideration obtains to a desire for that thing. But your explanation then differs from mine. You don’t want to explain what it is for a psychology to be perfect in terms of an independent conception of perfection. That’s the view you want to resist. Rather you suggest that a certain kind of psychology is perfect because it is both “the source of value and it desires, approves of, or endorses itself as such a source.” But does this say something informative? Consider the two halves of the proposal.
We already know what it is for a psychology to be the source of value. It is for it to be a psychology in which the transitions between beliefs and desires are those warranted by reasons. But if this is what it is for a psychology to be perfect, then we can hardly go on to explain what it is for a consideration to be a reason for desiring by appeal to the idea of a perfect psychology. For that would be to go in a circle. Nor does it help if we add the second part of the proposal. For to say that a psychology “desires, approves of, or endorses itself as such a source” is simply to say that it desires itself because it is a psychology in which transitions between beliefs and desires are those warranted by reasons. But, again, if what it is for a psychology to be perfect is for it to have such a desire, then we can hardly go on to explain what it is for a consideration to be a reason for desiring by appeal to the idea of a perfect psychology. For we would once again be going in a circle.
I know you must be thinking that I’m either thick or belligerent, but I just can’t see how to make the account sound like it gets off the ground. This isn’t to say that the alternative proposal, the one I offered, is trouble-free. It is simply to explain why I, at any rate, would want to resist the bootstrapping move that you seem to like.
Mike
Mark’s comment suggests to me that the source of our confusion may be an ambiguity in the notion of an alternative. Roughly, this can be understood either in terms of act-types or act-tokens. So let me try to say, in more detail, what I had in mind; perhaps that will help clear up the confusion.
Take a particular act (an act-token). Call it A. At the time A was performed, there were a number of other acts that the agent (whoever it was who performed A) might have performed instead of A. These are the alternatives of A. (Notice, on this way of thinking, the alternatives are, like A, act-tokens, not act-types.) We can, let’s suppose, compare the amount of utility produced by A with that which would have been produced by each of the alternatives. If any of the alternatives would have produced greater utility, then, and only then, A fails to maximise utility. Thus, whether A is a token of the type “acts which fail to maximise overall utility” depends on the alternatives. Call this, then, an “alternative-dependent” type. While it’s true that utilitarianism implies that acts of this type are always wrong, this doesn’t count as an absolute prohibition, on the Jackson-Smith definition, because it’s an alternative-dependent type.
Compare this with a moral view that says it’s always wrong to break a promise. This view also implies that there’s a type of act, “acts that break promises”, such that acts of this type are always wrong. But in this case, it’s not an alternative-dependent type; whether an act breaks a promise doesn’t depend on the other acts that might have been performed instead. So this is an absolute prohibition.
Michael, but I was already warming to the 3rd person; it comports well with my tendency to play to the crowd (of what, you and me?!) But 2nd it is.
What I meant to be challenging was just your own explanation:
The nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities because in a psychology that is perfect there is a transition from the belief that the exercise of rational capacities has that nature to a desire to exercise those capacities.
My proposal was that this is not yet an explanation. This explains why the nature of the exercise of our rational capacities provides us with a reason to desire the exercise of those capacities only if your perfect psychology standard is a good standard of reasons. Do we think it is desirable that reasons be just those considerations handled by the perfect psychology in this way? That question, the question whether we should be happy about the perfect psychology standard of reasons, or should, rather, find it lamentable, is perfectly understandable and can only be answered by appeal to that standard itself (unless, of course, there is some hidden, second, ‘real’ standard of reasons). Thus, the question is, In a perfect psychology, is there a transition from the belief that this is the standard of reasons to a desire to accept, preserve, employ, etc., that standard. Or is there rather a transition from the belief that this is the standard to an aversion to it, a deisire to reject it?
Now you’ll have to tell me why that isn’t a contentful question.
(Belligerent, no. Thick, definitely not. Fussy dresser? Well, yes…all of those shoes!)
It is a contentful question, but I don’t see how the answer to it has any bearing on whether reasons for desiring that p are those considerations which are such that, in a psychology that is perfect, where perfection is understood independently, there is a transition from the belief that those considerations obtain to a desire that p. The question seems rather to take that for granted. So it seems to me that what you need to explain is how the perfection of a psychology is to be understood so that it turns out that whether or not a psychology is perfect is in some way beholden to a positive answer to this question. Or, to put the point slightly differently, can you ask the question in some way that doesn’t assume that you’ve already got an independent standard of perfection?
Yes, or it seems so.
