Avoiding Consequentialism’s Demands

Consequentialism, many philosophers have claimed, asks too much of us to be a plausible ethical theory. Indeed, consequentialism’s severe demandingness is often claimed to be its chief flaw. I will try to show that consequentialism’s demandingness cannot be the theory’s downfall. I do not here aim to vindicate consequentialism or demonstrate that it is not too demanding. What I think I can show is that the demandingness of consequentialism cannot plausibly be the decisive objection to the view. This is the case because the demandingness objection, in any plausible form, requires that we already reject consequentialism before it can be persuasive. Unless we presuppose an ethically significant distinction between, for example, causing and allowing, or intending and foreseeing, any moral theory that is less demanding than consequentialism will permit us to cause or intend too much to be at all plausible. Thus, key components of consequentialism must already be assumed or argued to be false before the demandingness objection can get a grip. Thus the demandingness objection should not be what persuades us that consequentialism is false. The arguments I use to show this are not new, I borrow them from Shelly Kagan and others,  but they are underappreciated and have not been made in the most useful and general way.

Recall how Samuel Scheffler tried to amend consequentialism to accommodate the demandingness objection. If the problem is that one’s own interests are not permitted to be given enough weight, he in effect reasoned, then we should simply permit one’s interests to be given more weight. Thus, on his view, one could multiply one’s own interests by a certain number and be morally permitted to maximize the new weighted aggregate. This, or something much like it, is the most obvious response to the thought that consequentialism requires too much of us and that it does not make enough room for our projects and interests. (I here ignore some interesting technicalities of Scheffler’s approach that do not help him here.)

Shelly Kagan objected to this that it permits one to cause harm for the sake of one’s magnified interests as much as it permits one to allow harm for the sake of one’s magnified interests. One would have the permission to cause harm to some innocent person merely because the harm was less than N times as bad as the benefits to oneself of causing the harm. This sort of permission, Kagan rightly thought, is not sanctioned by commonsense intuitions about morality. The demandingness objection only comes into its own when we think of morality as not forbidding us to allow (or merely foresee) a certain harm. (Scheffler replied to this complaint and made interesting points in the process but I think could not rebut Kagan’s general claim.)

Bernard Williams argued that consequentialism threatens our integrity because it requires that we step aside from our most central personal projects merely because the interests of others outweigh one’s own interests. But again such a complaint only finds resonance when we presuppose that our projects require only that we not aid others. When we think of projects that require (for instrumental or intrinsic reasons) that we cause or intend harming other people for no better reason than that our outweighed personal project be promoted, then the integrity complaint against consequentialism is unpersuasive.

I do not know of a place where Kagan explicitly drew the moral from the failure of Scheffler’s approach that seems to me warranted. The moral is that the demandingness objection is implausible unless supplemented with a more nuanced account of when morality may ask quite a lot from us and when it seems to not be able to demand as much. It is simply false that morality, as commonsense understands it, may not demand quite a bit from us in relatively ordinary contexts. Most decent people do not much bristle at constraints such as the one to not bump off a rich uncle merely for personal gain, but in terms of the sacrifice of one’s interests such constraints are quite significant. To capture the intuitions that are pumped by the demandingness objection, together with the intuitions that one may not cause big harms merely to produce small benefits to oneself, the demandingness objection must be transformed into a new complaint. To vindicate this constellation of intuitions, one must vindicate the moral importance of distinctions such as the causing/allowing or intending/foreseeing distinction. And to do this is to switch the battleground between consequentialists and its foes away from the demandingness objection and to other ground. And only with a victory on this other ground could anything like a demandingness objection look attractive. Thus I think the demandingness objection presupposes that the decisive break with consequentialism has already been established before it looks attractive. Thus I think it should not be the reason people reject consequentialism. We should reject it on other grounds or not at all.

