Broadly construed, a theory is consequentialist (or teleological, if you prefer) iff it takes the deontic status of an action to be solely a function of some transitive ordering of outcomes in terms of their goodness or desirability. Now what I want to ask is: Can a consequentialist appeal to his or her considered moral convictions when determining how outcomes are to be ordered? That is, can a consequentialist defend one ordering over another on the basis that the former but not the latter yields intuitive moral verdicts when combined with the principle “an act is permissible iff no other available outcome ranks higher than its outcome”? I use to think that the answers to such questions were obviously "no" and “no,” but a comment made by Campbell Brown on my previous post has made me rethink my position.
Let O1 designate A1’s outcome, O2 designate A2’s outcome, and so forth. Clearly, the consequentialist cannot define the deontic status of an action in terms of some ordering and than define that ordering as follows: O1 outranks O2 iff A1 is deontically superior to A2, where a supererogatory act is deontically superior to a merely permissible act and a permissible act is deontically superior to an impermissible act. That would be circular. Let’s call this form of consequentialism:
Empty Consequentialism: Come to some fixed judgment about the rightness of actions that is entirely independent of any judgments we have about the desirability of outcomes, and then define the desirability of outcomes in terms of the rightness of actions.
Interestingly, though, ruling out empty consequentialism doesn’t commit us to answering “no” to the above questions, because there are still at least two further possibilities:
Foundational Consequentialism: Come to some fixed judgment about the desirability of outcomes that is completely independent of any pre-theoretical judgments that we have about the rightness of actions, and then define rightness in terms of the desirability of outcomes.
Coherence Consequentialism: Keeping the consequentialist principle “an act is permissible iff no other available outcome ranks higher than its outcome” constant, revise both our pre-theoretical judgments about the desirability outcomes and our pre-theoretical judgments about the rightness of actions in light of each other until reflective equilibrium is reached.
Unlike both the foundational consequentialist and the empty consequentialist, the coherence consequentialist doesn’t consider either her pre-theoretical judgments about the desirability of outcomes or her pre-theoretical judgments about the rightness of actions to be fixed starting points. Whereas the foundational consequentialist is unwilling to revise her judgments about the desirability of outcomes in light of any potentially counter-intuitive implications it may have when combined with consequentialist principle, the coherence consequentialist is willing to do so. And whereas the empty consequentialist is unwilling to revise her pre-theoretical judgments about the rightness of actions in light of any potential conflict that may arise when she combines her pre-theoretical judgments about the desirability of outcomes with the consequentialist principle, the coherence consequentialist is willing to do so.
Now if we think that coherence consequentialism is a legitimate, non-circular, and potentially informative form of consequentialism, then we should answer “yes” to the above questions. And it seems to me that coherence consequentialism is all these things.
So does anyone see a reason for thinking that coherence consequentialism is an illegitimate form of consequentialism?
(An aside: Coherence consequentialism employs the method of narrow reflective equilibrium, but it may itself be the product of wide reflective equilibrium, where all our judgments, including the consequentialist principle, are open to possible revision.)
Just a short comment. Brad Hooker is of course a rule-consequentialist but if I am not too mistaken about his view and have understood your description of coherence consequentialism correctly (not sure about either), I think he actually is a coherence (rule-)consequentialist in your classification.
He describes his aim of reflective equilibrium in the introduction of Ideal Code, Real world. And, whilst he then keeps the (rule-)consequentialist principle in tact all the way through, he seems to be trying to fit both the judgments about desirability of options to our pre-theoretical moral assessments (see his views on well-being and prioritarianism in chap. 2) *and* to change some of our pre-theoretical judgments to fit the outcomes of his theory (chap. 9).
Anyway, his view prima facie seems plausible enough not to be described as ‘illegitimate’, even though there probably are reasons for thinking that there is something problematic in all of this. At least I don’t see any obvious incoherences here.
Thanks Jussi. This is helpful. I hadn’t made the connection myself, but I think that you’re right: Hooker is, on my terminology, a coherence rule-consequentialist. This is yet more evidence that, contrary to many standard definitions of consequentialism (and teleology), the consequentialist/teleologist needn’t specify the good in a way that’s completely independent of (or prior to) the right.
