1. I would like to return today to some questions that have always puzzled me about the overall shape of Sidgwick’s ethics.
Did he think the principles of common-sense morality are self-evident? If not, what did he think was their epistemological status? Did he think that both egoism and impartialism are self-evident? In what sense if any did he think them contradictory?
These are not just questions for Sidgwick; they touch fundamental questions of moral philosophy. Convincing answers to them cannot be got from Sidgwick’s texts; there are too many tensions in his thinking to allow for that. I think he knew of this unclarity and uncertainty in some respects, notably in connection with the famous ‘dualism of the practical reason,’ but perhaps had not noticed it in others, notably in connection with the epistemology of common-sense morality.
This is not meant in a spirit of negative criticism. Far from it. Moral philosophers who try to take into account all the varied aspects of our ethical thought, rather than trying to reduce it implausibly to some single strand – philosophers such as Butler, or in a very different and bigger way Hegel – have always seemed to me to be the most interesting (they also tend to have an understandable philosophical interest in the history and anthropology of ethics). Sidgwick certainly belongs in this broad-minded and historically aware group. Furthermore, the obscurity that always seems to attend attempts at many-sidedness in moral philosophy may lie more in our ethical thought than in the attempts. If Sidgwick never succeeds in providing an overall picture that hangs together, is that because what is pictured does not hang together?
This is a question Sidgwick himself worried about throughout his philosophical life. It counts immensely in his favour that he never tries to force the issues in response to it, or to supply quick fixes where real tensions remain. But his eye is always on the mountain heights. He does not scale down his ambitions. Though his workmanship is invariably solid, he does not restrict himself only to those questions to which solid workmanship by itself can provide (or pretend to provide) an answer – lower hills that solid technique can hope to conquer. He is often said to be the first of the ‘professional’ or ‘academic’ moral philosophers. Obviously there is much truth in that; yet it is also somewhat misleading. Sidgwick shared with some other ambitious thinkers of the 19th century an intense preoccupation, even obsession, with the grand existential questions of ethics – they were as important to him as they were to Schopenhauer or Nietzsche, say. He thought of them as ultimate questions of philosophy, not as private concerns to be divorced from serious philosophy. Alas, his self-consciously.“scientific” mode of writing, which he took to such masochistic lengths, and which is one of his more dubious legacies, sadly obscures this. Along with his inordinate caution it may be the main reason why, as Bart Schultz put it in his lecture, he «is likely to remain, in large measure, a philosopher’s philosopher and a graduate student’s philosopher». It may be likely, but I hope it doesn’t happen. We collectively lose some self-understanding if a philosopher who expresses so accurately an unaffectedly the idealistic yet worldly, anxious and uncertain, ethical outlook of many intelligent modern people remains so little known.
2. Let’s consider Sidgwick’s account of what he calls the “intuitionist” method of ethics. It consists of investigating the intuitions contained in common-sense morality. Sidgwick explains that by an “intuition” he means an «immediate judgement as to what ought to be done or aimed at» (Henry Sidgwick, The Methods of Ethics, 7th ed., with a foreword by John Rawls, Indianapolis-Cambridge, Hackett Publish Company, 1981, p. 97). Intuitions appear as knowledge, and they are “immediate” in that they appear to deliver knowledge that is not derived from something else. Sidgwick notes that any method of ethics must ultimately rest on at least one such “intuition” – on a judgement as to what ought be done or aimed at which appears as evident in its own right; and which is, on that basis, regarded as «immediately known to be true» (In that sense, he notes, the egoistic and universalistic method are also “intuitional”). Nevertheless, when Sidgwick calls something an intuition he does not mean that it is knowledge. He thinks that intuitions «may turn out to have an element of error» (Henry Sidgwick, The Methods of Ethics, p. 211).
The intuitionist method seeks to identify our moral intuitions and to refine them from within, that is, from within reflective common sense. The intuitionist view, as against method, is the belief that these reflectively refined intuitions, or at least some of them or some part of them, do indeed constitute genuine knowledge in their own right, not derived from anything else. Many intuitionists would say that there simply isn’t anywhere else to derive them from.