18 Replies to “Avoiding Consequentialism’s Demands

  1. David,
    I was going to quote the last seven or so sentences, but since they are the last ones, I’ll forgo the quotation, since it is right there above.
    I see the point that to vindicate the objection we would have to show two things: (1) that there is some sort of asymetry between causing/allowing or intending/forseeing or some similar pair. And (2) that with respect to what consequentialism requires us to do, it sometimes asks us to harm ourselves more than is reasonable to bring about more good for others.
    But I’m not sure that this makes it right to say that the objection presupposes the falsity of consequentialism insofar as consequentialism rejects the required asymetry. I guess that I think this because as first presented the objection rests on a sort of intuitive judgement about certain kinds of cases. But, since intuitive judgements of this sort are defeasible, pointing to such judgements (even if they are inconsistent with consequentialism) does not create circularity or beg any questions. It is the start of an argument, and the argument is not done until we consider various strategies for explaining away the judgement as mistaken, balancing it against what other intuitive judgements we would have to give up to vindicate it, balancing it against various theoretical virtues that a consequentialist theory might have that a theory that vindicates the intuitive judgement would lack, and all of the other things we do in trying to reach reflective equillibrium with respect to particular intuitive judgements and more abstract theoretical judgements.
    I suppose that if it wasn’t like this any really knock down objection would be circular, since any such objection would be incompatible with the view it is trying to refute.
    Myself I like other regarding versions of the demandingness objection better than the standard versions involving sacrifice to the self, but they would have the same need of an asymetry of the sort you suggest is needed to make the objection go.

  2. Mark,
    Tell me if I am getting your point. Perhaps you are saying that we notice that in some cases C seems too demanding. That is the start of a hunt for a good account of that intuition. The best account, you seem to agree, would not say that the problem with C is as simple as being too demanding (rather something like being too demanding with respect to what we may allow, or some such). But it was the initial intuition about demandingness that put on this path to understand our complaint against C and this is for, in some sense, the demandingness objection to live on. If that is the thought, I don’t disagree, I just think that what is vindicated should not be thought to be the demandingness objection. Perhaps I am forced to understand the demandingness objection narrowly, but I am willing to live with that. What we learn is that the demandingness objection alone, unaided without other objections to C, cannot survive reflective scrutiny. I don’t need the point to be any bigger than that.

  3. Dave: While your response to this version of the demandingness objection seems correct, I’d actually always thought that the objection took a different form (what I’ll now mention as simply another *version* of the objection), namely, that C is too demanding because its maxim, “Maximize the good,” requires application into every corner of our lives, covering every possible action (or omission). I don’t see that this version of the objection, then, would necessarily fall prey to your general responses, for one might say that *regardless* of whether or not there are distinctions between doing and allowing, say, this worry still obtains, for it’s about the lack of any non-C personal space.

  4. Dave,
    Good, that is very helpful, thanks for your comment.
    One could think of a complaint against C that says its problem is that there is no area free from the strenuous demand to maximize. If one wanted to accommodate this sort of worry, a natural way to do it would be to say that one need not maximize in what one allows, merely in what one causes, or some such. The thought here is that we need some sheltered areas from C’s demands, not that there being areas in which C is as demanding as it is is a problem.
    That strikes me as an interesting worry against C. Perhaps others have put it that way, I suppose people at least tacitly certainly have. But the sort of remedy that Scheffler and Williams seemed to want was a sweeping, across the board reduction in demandingness, rather than pockets of protection. Well Scheffler certainly wanted that, Williams’s rhetoric can also seem to tend in that direction, but perhaps that is an uncharitable reading.
    In any case, this is a useful distinction in kinds of demandingness complaints, the distinction between seeking some shaded areas and an across the board muffling of C’s demands. My point in the initial post was aimed against the latter.
    There are two different ways, it now seems to me, that one might press your “anti-pervasive” version of the demandingness objection in response to my initial post. First one might say that it was demandingness’s thought all along that C inappropriately was as demanding in some areas (e.g. in what it permits one to allow) as it is appropriately in other areas (e.g in what it permits one to cause). That is, this version of the worry would build into the demandingness objection an understanding of the contours of the moral landscape that marks distinctions that C ignores. On this version, the demandingness objection builds in a distinction between causing and allowing (or some such) as I said it needed to, but this is not an unwanted supplement but part and parcel of the objection all along.
    The second version, I take it your version, would say that the “anti-pervasive” version of the demandingness objection need not presuppose any contours of the moral landscape (such as a morally important distinction between causing and allowing) but need only hold that we deserve some areas of shade or other from C’s demands. It could be held that there are several different ways to provide this shade and none are privileged by antecedent contours of the moral landscape. This version of the demandingness objection does look to me purely a demandingness objection not a “demandingness plus some other component that is hostile to C” objection. Thus it is more of a worry to me.
    Such an version of demandingness would need to explain why permitting us room to cause harm (rather than room to allow harm) would not be an equally acceptable way of making C less pervasively demanding without presupposing a morally salient causing/allowing distinction. This strikes me as a worry about this version, but I still need to think about it more.