Here’s my worry. I’ve been convinced by John Broome and Jamie Dreier that you can track any view about rightness in a “consequentialist” framework by allowing yourself to specify or rank outcomes in an agent-relative and time-relative manner.
But, since that would allow us to class any view as consequentialist, and since the term was coined to track a certain intuitive contrast between views (well, not really insofar as Anscombe seemed to have coined it and her application of it to Ross leads me to think it doesn’t track any intuitive contrast*), we now need a new word.
So I’d be inclined to keep the old word rather than coin a new one and just stipulate that we won’t allow agent and time-relative specifications and or orderings of outcomes.
If we don’t do that though, it seems like we don’t have a principled reason to resist Doug’s suggestion.
Am I missing something?
*I think though, that by the time Williams uses the term it does mark a certain intuitive contrast, and this is the contrast I think we need a word for.
I guess I’m confused about this. Consequentialism is not a view about moral epistemology. Maybe there’s a way to distinguish consequentialism from non-consequentialism by appealing to the “priority” of goodness. But the sort of priority in question is not epistemic. You can say that goodness is “logically” prior to rightness without committing yourself to any view about moral epistemology, right? So I guess what I’m saying is that I agree with Doug that there is nothing problematic with combining a coherence theory of justification with consequentialism, but I wouldn’t say that the result is a different sort of consequentialism.
Mark,
Thanks for the comment. Perhaps, it will be helpful if I explain where I eventually want to go with all this. But first let me say that I’m one of those consequentialists (but you can call me a teleologist) that defends consequentialism by employing the consequentializing strategy. For instance, I believe that, in order to accommodate constraints, the consequentialist should construe an act’s outcome broadly and hold that the ordering of outcomes is relativised to agents such that, for instance, the outcome where I commit one murder to prevent five others from each committing murder is outranked, on my relative ordering, by the outcome where I refrain from doing so. Now some philosophers object to this sort of strategy saying one of two things. First, some object on the grounds that it is somehow illegitimate for the consequentialist to appeal to our considered moral convictions about the rightness of actions in determining how outcomes should be ordered in terms of their desirability. They follow Rawls here in insisting that, in order to avoid circularity, the consequentialist/teleologist must define or specify how outcomes are to be ordered without appealing to our judgments about the rightness of actions. Second, some object (and I take you to be one of them) that consequentialism will “no longer give us any normative advice” if the consequentialist construes outcomes broadly and employs this consequentializing strategy in defending the view against potential counter-intuitive implications (see Mark van Roojen, “Plausibility of Satisficing,” pp. 162-5). In response to the first type of objection, I would argue that the consequentialist doesn’t have to eschew all appeals to our deontic judgments in order to avoid circularity; rather, the consequentialist need only eschew empty consequentialism. In response to the second type of objection, I would again point out that consequentialists are not straddled with the choice of either empty consequentialism or foundational consequentialism; they can instead adopt coherence consequentialism. Unlike empty consequentialism, coherence consequentialism can yield new and surprising normative information, for, in employing the method of reflective equilibrium, we may, for instance, find that we’ll need to revise many of our pre-theoretical intuitions about what’s right and wrong. But unlike foundational consequentialism, coherence consequentialism will be sensitive to our pre-theoretical deontic judgments and so will likely avoid many of the counter-intuitive implications that act-utilitarianism has.
Ben,
What do you want to say about empty consequentialism?
Even if empty consequentialism, foundational consequentialism, and coherence consequentialism are not different kinds of consequentialism, they are different kinds of views. And empty consequentialism is the sort of view that most consequentialists would and should want to distant themselves from — after all, I think that Foot is an empty consequentialist. So I want to spell out what the alternatives to empty consequentialism might be, and consider whether any of the alternatives can be “sensitive” to our pre-theoretical moral convictions.
I share Ben’s confusion. (That is, if Ben doesn’t mind sharing.)
Empty Consequentialism seems to describe a procedure for coming to a view, rather than describing a view.
