SITUATIONAL RELATIVITY

Certain versions of SC appeal to the value of a situation as the determinant of moral rightness. One such version of SC, developed by Thomas Hurka (1990), says, roughly, that an agent need not maximize utility when a situation is already ‘reasonably good’, i.e. when the value of the situation is already above some absolute threshold. Hurka says, for example, that ‘when [the value of] a situation is and will remain below the absolute threshold, an agent's duty is the same as under maximizing: she must do everything to move it towards satisfactory goodness. Once the threshold is reached, however, her duty vanishes. If a state of affairs is already, by absolute standards, reasonably good, she has no duty whatsoever to improve it.’Footnote ¹² The following version of SC, called ‘situational absolute level satisficing consequentialism’ (SALSC), is a reasonable precisification of Hurka's remarks:

The notion, captured by SALSC, that one need not improve a situation when that situation is ‘reasonably good’ is an instance of the notion that an act is morally right so long as it produces consequences (here, a situation) that are (is) ‘good enough’, i.e. that an act's being morally right does not (always) require maximizing, but only satisficing.

SALSC is admittedly somewhat imprecise as stated; the notion of a ‘situation’ is not entirely clear. Bradley suggests the beginnings of an explication: the situation that would obtain after an act is everything that would happen in the world in a very short period of time after the act is performed; further, the value of the situation is a function of the value of the things that happen within it.Footnote ¹⁴ I will follow Bradley in making the latter claim, though not in making the former. The former claim requires modification. Since any version of consequentialism, as a species of ‘consequentialism in general’, holds that the moral rightness of an act is determined entirely by consequences, the situation that is relevant to an act's moral status is better understood as everything that would happen in the world as a consequence of the act in a very short period of time after the act is performed.Footnote ¹⁵ Fortunately, nothing to be discussed in what follows will turn on this modification.

For, even leaving considerations about the nature of situations aside, it is clear that, as a theory about the necessary and sufficient conditions for an act's being morally right, SALSC simply will not do. Bradley presents the following general counterexample:

Such a result is clearly unacceptable; the completely gratuitous killing of several people is, intuitively, never morally right.Footnote ¹⁷ Further, Bradley is clearly right that SALSC implies that it is morally right in the sort of case described. For suppose that n = 100, and that the value of the situation prior to the act is 500.Footnote ¹⁸ The completely gratuitous killing of several people may produce significant disutility, such that the value of the situation after the act is only 200. Still, the value of the situation after the gratuitous killing is, in this scenario, greater than n. That is, the situation after the act has a value of at least n (100), and so, according to SALSC, the gratuitous killing is morally right. This is the wrong result, and so SALSC is false.

Bradley considers the following modification of SALSC, called ‘individualist situational absolute level satisficing consequentialism’ (ISALSC), which, he believes, has at least the virtue of not delivering such results:

He maintains that ISALSC avoids delivering the aforementioned results for the following reasons. The counterexample to SALSC just described involves the gratuitous killing of individuals. But (says Bradley) presumably the killing of a person will cause that person's welfare level to drop below n (as long as n is any plausible absolute threshold), and so the act of gratuitous killing will not, according to ISALSC, be morally right.Footnote ²⁰ Such gratuitous killings are therefore not endorsed as morally right by ISALSC.

But even assuming that this is right, there are other problems for ISALSC. As Bradley puts it, ‘ISALSC still licenses gratuitous harms.’Footnote ²¹ Suppose, for example, that each person's welfare level in a given situation prior to Smith's act is 200. Further, suppose that n = 100. Now let it be that Smith's act is not a gratuitous killing, but is instead a completely gratuitous bashing of Jon's kneecap with a crowbar. After Smith's act, the welfare level of everyone in the situation besides Jon remains at 200; Jon's welfare level, however, drops from 200 to 125. Jon's welfare level therefore remains at or above the threshold, n. In the situation after Smith's act, each person's welfare level is therefore at least n. But then, according to ISALSC, Smith's act of gratuitously bashing Jon's kneecap with a crowbar is morally right. This is false, and so ISALSC is false.