  5. David (Sobel)-
    I ditto David (Shoemaker)’s remarks: I’ve thought that the force of the demandingness objection comes not from showing that in case C, consequentialism is too demanding, and so consequentialism is too demanding. Rather, cases that are adduced to support the demandingness objection are meant to be instances of the more general worry that, as David put it, consequentialism denies that there is any of my personal normative space that is off limits to morality’s demands. I.e., I understand the cases that, e.g, Williams cites in his integrity paper, to function as illustrations of the general thesis, not evidence for it. Now, if the consequentialist can answer these cases and show that (individually) the cases fail to show that consequentialism is too demanding (perhaps because the intuitions being pumped in connection with each case do depend on doing/allowing, intending/foreseeing,etc.), then that’s a pretty strong defense for the consequentialist. But the demandingness objection, understood as a general thesis, doesn’t seem to me to depend on doing/allowing, intending/foreseeing, etc.
    Also, in the spirit of Flanagan’s “minimal psychological realism”, what’s your favored response to the empirical version of the demandingness objection: that human nature simply doesn’t exhibit the kind of altruism or impartiality that consequentialism demands? The empirical evidence does not look good for consequentialists, and I tend to think this is the more potent version of the objection. Would your response involve a subjective/objective consequentialism distinction, differentiating between criteria of rightness and decision procedures?

  6. David,
    I think you pretty well appreciate what I was trying to say.
    One way that perhaps we differ is about what counts as a demandingness objection. I’m inclined to say that if intutitions about some examples that consequentialism demands too much of someone in that case motivate the objection it can count as a demandingness objection, even if there may be other elements in what we wind up believing as we come to reflective equillibrium.
    I’m not sure that at the end of the process of seeking equillibrium I won’t just think that my initial intuitive judgement – say that asking a child to be killed by me for the greater overall good is forbidden – is one I won’t give up because I find the intuition so plausible that I cannot conclude that it is incorrect. And I guess that if that is why I find the objection persuasive (and not because I can come up with some additional independent reason to defend that conclusion) it is thoughts about demandingness that are doing the work. I do think that I will have to incorporate some asymetry story of the sort you suggest as well, but that too might be motivated by finding these sorts of intuitions about cases of this sort plausible. So I’m inclined to think this is still motivated by a worry about demandingness.
    (If this thread continues and I’m slow about replying to something directed to me, I’ll have only intermitant internet connectivity for a couple of weeks.)

  7. David,
    I’m pretty interested in this objection. The way I understand people, there really is no unique “overdemandingness” objection that people have in mind when they say Consequentialism is too demanding. They tend to mean either that (1) Generally, sometimes people are permitted to promote their own well-being over the greater good, (2) In a particular case, say, Jimmy didn’t do anything wrong when he sat in the hottub all night and he wasn’t promoting the greater good, (3) Generally, sometimes people are permitted to promote their loved-ones well-being over the greater good, (4) In a particular case, say, Jimmy didn’t do anything wrong when he bought his girlfriend a hottub to sit in all night and he wasn’t promoting the greater good in so buying, (5) Lastly, there are inaucuous actions, actions that are morally empty. These are actions like scratching one’s toes.
    Anyway, these seem to get dumped together and they are all pretty different. I don’t get though how an assertion of one of these, plus the conditional that, say, if (1) is true, then Consequentialism is false, assumes, in any important way, some account of the doing-allowing distinction.
    Michael,
    Doesn’t the “ought implies can” principle handle this worry: “that human nature simply doesn’t exhibit the kind of altruism or impartiality that consequentialism demands.” That is, for those who can’t because they don’t have the right nature, it’s false that they ought to.
    And I’m wondering, what’s wrong with taking a certain moral from instances of the demandingness objection like this: These objections show us that there are constraints on the correct value theory, constraints whose truth is entailed by the correct moral theory.