It’s one question whether the goodness of outcomes makes acts right or their rightness makes their outcomes good (well, maybe that’s a real question, let’s assume it is), and quite another how we theorists should set about figuring out which outcomes are good and which acts are right. So it’s quite plausible that we figure out which outcomes are good by the method of reflective equilibrium, including in our data which acts seem to be right and wrong, even if the consequences of actions make them right and wrong.
This is all to say that there is no problem at all combining coherentist epistemology with foundational metaphysics or normative structure.
No problem Jamie, I like sharing.
Doug, when you introduce empty consequentialism, you seem worried about a circularity problem. But as long as all the theses about goodness and rightness are just stated as biconditionals, there’s no circularity problem. A iff B means the same thing as B iff A.
Maybe we could understand empty consequentialism as a view involving the “in virtue of” relation. The view would be that an outcome gets its value *in virtue of* some facts about the deontic statuses of actions. I don’t know exactly how this would work. Maybe we could say that one outcome O1 is better than another O2 in virtue of the fact that, if A1 and A2 were the only alternatives, and O1 was the outcome of A1 and O2 was the outcome of A2, then it would be obligatory to do A1. This view would be incompatible with the view that what makes an act right is the fact that its outcome is best. It would be compatible with the consequentialist biconditional, *x is right iff x maximizes outcome value*. I guess the question would be what this in virtue of relation amounts to, and whether consequentialists should be worried about what gets to be true in virtue of what.
Does this seem like what you had in mind by empty consequentialism?
Doug’s proposals seem to me to make sense as distinct deliberative procedures. That is one well-entrenched way of understanding the difference between consequentialism and deontology (at least since Rawls). That is not a view about what it is *in virtue of which* an action is right, but a view about how to reason in practical situations to conclusions about what to do. Given that, it doesn’t seem confusing to me how he’s laid out the alternatives. One way of determining what the right thing to do is, is to go by (say) Rossian intuitions about what is right and regard whatever happens as a result ‘the best outcome’. Here, it’s of course true that the right produces the best outcome, but only in an empty sense. Another way to deliberate is to figure out what outcomes you think would be best, then do whatever produces them. Another: first figure out the outcomes, then factor in your Rossian intuitions about doing whatever produces the best. Is that so confusing? The only question is whether there’s something suspicious about the last as a form of consequentialism. But seems to me it could be called that, or a hybrid.
Of course, this so far leaves untouched ‘metaphysical’ questions about what makes acts right. But it is still an interesting question.
Thanks Robert. That’s a very helpful way of putting things. And hopefully that will clear up any confusion that I caused.
Robert, you think consequentialism since Rawls has often been thought of as a view in moral epistemology? I’m surprised to hear that. I guess you’re thinking of utilitarians.
I think we might distinguish, also, between the procedure a theorist adopts in constructing a theory and the procedure the theorist recommends once the theory is in place. An obvious example would be a two-level utilitarian who recommends following rules but himself always starts with consequences in his theory-building. Similarly, I might work out which consequences are good by reflecting on my Rossian intuitions, which I could regard as epistemologically basic, but when I’m finished building my theory I might just hand The Folk a list of the goods and recommend that they seek to maximize them.
I guess I share Jamie’s surprise about Robert’s comment, and about Doug’s endorsement of it. I don’t think I know any consequentialists who think that, when deciding what to do, it’s a good idea to “figure out what outcomes you think would be best, then do whatever produces them.” Not even Mill thought such a thing. On the other hand, I’m sure with a little digging I could find plenty of anti-consequentialists who attribute such a view to consequentialists.
Perhaps the following is a better way of making my point.
Suppose someone accepts the act-consequentialist’s bi-conditional: an act is permissible iff no other available outcome ranks higher than its outcome. Having accepted this bi-conditional there are three possible procedures one might employ in figuring out how to rank outcomes in terms of their desirability:
Procedure 1 (The Footian Procedure): Come to some fixed judgment about the rightness of actions that is entirely independent of any judgments one has about the desirability of outcomes, and then hold that one outcome outranks another iff its corresponding act is deontically superior to that of the other.
Procedure 2 (The Foundationalist Procedure): Come to some fixed judgment about the desirability of outcomes that is completely independent of any pre-theoretical judgments that one has about the rightness of actions.