Both versions of SALSC (SALSC and ISALSC) thus face a common problem: they sanction the causing of gratuitous harm. Put in terms more directly pertinent to these theories, the problem is that each one permits an act that gratuitously creates a situation that is worse (either in itself or insofar as things are worse for some individual(s) within that situation) than the situation that obtained prior to the act, so long as the resultant situation has a value still above the absolute threshold. And clearly – that is, intuitively – such a result is unacceptable. To avoid these difficulties, situational satisficing consequentialists (at least of the ‘absolute level’ sort) must maintain not only that agents are obligated to prevent the situation from falling below some absolute threshold (when possible), but also that agents are obligated not to gratuitously make the situation worse. The theories proposed so far do not deliver this verdict.

That being so, I propose that we allow the intuitions behind Bradley's objections here to push us in the direction of a new theory. The objections considered so far seem to bring out important intuitions about the moral rightness of specific sorts of acts. Acts that cause gratuitous harm are morally impermissible.Footnote ²² Our adjustments will attempt to accommodate these new intuitive data, with the aim of developing a theory that fits this evidence from intuitions about certain acts’ moral rightness.

An appeal to what I will refer to as ‘situational relativity’ can be of some assistance here. The general idea behind situational relativity is that there is something morally relevant about the comparative value of the situation that obtains after an act is performed relative to the value of the situation that obtains before it. A version of SC that appeals to situational relativity can capture the intuitions behind the objections posed by Bradley against SALSC and ISALSC. It can maintain that no morally right act gratuitously makes a situation worse overall, even if the situation already has a value above the absolute threshold.Footnote ²³ The following formulation of a version of situational satisficing consequentialism appears to accommodate this data:

According to SSC, one cannot permissibly make a situation worse by causing gratuitous harm or engaging in gratuitous killing. Since both of the counterexamples so far considered present scenarios where the situation is gratuitously made worse in such ways, SSC does not succumb to them. It is, in that respect, a better version of satisficing consequentialism.

It is also worth noting that SSC seems to better capture the thoughts of certain satisficing consequentialists, thoughts interpreted by Bradley as in some way supporting either SALSC or ISALSC. For example, Bradley quotes Hurka as maintaining that ‘if a state of affairs is already, by absolute standards, reasonably good, [an agent] has no duty whatever to improve it.’Footnote ²⁵ Shortly thereafter, he goes on to say (apparently interpreting this remark) that ‘Hurka suggests . . . that if the situation before an act has at least a certain value, an agent need not do anything.’Footnote ²⁶ Finding this latter claim implausible (because it allows an agent to permissibly let a situation's value drop below the relevant absolute threshold, so long as the prior situation has a value above that threshold), Bradley immediately adjusts his interpretation of Hurka's remark to have it that ‘what really matters is not the value of the situation before the act, but its value after the act’,Footnote ²⁷ and he thus arrives at SALSC. But I do not see that Hurka's remark suggests what Bradley takes it to suggest, and particularly I do not see that it suggests SALSC.Footnote ²⁸ Hurka says only that an agent in a situation with a value already over a certain threshold has no duty whatever to improve the situation. This suggests that (i) he explicitly takes the value of the situation prior to the act – and not just the value of the situation after the act, as in SALSC – to be relevant to determining which acts are morally right, and that (ii) he is at least non-committal regarding an agent's duties (if there be any) with respect to worsening (or not worsening) the situation, saying instead only than an agent need not improve a situation that is already good enough. Hurka makes exactly these qualifications elsewhere; he writes that ‘if the situation is already above the absolute threshold, absolute-level satisficing does not require an agent to do anything at all to improve it’.Footnote ²⁹ He thus seems to be insisting explicitly that the value of the situation prior to the act is relevant, and that situational satisficing consequentialism should only not require that an agent make a situation that is already above some absolute threshold even better. He suggests little, if anything, regarding the obligations of agents with respect to the worsening of (or with respect to allowing the worsening of) a situation – there is no obvious reason to take him to be suggesting that agents in situations of a high enough value are not obligated to do anything. And, in light of just how implausible it is (as Bradley notes) to suggest that an agent in a situation having a value above the absolute threshold can permissibly act to make the situation worse by engaging in gratuitous killing/harming, it seems that the most charitable interpretation of Hurka's remarks is along the very lines suggested by SSC. SSC has it, as Hurka appears to have it, that the values of the situations both before and after the act are relevant to determining moral rightness, and that an agent has no duty to make a reasonably good situation even better – and, importantly, it also has it that an agent cannot permissibly gratuitously make a situation worse.