  8. Christian,
    Well, yes, “ought implies can” handles it, but that’s exactly the point of the objection. The usual worry is that the consequentialist can’t handle it, to wit, that consequentialism requires that agents exercise a degree of altruism or impartiality that they are in fact unable to exercise. Hence,’ought implies can’ plus ‘agents ought to exhibit an impossible degree of altruism or impartiality’ yields that agents both ought and ought not be consequentialists.

  9. There is another way consequentialists can answer the demandingness objection[s]. If my memory serves me right, this is something Mill already thought of. One could see the demandingness objection as providing pressure to move from act-consequentialism to indirect consequentialism or rule-consequentialism. It may be that, on consequentualist criteria, we succeed in producing most well-being by pretty much acting as we do as moderetely both self-interested and altruistic. In the long run, if we all acted all the time on some more demanding altruistic requirements to further the general well-being, we could all loose our individual lives and integrity as Williams put it. It is hard to see how a world like that would have more well-being than ours. Like Mill put it, few of us can do much on the global scale, but we can do much locally to improve general well-being by starting from our own. Thus, consequentialism should not prescribe us to pursue well-being by the demanding route of each action directed to general well-being but indirectly through our ordinary, non-demanding actions.
    Rule-consequentialists can say even more. There is an interesting passage about this in Hooker’s Ideal Code, Real World on page 98. Hooker discusses the objection that rule-consequentialism’s rules collapse into act-consequentialism through all the exception clauses. That would be bad because then rule-consequentialism becomes also threatened by the demandingness objection which Hooker believes is a serious problem. But, Hooker points out that as in the cost-benefit comparisons of the rules one needs to also count the internalisation costs, the rules that would have collapsed into demanding act-consequentialism would be too costly to be get generally internalised – ‘if people were motivated to comply with these rules, they would be motivated regularly to make enormous self-sacrifices. The cost of getting such motivation internalised would be too great’. Therefore, less demanding code may save more in internalisation costs that is lost in the occasions where that code leads to less well-being conducive actions, and thus the less demanding code would be more ideal than the demanding.
    If the empirical assumptions behind this argument are true, then this sounds like a good consequentialist argument for consequentialism not being demanding.

  10. Christian,
    “I don’t get though how an assertion of one of these, plus the conditional that, say, if (1) is true, then Consequentialism is false, assumes, in any important way, some account of the doing-allowing distinction.”
    Well, it could, couldn’t it? Maybe the reason for thinking that people are sometimes permitted to promote their own well-being over the greater good is that when they do this they don’t do anything to harm anyone; they merely allow some harm to happen to distant others because they’re preoccupied with their own well-being.
    But even if that’s not the reason, I thought David intended his point to be broader. It’s that the demandingness objection relies on other objections to consequentialism, whether the objection is based in its failure of it to recognize the moral importance of causing vs. allowing, intending vs. foreseeing, OR some such objection, say, an objection to C’s requirement that we be impartial about whose good is being promoted. Doesn’t the assertion of (1), plus the conditional that if (1) is true, then C is false, already assume that C’s impartiality requirement is false?

  11. Sorry, I am falling behind because I am moving offices.
    Michael,
    I guess I don’t see how C could violate ought implies can. Surely C only asks us to do what we can do. If we really literally could not do something, then C would not ask it of us, just as it does not ask us to fly even when lots of good would result if we did fly.

  12. David,
    No no, I understand. Consequentialists shouldn’t (in my view) seek to jettison OIC. The challenge is this. Once we determine just how robust a sense of impartiality or altruism is consistent with human nature, the worry is that consequentialists have tended to favor, in comparison to their theoretical rivals at least, conceptions of morality’s demands that ask more of agents in terms of impartiality or altruism. So suppose that subjective act-consequentialism is the most demanding form of consequentialism in this regard. If the demands that it makes on agents in terms of impartiality or altruism are greater than human nature permits, then by OIC, subjective act-consequentalism should be rejected. Then consequentialists are compelled, ceteris paribus, to move to the next most demanding form of consequentialism to determine if it is consistent with human limits on impartiality and altruism, etc. The dynamic that results suggests that consequentialists will have to scale back their moral requirements if those requirements ask agents to exhibit impossible degrees of impartiality or altruism, and perhaps in so doing, the resulting view is not recognizably consequentialist.
    (Incidentally, I assumed that this whole discussion is predicated on maximizing consequentialism. The satisficing version may be less subject to this objection.)