Procedure 3 (The Coherentist Procedure): Keeping the act-consequentialist’s bi-conditional constant, revise both one’s pre-theoretical judgments about the desirability of outcomes and one’s pre-theoretical judgments about the rightness of actions in light of each other until reflective equilibrium is reached.
The act-consequentialist (the person who holds that acts are right solely in virtue of the desirability of their outcomes) can’t adopt the Footian Procedure if she wants the act-consequentialist bi-conditional to yield informative results, but this doesn’t mean that she must adopt the Foundationalist Procedure, for the Coherentist Procedure is a viable alternative that can yield informative results. Moreover, the Coherentist Procedure has the advantage of being likely to yield a substantive normative view (one that combines consequentialism with a substantive account of how outcomes are to be ordered) that avoids many of the counter-intuitive implications that plagues act-utilitarianism.
I posted my last comment not having seen either Ben’s or Jamie’s latest comments. I didn’t mean to endorse the suggestion that I was distinguishing between three procedures for deciding what to do. I mean to be distinguishing between three procedures for determining how outcomes are to be ranked, as should be clear from my previous comment.
Thanks Doug, that helps. I agree that Procedure 3 seems most reasonable.
I didn’t think my one-liner charicatures would be accepted by anyone, consequentialist or not. I was just trying to make a simple point that Doug’s alternatives didn’t seem puzzling thought of as procedures for ranking (at some level–it doesn’t have to be a theory of first-order practical reasoning), and that is one way consequentialism has been distinguished from deontology. In any case, the pull of consequentialism for me requires viewing it as such a procedure …because it is puzzling how bringing about more good rather than less could be the wrong thing to do.
This comment is off to the side of the main debate, but I don’t think Foot is a Footian consequentialist (since I don’t think she is a consequentialist at all) at least in the paper on virtue and consequentialism.
What I take her view to be is that rankings of states of affairs for better or worse are always made from the perspective of benevolence. And whenever two people rank the same states of affairs for better or worse they should agree about which is better and which is worse.
But, the other virtues besides benevolence constrain when we may rank states of affairs. I can’t rank a state of affairs as better than another if I have to commit an injustice to bring it about. But another person might be able to rank that same state of affairs since you would not need to do an injustice to bring it about. (I think ideas about the unity of virtue play a role in her thinking this.)
Thus each agent will have an ordering of states of affairs from better to worse, where the ordering will be determined by benevolence, but not every state of affairs will be ranked. Which ones will be determined by what the other virtues permit. And since the other virtues will involve constraints on doing certain actions, these will produce unranked states of affairs that differ for each agent, depending on whether those actions are actions of that agent or the actions of another. (So you get a gappy utility function from outcomes to values, where some outcomes get no value.)
The right action when an available outcome is unranked will not always be the action which brings about the best available outcome. Sometimes the right action is neither better nor worse (from the point of view of benevolence) than the one that is forbidden (since such comparisons would require both to be ranked). So her view is not consequentialist, at least as she presents it.
At least this is my reading of her view.
Mark,
You say, “Sometimes the right action is neither better nor worse (from the point of view of benevolence) than the one that is forbidden (since such comparisons would require both to be ranked). So her view is not consequentialist, at least as she presents it.” It sounds to me like she accepts the act-consequentialist’s bi-conditional: an act is permissible iff no other available outcome ranks higher than its outcome. For instance, it seems that it will be permissible to refrain from maximizing aggregate utility iff the outcome where one fails to maximize aggregate utility is not outranked by the outcome where one maximizes aggregate utility, as where maximizing aggregate utility would involve the commission of an injustice (or some other virtue infringement). Note, of course, that if we can’t rank O1 vis-à-vis O2, then O1 is not outranked by O2. Nevertheless, I admit that Foot is a non-consequentialist. What I earlier and misleadingly called “empty consequentialism” and now call “the Footian Procedure for determining the ordering of outcomes” is not a procedure that the act-consequentialist (defined here as the person who holds that acts are right solely in virtue of the desirability of their outcomes) can adopt. Thus the fact that Foot believes that we can only determine how outcomes are to be ordered by employing something like the Footian Procedure is what makes her a kind of non-consequentialist.