Unfortunately, in spite of these virtues of SSC, the view is still susceptible to counterexample. Though it does not allow gratuitous harming or killing, it seems to fall prey to an instance of Bradley's more general objection to (almostFootnote ³⁰) all forms of satisficing consequentialism: it sanctions the gratuitous prevention of goodness by allowing an agent to gratuitously ‘bring about a submaximal outcome in order to prevent a better outcome from obtaining’.Footnote ³¹

To see this, suppose that n = 100 and that the value of the situation prior to the act that we are about to consider is also 100. Suppose, further, that Smith, in this situation, is just sitting idly on his couch. His daughter skips into the living room with a cookie that she has just baked and sets it on the table in front of him. ‘I've baked this cookie for mommy,’ she tells him, ‘just because I love her. I'll leave it here so she can eat it when she gets home.’ Now Smith reflects. He knows that his wife would be extremely delighted by this surprise gift, and he knows that he himself would take some mild enjoyment in seeing his wife so delighted. Still, he would also take some equally mild enjoyment in eating the cookie himself. Further, he knows that his daughter is extremely forgetful, will forget (unless reminded) that she ever made such a cookie, and will only be reminded if in fact his wife receives the gift and remarks upon it to her. That said, no one will be harmed if Smith eats the cookie. True, his wife will miss out on some potential delight, but since that delight is not yet hers, Smith will not be actively harming her; and likewise for his daughter. So, Smith eats the cookie.

Smith had at least two alternatives before him. He could have eaten the cookie, or he could have simply remained idle and not eaten it. These alternatives and the values of their respective consequences can be represented as follows (let ‘Total’ = the value of the situation after each act):

As it turns out, the value of the situation prior to the act is 100, which is at least n, and the value of the situation after Smith's act of eating the cookie is 105, which is at least 100. This means that, according to SSC, what Smith did was morally permissible. But it was intuitively not morally permissible; it was selfish, and even quite cruel, for Smith to eat the cookie. This is especially so given that he would have experienced no displeasure in not eating the cookie, and could have simply remained on the couch, watching and waiting until a more pleasant overall situation – one which includes the same level of mild enjoyment for himself – unfolded before his eyes. SSC has thus allowed Smith to actively ‘steer the world towards a worse result’Footnote ³² than it would have otherwise reached, and for no good reason at all.Footnote ³³ This is the wrong result, and so SSC is false.

Since we earlier considered an individualistic version of SALSC, it is worth noting that an individualistic version of SSC fares no better with this example. An individualistic version of SSC might be stated as follows:

Smith's act in this scenario does not maximize utility. However, in the situation that obtains after his eating of the cookie, each person's welfare level is at least equal to its level in the situation prior to the act – again, no one is harmed by Smith's action. ISSC therefore delivers the result that his eating of the cookie is morally permissible. So, ISSC is also false.

SELF-SACRIFICE

SSC (and its individualistic variant, ISSC) has allowed Smith to steer the world toward a worse result, and for no good reason at all. Smith has gratuitously prevented a better outcome by bringing about an outcome that is less than the best, and SSC has endorsed it. This much is enough to show that SSC is false. But the preceding considerations do not show that it is always wrong to prevent a better outcome by bringing about a lesser one. The latter may be permissible when bringing about the better outcome requires enormous personal self-sacrifice.

Thoughts like these are common in the literature on satisficing consequentialism. A view that says something much like this about the permissibility of bringing about sub-optimal outcomes, for instance, is endorsed by John Turri in his defense of satisficing consequentialism. Turri explicitly endorses what he calls the ‘Anti-Maximizing Principle (AM)’:

Turri apparently endorses AM on the grounds that ‘morality will [at times] demand no more than that which can be accomplished without sacrificing anything of appreciable importance’.Footnote ³⁵ On such a view, there are at least possible occasions where the fact that an agent would be required to sacrifice something of ‘appreciable importance’ in order to bring about the best overall outcome makes it permissible for the agent to bring about some suboptimal outcome instead. Insofar as such an agent ‘prevents’Footnote ³⁶ a better outcome by bringing about an outcome that is less than the best for this reason (the thought goes), the agent does not gratuitously do so. The view suggested by Turri thus says that on at least some occasions, non-gratuitously (in this sense) preventing a better outcome by bringing about a lesser is morally permissible.