  13. Michael,
    I would say that maximizing C is the view that one morally must create the largest amount of good that one can. OIC is built into such a proposal. If we are capable of less than some have hoped, then C automatically becomes less demanding, and not from external pressure, but for reasons stemming from the C thesis itself. Thus I don’t see how, if we have fewer options than some have hoped, this could threaten C. Indeed if psychological egoism is true, we will trivially live up to C.

  14. Michael,
    I suppose I ditto David Sobel. I think OIC is built into Maximizing C in that it is a constraint on the truth of any moral theory of rightness. Michael says, “the worry is that consequentialists have tended to favor, in comparison to their theoretical rivals at least, conceptions of morality’s demands that ask more of agents in terms of impartiality or altruism.” And I think something like this is true, C demands alot, but only that when we are able to promote the best, that’s the right thing to do. But I don’t think C demands more by way of what dispositions (altruism) exist than any other theory.
    Kyle,
    I think an objector might accept a doing/allowing distinction and argue against C by way of the demandingness objection. They might even accept that one is not permitted to cause harm to others for the overall good, but they are permitted to allow harm for the greater overall good. Sure. I just don’t think they have to by virtue of asserting the demandingness objection. That is, the doing/allowing distinction is not a precondition of rationality for someone asserting the demandingness objection. It is not built into any of (1) through (5) I listed above, nor is it forced on anyone asserting (1) through (5) by virtue of this acceptance and deductions from common background knowledge the objector should share with her opponent. That’s all I mean really. And I really don’t think C requires impartiality, strictly speaking, not if our loved ones can sometimes be the ones that need our attention the most.
    Jussi,
    I don’t think I understand why C’s being demanding, supposing the demandingness objection is okay, gives us a good reason to be rule-consequentialist. How is rule-consequentialism less demanding? I also got to say that I am pretty convinced that rule-consequentialism just reduces to, in some important sense of ‘reduces to’, to C discussed above.

  15. David,
    I think I’m with Mark on this one.
    In response to Mark you say: “What we learn is that the demandingness objection alone, unaided without other objections to C, cannot survive reflective scrutiny.”
    Consider the following fantasy. We are all worried about cases where C seems to demand too much of us. We can’t get into reflective equilibrium. Someone comes along with a new theory X that seems to get the right results in those cases. Maybe X incorporates some sort of doing/allowing distinction that enables it to give those results, but it doesn’t matter – it will obviously have some feature or other that is incompatible with C. We all embrace X and reject C, and live happily ever after in reflective equilibrium.
    Doesn’t it seem fair to describe what happened as C getting refuted by a demandingness objection? It’s not the demandingness objection *aided by another objection* – it’s the demandingness objection *supported by X*. Of course, X might also support other sorts of objections to C, but that’s not relevant to whether demandingness is a problem on its own. (I think I might just be reiterating Mark’s point.) Enjoy your new office!

  16. Actually Ben adds in an extra component to what I said that I should have made explicit but neglected to mention.
    Much of our thinking about whether a theory is adequate is sensitive to the other candidates available. Whether a theory is refuted by an objection might well depend on whether people are aware of alternative theories to which the objection does not apply, or does not apply as strongly. So one way of taking David’s original point without granting all of it is to say that it is only when theories are available which can make use of doing/allowing, acting/refraining distinctions or the like before the demandingness objection makes it an attractive option to abandon consequentialism.
    I think Ben’s comment does a nice job of making this point.