Does Sidgwick deny that common sense delivers moral knowledge in its own right? I have said elsewhere that I find it very hard to be sure ( John Skorupski, Three Methods and a dualism, in Henry Sidgwick, edited by Ross Harrison, “Proceedings of the British Academy – 109”, Oxford, Oxford University Press, 2001, pp. 61-81). On one plausible interpretation he does deny it. He certainly asserts that “to an important extent… moral intuitions commonly so-called” do contain an element of error. On this interpretation, he goes quite a bit further: he thinks that no principle of ordinary morality is immediately known to be true; such principles, or rather their more precise correlates, can only be known by being derived from the impartialist principle which grounds utilitarianism and is immediately known. Here is the famous passage in which Sidgwick sets out this principle:
I obtain the self-evident principle that the good of any one individual is of no more importance, from the point of view (if I may so) of the Universe, than the good of any other … And it is evident to me that as a rational being I am bound to aim at good generally, – so far as it is attainably by my efforts – not merely at a particular part of it. From these two rational intuitions we may deduce, as a necessary inference … that each one is morally bound to regard the good of any other individual as much as his own... (Henry Sidgwick, The Methods of Ethics, p. 382).
So here Sidgwick derives a moral conclusion from two premises which he thinks are self-evident. The first is what I referred to as the impartialist principle: notice that it is an impartialist intuition about the good. Sidgwick’s second premise poses some problems of interpretation. It connects the good to what “I am bound to aim at.” And it seems clear, in context, that Sidgwick doesn’t just mean that the good impartially conceived is one of the things I am bound to aim at; he means that it is the one and only thing I am bound to aim at (cp “not merely at a particular part of it”). Furthermore he seems to take it for granted that ultimate truths about what I am bound to do are always about what I am bound to aim at. There are, for example, no ultimate truths about what I am bound not to do, irrespective of what I should aim at.
Notice, next, that the conclusion Sidgwick thinks we may deduce is something about what we are morally bound to do. We can assume that when Sidgwick talks about what we are morally bound to do he means what we have a moral obligation to do. So how does his deduction work? When he says that it is evident to him that as a rational being he is bound to aim at good generally, does he here also mean by ‘bound’ that he has a moral obligation to do that? If so, then his two rational intuitions do entail his conclusion. But now there is a difficulty. The second intuition on this reading is far more contentious than the first. Sidgwick slips it in
without apparently realising how contentious it is. Many people (including me) do indeed find it ‘self-evident’ that no-one’s good is more valuable than anyone else’s. Bernard Williams had some perceptive fun about “the point of view of the universe”. But it’s easy enough to eliminate this phrase without loss of meaning (if not without loss of background cultural assumptions). Many of us think the good of any one person is neither more nor less valuable than the good of another.We think this even if the good of some people is rightly more important to us, or more our responsibility, than the good of others. Would Williams have wanted to deny this? I doubt it (Bernard Williams, “‘The Point of View of the Universe’: Sidgwick and the Ambitions of Ethics”, in Cambridge Review, May 7, 1982).
That is Sidgwick’s first premise. In contrast, very few, when they reflect on it, find it ‘self-evident’ that they have a uniquemoral obligation to aim at the general good. They don’t find it obvious that the good is related to moral obligation in that extremely simple way. The conclusion Sidgwick draws from his two premises sounds act-utilitarian. I think that is what he meant by it. But as we by now all realise, act utilitarianism understood as a doctrine of moral obligation is definitely in sharp conflict with the morality of common sense. To remind you of some well-known points: common sense accepts the idea of supererogation. There are many cases in which it denies that I have a moral obligation to do the very best I could do, however admirable of me it would be to do it. (That is something John Stuart Mill liked to stress, and it is one reason why he cannot be cast as a moral act-utilitarian.) Another important point is that moral common- sense is not impartial. It holds that we have special, agentrelative obligations to people who stand in certain relationships to us – our children say, or our fellow-citizens. Ordinary morality holds that these are some of our weightiest moral obligations. And there are plenty of other familiar ways in which commonsense morality can come into conflict with act-utilitarianism.