In any case, I wonder whether, in light of the viability of the Coherentist Procedure, you still think that the consequentializing strategy for defending consequentialism against non-consequentialism results in a kind of consequentialism that provides no normative advice. Thoughts?
Robert,
I didn’t think my one-liner charicatures would be accepted by anyone, consequentialist or not.
Ok.
I was just surprise that you thought (or seemed to think) philosophers since Rawls have often understood consequentialism as a deliberative procedure.
In any case, the pull of consequentialism for me requires viewing it as such a procedure …because it is puzzling how bringing about more good rather than less could be the wrong thing to do.
I don’t get it. Why would it not be so puzzling how bringing about more good could be wrong if consequentialism weren’t a deliberative procedure but (say) an account of what the right-makers are?
Doug,
Two comments, the first on Foot, the second on the main issue.
On Foot, you write, “It sounds to me like she accepts the act-consequentialist’s bi-conditional: an act is permissible iff no other available outcome ranks higher than its outcome.”
That’s right, though she rejects the idea that an act is permissible if and only if it is in the set of most highly ranked available actions. Since I’m not sure that one or the other conditional is more constitutive of consequentialism than the other, I tend to think consequentialists should be thought of as accepting both. Otherwise, someone who thought you can never rank outcomes would count as a consequentialist.
On the issue of whether the consequentializing strategy must give no advice, I’m going to ramble through what I’ve been mulling over in response to the whole thread and see if that helps answer the question.
My unhappiness with the consequentializing strategy of incorporating what seem to be non-consequentialist views is that it hides the non-consequentialists’s reasons for wanting to reject consequentialism of the traditional sort. It isn’t that even “empty” consequentialism doesn’t give advice. Each theory (whether empty, old-fashioned, or coherentist) when filled out with an ordering of outcomes gives very particular advice. But since I think different way of capturing that advice can more clearly display the rationale for acting one way than another, I think there is a reason to choose more traditional ways of characterizing non-consequentialist views, over the agent/time-relative consequentialized versions. I think that the consequentialized version doesn’t always allow the nonconsequentialist to present the reasons for certain constraints in such a way that someone can see their intuitive pull.
I actually think it is related to the idea which (as I understand you) leads you to think of your division into empty, foundationalist and coherentist consequentialism, yields three kinds of consequentialism, rather than just being three different moral epistemologies for theory choice. I think you think that a moral theorist who who adopts each method will wind up with a different theory insofar as each method yields a different ordering of outcomes.
Roughly, as I understand you, you think that taking intuitions just about what is better than what will give us one ordering — perhaps something like traditional utilitarianism, or G. E. Moore’s view. Taking intuitions just about what is right and constructing an ordering to capture them will lead to an ordering of outcomes that is agent and time relative and mimics Rossian or Foot-like views about what is right. And you think that a theorist who balances these intuitions against one another will wind up with a distinct view. I guess the idea is that if we start with an intuition that fewer deaths are better than more deaths and the intuition that killing one person to save several is wrong, we now have to figure out how to construct an ordering we’ll be happy with and that gives each intuition its due. And, when we do so we get an ordering of outcomes that is distinct from the orderings generated by the other two methods.
My main problem with this is that an intuitive level these two intuitions don’t seem to me to compete unless I’m already committed to capturing the second one in my ordering of outcomes from better to worse. So when I try to put everything in reflective equilibrium, I hang onto the thought that fewer deaths are better than more, that killing one to save several is wrong, and give up on the idea that right actions must lead to a better outcome. On the other hand, I feel the pull of the idea that as outcomes get worse we have more reason to avoid them, so I tend to think constraints that give way to avoid extreme badness are the way to go.
Thus, I can see that one might think that the coherentist way of generating an ordering would generate a different ordering than the other two approaches. But when I try to get coherence among my own intuitions I give up a commitment to the idea that right actions make things better. Or (and this comes to the same thing really) I think that the sense in which it is true that right actions make things better is not the same sense of better as the one in which fewer deaths are better than more. In other words, I can read the former judgement as employing an agent-relative notion of better, and using that notion I won’t deny it. And I can read the latter as employing the agent-neutral sense of better that I have no trouble with. Thinking of it this way, there are not really two conflicting intuitions to be put into equilibrium.