  17. Mark and Ben,
    I am not sure I am following you guys. It seems to me that we all agree that Scheffler-style attempts to remedy demandingness fail because they do not add in a doing/allowing distinction (or some such). But you guys want to say that, I take it, despite this, the demandingness itself, not the demandingness plus some other independent complaint, can still serve as a self-standing good complaint against C. That is what I still don’t see.
    Ben says, “Maybe X incorporates some sort of doing/allowing distinction that enables it to give those results, but it doesn’t matter – it will obviously have some feature or other that is incompatible with C. We all embrace X and reject C, and live happily ever after in reflective equilibrium. Doesn’t it seem fair to describe what happened as C getting refuted by a demandingness objection?”
    I don’t see why we should think this. In answering the question of what the telling objection to C is, does it not matter what relation X has to the original objection. Surely X occurring to us has to be more than merely causally the upshot of having the original worry for it to be the case that the vindication of X is a vindication of the original objection.
    Suppose generally, one has a complaint A against a theory and all by itself (unsupplemented with non-A complaints against C) A does not offer a plausible alternative to C. But there is another complaint against C, call it B, and A plus B together make a plausible package alternative to C. Finally, and this I am adding without previous support, suppose B did look tempting all on its own (without A) and could generate many of the intuitions we want to capture even without A. Why should we then think of the plausibility of the package of A and B as a vindication of the A objection alone? Wouldn’t it be much more tempting to think that A is a bit of a free rider in the plausible alternative to C?
    As I was thinking of it, demandingness is a distinct kind of complaint, different from other complaints against C. I was wondering if that distinct complaint counted as a good objection to C. My thesis is that that distinct complaint is only plausible after another independent and distinct complaint against C is made plausible, the causing/allowing distinction (or some such). This it seems to me is the lesson of the failure of Schaffer’s approach. Merely making C less demanding, all by itself, does not yield a plausible theory. We must make it less demanding in a way that presupposes an important moral distinction between causing and allowing (or some such) to make a plausible view.
    Now perhaps, I am not sure, you guys are saying that the good point in the demandingness objection helps us see the attractions of the distinction between causing and allowing and so focusing on demandingness is helpful in seeing a plausible alternative to C. I don’t disagree with that. I just think that the demandingness of C is not, by itself, a good objection to C, even if thinking about demandingness helps us see what is really a good objection to C.

  18. David,
    I think I do think something like “that the good point in the demandingness objection helps us see the attractions of the distinction between causing and allowing and so focusing on demandingness is helpful in seeing a plausible alternative to C.” (But I have to agree with earlier folks who claimed that there might be several demandingness objections.)
    A person might take three attitudes toward C:
    (1) Accepting C.
    (2) Agnosticism towards C.
    (3) Rejecting C.
    I take it that one could adopt either of (2) or (3) because one thought that in certain kinds of cases C demands too much of someone, and that one could take those attitudes without being committed to any particular alternative theory of morality.
    But now suppose that one then casts around for an alternative to C, and notices that some of these alternative make a distinction between doing and allowing (or something of the sort) and that this theory is therefore in a position to say what about the example in question makes it too demanding, and also is consistent with intuitive judgements one accepts about some other cases. Let this theory be X.
    I think that having noticed that X is an alternative that treats the example more in line with what one finds intutitively plausible, one might now have more reason to reject C than before. And the original intutive judgements about demandingness can give one reason to adopt X rather than C. So I think it is true that the availability of the doing/allowing distinction can play a role in one’s reasons for rejecting C.
    But I don’t think this shows that this makes the demandingness objection circular, or that it presupposes some such distinction. First off, one could have rejected C while remaining unsure of which alternative to it was correct, just because one finds the intuitions about particular cases used in pushing the objection more plausible than C. And secondly, even if it is the availability of X that leads one to reject C, that doesn’t show that the objection presupposes X or its component theoretical claims. Usually, philosophical theorizing is a matter of inference to the best explanation with each theory doing better on some measures and worse on others. Often we won’t reject a particular theory without an alternative to it. That does not make all objections to philosophical theories circular, even if we won’t reject a theory on the basis of an objection without some alternative theory not subject to that objection being available.
    One of the truisms (which may be false) about the history of astronomy is that the addition of epicycles to the existing theories continued in the face of evidence that the theory did not meet the observations until a rival system was available, at which point the original theory was rejected and the new one adopted. This would not show that recalcitrant observations were not evidence against the theory, nor that objections to the theory based on such observations presupposed the alternative theory, even if people would not reject the original theory until the alternative was available.
    I think that the same should be said here.
    Hope this helps make clear my view (and maybe Ben’s).

Comments are closed.