From an act-utilitarian standpoint common-sense morality contains no immediate moral knowledge at all. The only moral truth is that one has a moral obligation to promote the good, impartially conceived, to the utmost degree; and that claim is one from which moral common sense clearly dissents. In fact, from an act-utilitarian standpoint the intuitionist method, at least if it is understood as a method of getting at moral truth, is a complete waste of time.
That of course is something a consistent act-utilitarian need not deny and may want to insist upon. The relevant point for us is that it is certainly not what Sidgwick insists upon. Instead, his eventual conclusion is that the rules of common sense morality can be in some sense explained as following from, or subordinate to, the impartialist conception of the good; perhaps because the moral sense is unconsciously impartialist, even though it is consciously partialist, or perhaps because the best psychological theories of workings of the “moral faculty” show it to be in some way causally responsive to empirical facts about what promotes general good (Henry Sidgwick, The Methods of Ethics, pp. 496-497, 481 ff.).
If these lines of thought are supposed to vindicate the intuitionist method, in the sense of showing that the intuitions it discovers in moral common sense do by and large count as knowledge, even if they contain unclarity and error, then they cannot be combined with act-utilitarianism. An act-utilitarian can at most see ordinary morality as providing useful heuristic rules for determining what one has a moral obligation to do, not as providing knowledge of what one ought to do. Even that is stretching it, since if these rules have evolved through some sort of responsiveness to what would promote general good if generally followed, there’s no reason to think they provide guidance as to what I in particular should do now.
There is of course another view: rule utilitarianism. This says that I am morally bound to follow rules whose general acceptance would produce the most good generally. Now that proposition is hardly an admissible interpretation of the second rational intuition, as Sidgwick formulates it in the passage I quoted. It would, however, make a great deal more sense of his broad-brush endorsement of common-sense morality, since it is much more plausible to argue that general compliance with common sense moral rules most effectively promotes the good than it is to argue that the individual maximiser of good should be bound by them.
Note however that on the rule-utilitarian view the obligations of ordinary morality, for example the obligation of honesty, or obligations of gratitude, or the obligation not to take people’s property without their consent, cannot be immediately known. The only obligation that can be immediately known is the rule-utilitarian obligation: to follow rules whose general acceptance would produce the most good generally. Therefore unless the rule-utilitarian argues implausibly that his own principle is contained in common-sense morality, he too has to deny that common-sense morality contains any immediate knowledge at all. But for him the intuitionist method need not be as much of a waste of time as it is for the act-utilitarian. For if it’s empirically plausible that our moral dispositions in some way track the facts about what would promote general good if generally followed, then the rule-utilitarian will think it’s useful to identify them accurately – even if they do not constitute immediate knowledge.
Either way, whether we make Sidgwick an act or a rule utilitarian (I am not suggesting these are the only possibilities, see below) it looks as though he should hold that moral common sense provides no immediate knowledge; if it provides knowledge at all, it does so only by an indirect route that goes well beyond its own terms. It is an indicator of moral truths rather than an intrinsically authoritative source of them. But that conclusion
is already inconsistent with moral common sense – for moral common sense holds that we can know our duty by immediate conscientious reflection on the facts as known to us. This doctrine of the authority of conscience, or of moral self-governance, is dear to many modern hearts, though it has older roots (I discuss it in more detail in “Welfare and Self-Governance”, Ethical Theory and Moral Practice, 9, 2006, pp. 289-309, and in “Morality as Selfgovernance - has it a future?”, Utilitas, 16, 2004). All of which makes Sidgwick’s general complacency about the ease of reconciling intuitionist and utilitarian methods quite surprising.