I’m not sure this is a good answer to your question, but I think it might be.
Mark,
Thanks for the detailed reply. First, regarding consequentialism and the ability to rank outcomes, I don’t think that the act-consequentialist is committed to completeness. Indeed, I think that even someone who held that you could never rank outcomes could count as an act-consequentialist provided that she held that all acts were permissible precisely because one outcome never outranks another. Nevertheless, we both agree that Foot is a non-consequentialist.
Second,regarding the other matter, I think that we might now be talking past one another. My post concerns how best to arrive at a substantive form of act-consequentialism, whereas you seem more concerned to argue that when we employ wide reflective equilibrium, we’ll find that there are reasons to think that we’ll arrive at non-consequentialism as opposed to consequentialism—that, for instance, we (or at least non-consequentialists) are more likely to give up the intuition that it is never wrong to bring about the best outcome (if we/they have it) than to give up the intuition that, other things being equal, an outcome with more deaths is worse than one with fewer deaths. I, of course, think that this is the more important issue. But before we can ask which theory does better on wide reflective equilibrium we need to have a particular substantive version of consequentialism in hand. My thesis is that the one that stands the best chance on wide reflective equilibrium is the one we arrive at by employing the Coherentist Procedure.
Now besides your important arguments regarding the relative merits of agent/time-relative consequentialism and non-consequentialism, I thought that you also wanted to claim that if we individuate options in a sufficiently fine-grained way as the agent/time-relative consequentialist would have us do so, then any choice can be described as bringing about the best outcome and that, in that case, “the theory no longer gives us any normative advice.” This is what I see as relevant to the issue about how to arrive at the most plausible substantive version of consequentialism. We certainly don’t want to arrive at a version that provides no new information (or normative advice) beyond what pre-theoretical moral judgments we already have, as is the case with empty consequentialism. But it now seems that you agree that agent/time-relative consequentialism (if thought of as a product of the Coherentist Procedure rather than the Footian Procedure) will offer plenty of new and interesting normative advice. It’s just that you think that something will get lost in the process.
Jamie, for me the intuitive pull of consequentialism comes entirely from the variety of imaginary cases that crowd the literature (“You are in a boat and can save x number here or y number there, but not both…’). Those thought experiments seem to ask me to rank my alternatives, and in those cases, I find it very difficult to see how I can do that in any way other than some form of consequentialism tells me to. I don’t find anything like that intuitive appeal when the question concerns the right-making properties of actions. That said, it may just be an autobiographical point.
Doug,
I don’t think we are talking past each other. I was trying to get across that I see why you think the reflective equilibrium process might yield a unique version of consequentialism. But whether it does or not will depend, it seems to me, on the actual process of reasoning one would go through when one tries to reach such equilibrium. I just suggested that when I myself go through the process, I wind up jettisoning the consequentialist bit, so when I do it I don’t wind up with a different version of consequentialism. I think that the next move for the consequentialist who likes this strategy is to say a bit about how other ways of resolving apparent tensions between intuitively attractive ideas would lead to a consequentialist view. Of course presenting such a picture might change what my reflective equilibrium would look like too. (Or if not mine, someone less stubborn.)
On the providing normative advice point: I guess I think that in principle, if the ordering is completely filled out to construct a complete theory, it would provide guidance. But in real life, to the extent that I accept the agent/time relativized version of consequentialism I accept that there is such an ordering that captures which actions are right and which actions are wrong. But I can’t consult that thought to figure out what to do unless I have a pretty determinate idea of which ordering that is. But I don’t yet. When I try to figure that ordering out I go back to thinking of betterness and worseness in an agent-neutral way and then think about what role that goodness should have in my decision-making in a life where I have other obligations besides maximizing overall agent-neutral good. So in real life the relativized teleological/consequentialist way of thinking does not give me guidance.
