In reading Peter Singer’s “Famine, Affluence, and Morality”, one might extrapolate a relationship with the universalizability of an action that echoes Henry Sidgwick’s claims on universalizability in The Methods of Ethics. In contrast to a deontological tradition – which understands moral good as being inherent to an action in itself rather than the outcome of that action and therefore must consider all actions of a certain kind of be categorically (universally) good – Singer and Sidgwick both argue about actions that derive their goodness or permissibility not categorically but rather because others are not doing them. For Sidgwick, making a false promise is morally permissible only because most others are not doing the same; For Singer, donating a large portion of one’s wealth is morally required because others aren’t doing the same. Such a reading – in its demonstration of an apparent analogy between the two and their implied relationship between the moral individual and society at large – prompts the question: does Singer’s argument imply the esoteric morality for which Sidgwick advocates?
I argue that, though esoteric morality may be a major component of Sidgwick’s Methods of Ethics, his argument around the occasional permissibility of usually impermissible actions – demonstrated here by the case of the false promise – does not inherently hinge on esoteric morality. Singer’s argument, therefore, does not either. The thought that it might, I believe, comes from the conflation of fundamentally different sorts of claims, where the confusion comes from a framing that foregrounds the relationship with the other. I argue this point by first reconstructing Singer’s argument, then by providing the example of the false promise in Sidgwick (as a case in which something is morally permissible but only because it is not done by others), along the way providing a breakdown of Sidgwick’s exploration of a rational moral theory (a ‘Method of Ethics’) and its relationship with Common Sense and Moral Intuition. From this relationship I will derive Sidgwick’s argument for esoteric morality, after which I will put this formulation in conversation with the previously discussed case of the false promise and Singer’s argument.
A brief reconstruction of Singer’s argument
Singer’s argument is, perhaps, best illustrated by the example he uses of the drowning child. “If I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out. This will mean getting my clothes muddy, but this is insignificant, while the death of the child would presumably be a very bad thing” (231). While this is a scenario in which one’s ability to prevent death is plainly visible and immediate, there seems to be no apparent logical reason that the nature of a sensory experience leading up to saving a life should be a morally relevant consideration when deciding whether or not to do it. Whether the act of saving life consists of wading into muddy water and saving someone right in front of you or donating a couple of hundred dollars to charity and saving someone across the ocean does not make a difference: either way, those costs are incomparable to the value of a life. Distance, in other words, is not a morally relevant consideration. Therefore, Singer’s argument is a simple one. If death and suffering are bad, and it is within one’s power to prevent something from happening without losing something of comparable moral importance, then one ought to prevent death and suffering whenever they can.
Of course, this is a radical departure from conventional moral intuitions: one typically would not blame anyone for saving the extra $200 dollars from their paycheck they did not spend this month; one typically would not blame anyone for spending money to travel or having a nice dinner every once in a while. But Singer’s argument would hold that one is required not to do these things, and instead to save lives with those resources.
In a sense, this argument does hinge on others not behaving in this way: the suffering and death of so many people around the world could be averted if more people were to try to stop it with the resources they had, it is the fact they don’t that means that the considerations of some lives (like where to travel) are not morally comparable to the considerations of others (on the level of basic subsistence). At its core, though, Singer’s argument is one about moral comparability – this framing in terms of the other is tangential to the crux of Singer’s logic.
Sidgwick’s various arguments and their relationships
Sidgwick acknowledges the Common Sense intuition that Fidelity and the upholding of one’s promises is valuable in itself. The way Sidgwick interacts with this intuition is through a sort of reasoning that he establishes from the very beginning of his treatise. He writes: “‘Method of Ethics’ is explained to mean any rational procedure by which we determine what individual human beings ought – or what is right for them – to do” (1). Thus, any aspects of a morality of Common Sense – like the intuition that one should not make a false promise – should be evaluated by putting it into a scientific form (the form of rational procedure). That scientific form necessarily consists in self-evident propositions (338). This exercise reveals whether the premises of our common moral reasoning, when made explicit, are coherent. Yet, in the case of the promise, rationalizing the Common Sense intuition into the proposition “one ought to keep their promises” must be qualified. There are two approaches to this sort of qualification:
I. Listing the manifold situations under which a promise is not binding and qualifying the principle according to those situations. The binding nature of a promise is contingent on a particular sort of agreement between parties involved being met – the promise, for instance, might lose its binding nature “if it was made in consequence of false statements… or if important circumstances were concealed… or we were in any way led to believe that the consequences of keeping the promise would be different than what they turn out to be; or if the promise was given under compulsion” (353). This fails Sidgwick’s test: a principle that must be qualified in such a complicated way can hardly be considered to be self-evident at all!