In an earlier paper I considered another possible interpretation. On this interpretation Sidgwick does not deny that common sense morality contains some intuitive knowledge. He is prepared to accept that common moral intuition does yield some underived or immediate knowledge of normative truths, insisting only that its knowledge is unclear and imprecise, and contains an “element of error”: «Nothing that I have said even tends to show that we have not distinct moral impulses, claiming authority over all others, and prescribing or forbidding kinds of conduct as to which there is a rough general agreement, at least among educated persons of the same age and country. It is only maintained that the objects of these impulses do not admit of being scientifically determined by any reflective analysis of common sense. ... the Morality of Common Sense may still be perfectly adequate to give practical guidance to common people in common circumstances: but the attempt to elevate it into a system of Intuitional Ethics brings its inevitable imperfections into prominence without helping us to remove them» (Henry Sidgwick, The Methods of Ethics, last paragraph of his “Review of Common Sense”, Book III ch. XI, pp. 360-361).
Here Sidgwick does not, it is true, explicitly say that these "distinct moral impulses" amount to intuitive knowledge – only that they claim “authority over all others” and that they are "perfectly adequate to give practical guidance". But his only point against them is that their objects cannot be "scientifically determined" – made precise – by reflective common sense alone. So unless Sidgwick has surprisingly overlooked the obvious point that there can be knowledge that is not precise knowledge, he could perfectly well grant that common sense morality yields such imprecise yet immediate knowledge, which is adequate for (most) practical purposes.
Take the obligations of gratitude, which feature in Sidgwick’s discussion of the morality of common sense. He points out the unclarities in their «nature and extent» (Henry Sidgwick, The Methods of Ethics, p. 260). But pointing out such unclarities is perfectly consistent with accepting that I know the following: (a) that if a person has done me a lot of good out of sheer good will, I have reason to be grateful and show it (b) that lack of gratitude, taken far enough, can become blameworthy. The unclarity, or rather imprecision, of (a) and (b) are evident, but so is their truth. Would Sidgwick, if challenged, deny that we immediately know that truth? Would he say that we have to derive (a) and (b) in some way from the impartialist principle to know them? On the second interpretation, he would not. On this interpretation Sidgwick's view is that common sense morality contains intuitive knowledge, though through lack of precision it cannot qualify as 'scientific' or 'philosophical' knowledge. And he could add that common moral intuitions can be defeated in one way or another by considerations coming from impartial good – he might say that moral common sense itself accepts that.
On this second interpretation Sidgwick’s position would be in my view much more plausible than it is on the first interpretation. However on rereading relevant passages I am less inclined than I was to think it is what he might have meant. My reason is that if this is to be Sidgwick’s position he might be expected to have done one of two things he notably fails to do. On the one hand, of the two allegedly self-evident premises that I quoted above he might have kept the first, impartialism about the good, while dropping the second. Thus he would not have claimed that I am morally bound to aim at good generally. He might instead have tried to replace that by postulating a more indirect teleological connection between his utilitarian theory of the good and common-sense morality, a connection that would still show how intuitions about moral obligation can be defeated by facts about consequences for the good. On the other hand, he might have accepted both of the allegedly self-evident premises, and concluded that there is not just a dualism of practical reason but a three-way conflict.
I think he should have done the first of these things. But let’s consider the second; and first let’s consider the famous “dualism
of the practical reason” as he presents it.
3. There is in fact some difficulty in understanding what Sidgwick had in mind. His general point is that egoism and impartialism are both self-evident, or appear to be, but that they may be contradictory. Thus practical Reason feels:
a “vital need” of proving or postulating [a] connexion of Virtue and self-interest, if it is to made consistent with itself. For the negation of the connexion must force us to admit an ultimate and fundamental contradiction in our apparent intuitions of what is Reasonable in conduct; and from this admission it would seem to follow that the apparently intuitive operation of the Practical Reason, manifested in these contradictory judgements, is after all
illusory (Henry Sidgwick, The Methods of Ethics, p. 508).
The contradiction that Sidgwick seems to have in mind is that egoism and impartialism may dictate incompatible actions; at this point he famously considers the legitimacy of postulating a Supreme Being who ensures that they never do – but stops short of endorsing it. Well, if he’s ready to consider that, then why shouldn’t he be ready to consider a Supreme Being who reconciles not just egoism and impartialism, but each of those with common sense morality? Something like this seems to me to be his position.