If your reflective equilibrium method yields a pretty determinate theory, then “no guidance” objection isn’t so strong anymore, though I suppose we might still argue about whether this was the best way to present the view. I think I think that the issue here is not about the truth of the view, since I take these to be alternate presentations of the same view, at least about which actions are right. I think, however, that the non-teleological presentation of non-consequentialist views (to talk my preferred way) is most apt to suggest why these are the right actions.
Mark,
I’m confused. How do you jettison the consequentialist bit and wind up with any version of consequentialism at all? If you jettison the consequentialist bit, you wind up with nonconsequentialism. What am I missing? In any case, on the Coherentist Procedure as I’ve stated it, you must keep the act-consequentialist’s bi-conditional constant and revise your pre-theoretical deontic judgments and your pre-theoretical evaluative judgments until reflective equilibrium is reached. Otherwise, as I just noted, you wouldn’t necessarily wind up with a version of consequentialism, let alone the most plausible version. So you’re not permitted to jettison the consequentialist bit it at this stage of using narrow reflective equilibrium to determine the substance of consequentialism.
I agree that the next step after that would be “to say a bit about how other ways of resolving apparent tensions between intuitively attractive ideas would lead to a consequentialist view.” Perhaps, the consequentialist can appeal to maximizing rationality or the teleological conception of reasons or what have you. But that seems to be wide reflective equilibrium, where the consequentialist principle is not itself privileged and is open to rejection. In deciding what the most plausible version of consequentialism is, though, we can’t allow the consequentialist principle to be jettisoned.
You say, “I accept that there is such an ordering that captures which actions are right and which actions are wrong. But I can’t consult that thought to figure out what to do unless I have a pretty determinate idea of which ordering that is. But I don’t yet.” Isn’t that because you haven’t yet gone through the Coherentist Procedure and reached reflective equilibrium? If you had, wouldn’t you know exactly what that ordering was? Once you go through the Coherentist Procedure you do get a determinate ordering that yield new normative advice. It’s only if you stick with the Footian Procedure that you’ll fail to get an ordering that provide new information.
You also say, “I think, however, that the non-teleological presentation of non-consequentialist views (to talk my preferred way) is most apt to suggest why these are the right actions.” To determine whether you’re right or not, there will need to be a lot of further debate, debate about whether the rationale for constraints is most plausibly construed as being agent-centered, as I would suggest, or victim-centered as Kamm and Brook have argued, debate about whether the teleological conception of reasons is correct, debate about whether it is puzzling to presume that could be impermissible to bring about a better world, etc. But that’s my whole research project: to argue that wide reflective equilibrium yields a relativized teleological/consequentialist view. All I wanted to do in this meager post was to suggest that Coherentist Procedure was the best procedure for arriving at a substantive consequentialist view and that the resulting view does yield new normative advice.
Doug,
Two quick replies to two things you wrote.
You wrote:
I’m confused. How do you jettison the consequentialist bit and wind up with any version of consequentialism at all? If you jettison the consequentialist bit, you wind up with nonconsequentialism. What am I missing?
Perhaps that I’m not disagreeing with you as much as you think.
What I meant to be saying is that when I try to put all the prima facie plausible claims together, I wind up rejecting consequentialism. But I’m a non-consequentialist, so that is no surprise. I was saying that I’m not a good test case for what a committed consequentialist applying your methodology would come up with. The real test is what a committed teleologist comes up with using this methodology. And I take it that is your long-term project.
And you wrote:
Isn’t that because you haven’t yet gone through the Coherentist Procedure and reached reflective equilibrium? If you had, wouldn’t you know exactly what that ordering was?
That’s right. I was trying to explain the sense in which the ‘no normative advice” point might still be correct given my admission that that the relativized teleological strategy does generate advice when filled in. And my thought was that in real life I don’t expect to ever get a completely worked out theory, so I won’t ever actually accept a complete filling in of the ordering. In that situation, it might be that non-teleological specifications of what to do (and why) will serve me better in figuring out how to deal with as yet unconsidered cases than an only partially filled in ordering.
And I agree that the debate about the rationales for constraints is relevant to figuring out if I’m right about the best way to construct a theory that explains why certain actions are right.
So while we disagree about our favored positions, I don’t see that much disagreement about what is relevant to the argument between them.