Lets say we are trying to construct this perfect psychology. At any point in our account, it will have some content that will describe a profile for this psychology; call it, for lack of a better alternative, an ‘M-psychology’, in which ‘M’ give the specification for how this psychology works at the point in the construction we are at. Now in specifying the content of M, one thing I am insisting we’d have to consider, among all of the things we have to consider, is the question of whether within M, at whatever point in constructing M we are, there would be a transition from the belief that M provides the standard of reasons to a desire to accept, and so on, etc., it as the standard. We’d have to consider this because we want to know whether M is a good standard. And if there is no such transition, then M is not the psychology we want yet; it’s not the psychology whose workings provide the standard we are after.
Doesn’t that work?
By the way, if it does, notice that, as I said initially, there is in fact no independent value, and so, for instance, the Good Will is not an independent source of value.
Is this right? All it takes for a psychology to be perfect is for it have a property P such that:
(i) it is P in virtue of being M, where M describes a set of transitions between beliefs and desires
(ii) if one’s psychology is P, then there is a transition from the belief that one’s psychology is P in virtue of being M to the desire that it is M
(iii) being M includes the transition described in (ii)
Not (iii). That is, being M shouldn’t include (ii), but (ii) should neverthless be a necessary condition of some psychology’s being P in virtue of being M.
We’ve gone in a rather different direction from ‘what Kant thought,’ but on that issue, I strongly recommend Reath’s review of Guyer’s ‘Kant on Freedom, Law, and Happiness’ (Ethics 114 (2003): 127). I can’t do it justice here, but Reath distinguishes two roles that rational agency plays in Kant’s ethics and does a very careful job indicating whether Kant’s derivations of the moral law from rational agency end up being “teleological”, etc. Do check it out.
My immediate and intuitive mental association connects priority of the right with Kant.
The sage of Konigsberg of course distinguished between inclination and value. Korsgaard, his latter-day disciple follows him, “[I]n Kant’s theory our values are created from psychological materials, from the natural bases of interest and enjoyment, rather than from nothing. Here as elsewhere in Kant’s theory, reason works by imposing form on matter that it finds…[via] really only one principle – the principle that we should choose our maxims as universal laws.”
Morality seems largely ifn’t exclusively the enterprise of going from psychological materials to maxims. (How practical reason “confers” value on everything). Such conferring of value and justification thereof is what the Kantian project is all about. John Rawls’s entitling his lecture Kantian Constructivism in Moral Theory dovetails quite well with Korsgaard’s statement hereinabove quoted that “our values are created from psychological materials…rather than from nothing.”
Hilary Putnam was quite right in his Collapse of the Fact/Value Dichotomy to point out how the logical positivists ironically by appropriating Kant’s notions of analytic and synthetic truths ended up dismissing ethics as assemblages of what cannot ever be justified by reason. Korsgaard is absolutely correct to distinguish psychological materials from nothing, but those materials seem to come – as Putnam and others have stated – already laden with value and only valuable by us because of that, not owing to an assuredly (if assumedly, for my purposes here) faulty metaphysics/psychology of faculties.
Not many of us would be willing to, I think, put down ethics like Kant put down mathematics – as a set of synthetic a priori truths. But if one accepts the Kantian way of ethics, isn’t it the case that we’ve to import the form of his metaphysics, the stuff about synthetic a priori truths?
(As an aside, but not really, Putnam senses “something in Kantianism that pushes philosophers in this direction…of skepticism about value realism”, which he finds also in Habermas’ distinguishing so much between norms and values that he rather marginalizes the latter.)
“Should Kantians become teleologists?” My inclination is to ask them, but reason dictates otherwise. Unquestionably, they’re at one level – the working one, at all events – hardcore into the telos of following the maxims, no? What the maxims say of course in actual situations to do is a subject of endless debate. (Kant’s ethics is formalist bunk!). But soberly one’s reminded of how Hannah Arendt wrote of Adolf Eichmann’s favorite philosopher being one Immanuel Kant. Undeniably, there is something dangerous in teleological Kantianism, for Eichmann went to his grave thinking he’d been a good Kantian . Is that what happens when a Kantian selects the wrong set of ethical absolutes from which to fashion her psychological material into maxims?
Does that advert to Kant’s own confessed troubles trying to show how we know the absolute call of duty, difference between good and evil, between categorical and hypothetical imperatives other than by asserting that’s how things ARE? As Iris Murdoch wrote, “When Kant tries (in the Grundlegung) to establish these ideas upon a more profound basis he admits the circularity of the argument.” She is right to link it to Ludwig Wittgenstein’s much later post-Kantian proclamation, “Ethics is transcendental”, which beautifully savages the notions of maxims and such in favor of ethics as nothing more – but nothing less – than an attitude, mood, or temper.