II. Limiting the binding nature of a promise to only apply to cases in which “the amount of harm done [by] disappointing the expectations” (354) is greater than the harm done by keeping the promise. Note that this is, in a sense, a similar sort of movement to Singer’s: here, an action often considered impermissible is translated into the realm of permissibility; in Singer, an action often considered permissible but not required is translated into the realm of requirement. This account makes more explicit Sidgwick’s notion that the intuition that a promise ought not be broken does not come from an absolute duty, but rather on an outcome-based moral approach. “[W]e perceive that the obligation must be regarded as contingent on the reliance that another has placed on my assertion: that, in fact, the breach of duty is constituted by the disappointment of expectations voluntarily raised” (354). The problem with making a false promise, for Sidgwick, does not lie in some inherent categorically true property of the duties one has when promising; rather, it lies in the outcome of one’s actions around the promise. In this case, a principle is reached, but it is – as Sidgwick points out a tendency towards (342) – a principle that is simply the utilitarian standard: evaluating the morality of an action on the basis of the happiness created by the outcome.
While (I) can be immediately eliminated, it is important to evaluate the proposition in (II) according to Sigdwick’s conditions for a self-evidence: that its terms are (a) clear and precise (338), that (b) its self-evidence can be ascertained by careful reflection (339), and that (c) it is mutually consistent with other self-evident principles (341). It seems that the fundamental reasoning in (II) is that one ought to act towards an outcome with less harm over an outcome with more harm. This is (a) clear and precise, lacking the many qualifications that convolute the proposition in (I). Sidgwick ascertains its truth by (b) a “careful reflection” on what an ethics is – a rational procedure – that leads him to come to the conclusion that a rational procedure can only be devised based on the outcome of actions*, and of course, if this is the method, then a better outcome is better than a worse outcome. The proposition is (c) mutually consistent with other principles Sidgwick holds to be self-evident: the Axiom of Rational Benevolence, which holds that because from the point of view of the universe, the good of one individual is no more important than any other and that because “‘acting rationally’ is merely another phrase for ‘doing what we see to be right’” (375), a rational being must therefore have as much concern for the good of any other individual as oneself. Deciding whether or not to keep a promise, according to proposition (II), ought to be done with the same level of concern for the good of any other individual as oneself.
Though this is the core of Sidgwick’s argument, it is worth noting that a component of the calculus about whether breaking a promise will cause more or less harm than keeping it is how others behave in promise interactions. A promise, Sidgwick points out, must be “understood by both parties in the same sense. And by the term 'promise' we include not words only, but all signs and even tacit understandings not expressly signified in any way, if such clearly form a part of the engagement. The promiser is bound to perform what both he and the promisee understood to be undertaken” (304). If everyone were to make false promises, this would be an impossibility – the frameworks of consent and tacit agreement on which the system of the promise is contingent would fall apart. As Sidgwick puts it, “If any of these conditions fails, the consensus seems to become evanescent, and the common moral perceptions of thoughtful persons fall into obscurity and disagreement” (311). This must factor into the calculus: if the harm done by disappointing expectations is not just a harm to the promisee, but rather harm to the concept of the promise in such a way that it would be impracticable to even make a promise at all because promises are broken too frequently to even be “understood by both parties in the same sense” (304), then surely that is a harm to great to make a false promise or break a promise made. Therefore, the permissibility of the breaking of a promise is contingent on the fact that others do not do the same thing.
One might hesitate to come to such a formulation might seem to violate requirement (c) from above: there is an air of mutual inconsistency between this perspective and Sidgwick’s Axiom of Justice, which states that “whatever action any of us judges to be right for himself, he implicitly judges to be right for all similar persons in similar circumstances” (379). However, it seems that the reasoning around other persons in the case of the promise hinges on the very fact that they are not similar people in similar circumstances. From the perspective of a rational promise-breaker, similar people in similar circumstances – that is, circumstances where the act of doing so would not significantly challenge the standing of the logic of the promise in society, therefore prompting people to break a promise because keeping it would cause more harm than good – could permissibly break their promises. People in different circumstances – which could mean circumstances where the idea of the promise is at risk and that plays a significant role in a utilitarian calculus of the outcome, therefore prompting people to keep a promise because breaking it would cause more harm than good – can’t permissibly break their promises. This line of reasoning is therefore mutually consistent with the Axiom of Justice.