Yet surely the contradiction between egoism and impartialism is profounder, holds necessarily, and cannot be eliminated by any purely religious postulate. Egoism and impartialism, as Sidgwick presents them, are not just views about what ought to be done, they are views about why it ought to be done. If we simply presented them as follows, then certainly there would be no immediate contradiction:
You ought to do what will most promote your good
You ought to do what will most promote the general good.
A contradiction would arise only if they dictated incompatible actions. And not necessarily then. We need some further assumptions to get a contradiction from “You ought to do X, you ought to do Y, doing X is incompatible with doing Y”. For example: “if you ought to do X and doing X is incompatible with doing Y then it’s not the case that you ought to do Y”. Such assumptions have been disputed by believers in moral dilemmas. But this is not all that Sidgwick argues. In each case he argues something stronger, about what makes an action obligatory or right, as follows:
The fact that makes an action right is that it will promote your good (taken as a whole).
The fact that makes an action right is that it will promote the general good (taken as a whole).
Egoism and impartialism, as one might say, make totalising claims. Their claims are directly contradictory unless the facts they mention are one and the same fact. That is, their claims are consistent only if your good as a whole is identical with the general good as a whole. At this point those who know about the other great British moral philosopher of Sidgwick’s time, Thomas Hill Green, will immediately recognise the thesis that is required. We are out of Sidgwick territory and into Green territory. The claim that is required to avoid contradiction is not a claim
about a Supreme Being at all, but an absolute idealist claim about our common identity. Green argued for it by identifying a person’s good with what that person truly desires and then arguing that what a person truly desires is the general, or as he put it, the common good. Our good is the common good because we realise ourselves most fully by selfless (in one sense!) dedication to the common good.
Sidgwick made many good points about Green’s doctrine of the identity of individual and common good, and I certainly have no wish to defend it. My point is that his own view is not in a better state. It’s not enough for a Being to exist which ensures that an action promotes my good if and only if it promotes the general good.
It’s too glib to say that without God Sidgwick’s dualism is contradictory. We should say that unless Green’s idealist identification of individual and common good can be somehow vindicated the “dualism” of practical reasons, as conceived by Sidgwick, is contradictory. So if Green’s idealism is indefensible, then on Sidgwick’s account there is a contradiction in practical reason. In his own words, «the apparently intuitive operation of the Practical Reason, manifested in these contradictory judgements, is after all illusory» (Henry Sidgwick, The Methods of Ethics, p. 508). By his own criteria for moral axioms, we are left with no moral axioms at all. Roger Crisp and Bart Schultz both discuss Sidgwick’s moral epistemology in their contributions. Sidgwick sets up four criteria for “moral axioms”. One of these is that they should not conflict «with any other truth». Egoism and impartialism fail that test: they cannot both be true (Henry Sidgwick, The Methods of Ethics, pp. 338-343).
4. What should we make of this? The problem does not lie in Sidgwick’s admirable effort to take full account of all the sources of ethics: the distinct claims of morality, of an impartial theory of the good and of ‘egoism’ – or as one might better say, for reasons I’ll come to directly, the domain of personal or agentrelative values. These three – the good, morality, and personal vales all make claims that are real and genuinely distinct in their sources, and hence they must all be taken into account in a sufficiently broad-based moral philosophy. To investigate properly
why Sidgwick’s attempt at this task – his large and fair minded attempt to grasp the ethical as a whole – leads him into contradiction and failure (as he thought himself), would require far more space than I have here. However one root of it, it seems to me, lies in Sidgwick’s narrowly rationalistic meta-ethics – his rather one-dimensional concern with “self-evident axioms of practical reason”.
Consider ‘egoism’. Is it self-evident that the only source of practical reasons is one’s own good? Far from it: it’s not even true. It is true that the fact that an action will promote one’s own good is a reason to do it; but it is false that it is the only reason to act. An impartialist can agree with both points. But further, one’s own good is a source of reasons for action distinct from impartialism: Sidgwick is right about that.