I will now return to the formulation that the permissibility of the breaking of a promise is contingent on the fact that others do not do the same thing. This brings out a fundamental part of Sidgwick’s notion of how an Ethics ought to operate: “we study Ethics, as Aristotle says, for the sake of Practice: and in practice we are concerned with particulars” (215) – Sidgwick is very willing to hold that one genre of action would not be permissible universally, but is in a particular case. He cautions that:
Could we indeed assume that the scientific deduction of such a system [a scheme of conduct] would ensure its general acceptance; could we reasonably expect to convert all mankind at once to Utilitarian principles, or even all educated and reflective mankind, so that all preachers and teachers should take universal happiness as the goal of their efforts as unquestioningly as physicians take the health of the individual body; and could we be sure that men's moral habits and sentiments would adjust themselves at once and without any waste of force to these changed rules: then perhaps in framing the Utilitarian code we might fairly leave existing morality out of account. But I cannot think that we are warranted in making these suppositions. (468)
Because Sidgwick’s utilitarianism demands that we evaluate the morality of an action based on its outcome, the morality of even the action of adopting utilitarianism ought to be evaluated as such, too. He is open, therefore, to the idea that other sorts of moral reasoning – like the morality of Common Sense – might bring about a better outcome: “IF, then, we are to regard the morality of Common Sense as a machinery of rules, habits, and sentiments, roughly and generally but not precisely or completely adapted to the production of the greatest possible happiness for sentient beings generally” (475) it might be the best means to arriving at a moral outcome. Hence, Sidgwick’s theory is one that has a divide between what should be epistemologically and morally accepted (i.e. is an esoteric morality): epistemologically, utilitarianism is the rational procedure to evaluate the morality of actions, but morally, the difficulty of doing a hedonistic calculus or getting people to accept utilitarianism means that Common Sense will nonetheless lead to the best outcome.
This argument, however, is a separate one from the argument that Sidgwick uses in the case of the promise: assessing the binding nature of a promise based on the outcome brought about by keeping or breaking it is, in fact, doing so based on utilitarian principle – the way Sidgwick arrived there was not by using Common Sense as a proxy (as esoteric morality might suggest) but instead by attempting to derive from Common Sense a rational procedure. The result is something that should be accepted, in Sidgwick’s eyes, both morally and epistemologically.
The reading that Sidgwick’s allowance for the permissibility of typically impermissible action is connected to his esoteric morality seems to come from the fact that esoteric morality could lead one to similar-sounding conclusions when thought of in terms of a relationship with the other. Accepting esoteric morality can lead to the perspective that different people in different circumstances ought to accept different methods of Ethics: (e.g.) for some people in some circumstances, it is morally best to use moral approaches that lack epistemological grounding; for some people in some other circumstances it is morally best to accept the epistemological truth. This can sound similar to the claim that: for some people in some circumstances, it is morally best to do something generally impermissible; for some people in some other circumstances it is morally best to do the generally permissible thing. But these claims are not the same: one is about how one should generally decide on action (in contrast to how others do), another is about whether or not one should do what others generally think of as permissible. As shown above, the logic that underlies them is separate: neither implies the other, either one could reasonably be held without holding the other.
It is just the common framing around what one can do as compared to others that makes these two claims within Sidgwick seem dependent on another. Upon realizing their separation, the question on whether Singer’s claims depend on Sidgwick’s esoteric morality becomes moot. The analogy between Singer and Sidgwick is limited only to the similarity between their arguments in that they translate conventional interpretations of the morality of an action in the same direction: Singer thinks typically permissible action is required and Sidgwick thinks typically impermissible action can be permissible, both of which are entirely separate positions from esoteric morality.
*Reconstructing the argument here is outside the scope of this project.
Bibliography
Sidgwick, Henry. The Methods of Ethics. (1874), Seventh Edition ed., London, Hackett, 1981.
Singer, Peter. “Famine, Affluence, Morality.” Wiley, vol. 1, no. 3 (Spring, 1972), Spring, 1972, pp. 229-243, https://www.jstor.org/stable/2265052 .