In what way is it distinct though? To answer this question we must move away from alleged axioms of “pure practical reason”, and turn to reasons for feelings, sentiments, inclinations. It is essential to recognise that there are such reasons, call them evaluative reasons, and that they are distinct from practical reasons. For a person’s good, as I’ve argued elsewhere, is characterisable as what there is reason for that person to desire – where “reason to desire” should be thought of as belonging to the category of evaluative not practical reasons.("The Concept of a Person’s Good", on my website)
How then do evaluative reasons connect to practical reasons? One way is through the following bridge principle:
If there's reason to feel there’s reason to dp what feeling characteristically disposes one to do; that is, to express dp in its characteristic way.
Here are some examples. Out of sheer goodness of heart, someone does me an unrequested good turn. That fact certainly gives me reason to feel grateful. And because I have reason to feel grateful to him for his good turn, I have reason to express that gratitude, for example by thanking him or giving him a present or by returning the favour. Suppose on the other hand that he did me some undeserved harm. In that case I have good reason to feel resentful. And because I have reason to feel resentment I have reason to express that resentment, by recrimination, insistence on apology or even by seeking compensation. Likewise, if there's reason to be frightened of something then there's reason to avoid it; if there's reason to be bored by something then there's reason not to attend to it.
Now as Anscombe once noted the primitive sign of desiring something is trying to get it. So by the general principle we have the following special case:
If there's reason for me to desire X then there's reason for me to pursue X, to try to get it.
If we substitute the account of a person’s good in terms of reason to desire into this special case, we have the following result:
If X is a part of my good then there's reason for me to pursue X, to try to get it.
This is one half of the formal principle of egoism; crucially however, it is a conditional only, not a biconditional. The truth in egoism is that there is a special sort of reason to pursue one's own good; and this follows from the definition of a person's good together with the general bridge principle linking reasons to feel with reasons to act. That, I suggest, is the right location for it on the map of practical reasons. My reason to pursue my own good fits in as part of my reason to pursue various personal ends, all of which are characterisable in terms of the various things there is reason for me variously to feel.
We can now add that the right account of moral obligation must also bring in evaluative reasons because it must bring in the sentiments involved in the response of blame. One cannot understand what moral obligation is without understanding that when a person fails to do what they are morally obliged to then that, in the absence of extenuation, is a reason for those feeling of blame towards him. Here too then we need the concept of evaluative reasons – reasons to feel. The source of moral knowledge is not some purely reason-driven “moral intuitions” but insight into what there is reason to blame.
And yet, when we come to impartiality in theory of the good we find that is not based in any of these ways on the reasonable sentiments, on what there is this or that reason to feel. Ethical impartiality is something the sentimentalist tradition in moral philosophy has always had difficulty in coping with. For here pure practical reason does have a role. Here the best authority is another of Sidgwick’s masters: not Aristotle, Butler or Mill, but Kant. As Kant says, pure practical reason is simply the pure or disinterested will, a disinterested will that is not be identified with any particular inclination.
These are sketchy remarks which I’m not proposing to pursue in this lecture. I simply want to end by noting that they present a sketch of ethical that has something of the tripartite structure Sidgwick presents in his three methods. Sidgwick’s structure suffers from being forced into a rationalistic strait-jacket. My criticism of that is not that there is no such thing as pure practical reason, but that Sidgwick focuses on it too exclusively, and fails to take into account the rational role of the feelings in ethical judgements. For that reason he does not see the full nature and distinctiveness of what he misleadingly calls the methods of “egoism” and “intuitionism”.
This is not to deny the greatness of his achievement. As I said at the beginning, he belongs in the select group of moral philosophers who try to maintain an open-minded view of the ethical in its whole breadth. Because he grasped the distinctness and variety of the ‘methods of ethics’, and pursued their distinctive characteristics with such admirable care, he made one of the lastingly important contributions to moral philosophy.
Published with the author's consent
©2008-2009 Hortense Geninet