There are, it seems, good intuitive reasons to think that something like this view is right. Suppose, for example, that an agent has two options before him. On the one hand, he can give everything he owns to charity. The money and goods which he gives will be put to use to make significantly better the lives of three people in a distant, impoverished land, but he himself will become entirely destitute, live the rest of his life in abject poverty, and suffer serious misfortune related to malnutrition and disease. On the other hand, the agent can give an extremely large portion of what he owns to charity, with the result that the lives of two distant people are greatly improved, while the third person is now helped only minimally. On this alternative, the agent himself will retain enough to live a reasonably decent – though far from extravagant or what most would call ‘well-off’ – life. As it turns out, if the agent takes the first alternative, the additional person helped will (unbeknownst to the agent) live a life that is just slightly more pleasant than the one that the agent himself will live if he takes the second alternative. This being so, taking the first alternative will ultimately result in an outcome that is just slightly better overall.Footnote ³⁷ But surely the agent's taking of the second alternative, where he still gives massive amounts to charity and retains only enough to get by decently, is at least morally permissible. It is not the case that the only morally permissible option for the agent is that of sacrificing his well-being entirely (undergoing poverty, disease, and malnutrition) in order to make it the case that there is a life that is just slightly more pleasant than his would be.Footnote ³⁸ While taking that option may be permissible as well, and perhaps even honorable, it is not morally required. Thus, some actions are morally permissible despite the fact that an available alternative would promote a better outcome. This is at least so when performing the act with the better outcome requires significant self-sacrifice.

Along similar lines, Peter Vallentyne notes that there is a ‘common-sense’ notion that some acts are supererogatory supporting the idea that it is sometimes morally permissible to perform an act with less than the best consequences. A supererogatory act is one that goes ‘above and beyond’ what is morally required.Footnote ³⁹ Common sense – or, at least, generally shared moral intuition – seems to suggest that there are such acts. Vallentyne notes that, ‘according to common sense, devoting all one's income and time to help others is supererogatory’.Footnote ⁴⁰ In cases where such devotion is an available alternative, ‘common sense . . . holds that some permissible actions are better than others’Footnote ⁴¹ – that is, it holds that some non-best acts are morally permissible. The scenario presented above appears to be one example where this is the case. All of this suggests that the intuitive data here are on the side of a version of satisficing consequentialism that allows enacting non-maximizing alternatives in at least some situations. The data are at least not on the side of a version of maximizing consequentialism. The data support the notion that some morally permissible acts are not (overall) the best, particularly in cases where significant self-sacrifice is involved.

Bradley attempts to develop a version of satisficing consequentialism that accommodates such considerations. The crucial notion to which he appeals, suggested by the above remarks, is the notion that an agent can perform a morally permissible act that does not maximize utility when it is both ‘good enough’ (has consequences with a utility of at least n) and performing the maximizing act would be ‘worse for’ the agent. He calls the version of SC that he develops on the basis of these considerations, suggested to him by Garrett Cullity, ‘Cullity's self-sacrifice absolute level satisficing consequentialism’ (CSSALSC):Footnote ⁴²

Bradley contends that CSSALSC, despite its initial appearance of promise, falls prey to something like his general objection to all other forms of satisficing consequentialism: it says that it is morally permissible for an agent to prevent some better outcome by bringing about some lesser outcome, even when doing so is (intuitively) clearly wrong. The prevention of goodness that CSSALSC allows is not gratuitous, however; it is, instead, a prevention of goodness that is done for a reason that is intuitively just not good enough, i.e. for a reason that is at least not sufficient for making the prevention of goodness morally permissible.Footnote ⁴⁴ Call this an inappropriate prevention of goodness. Bradley presents an example to this effect as follows:

The claim here is essentially this: act A2 is intuitively morally impermissible, and yet CSSALSC says that it is permissible, because A2 has a utility of at least n (20), and any better alternative (A1, for example) is such that it is worse for the agent, Jack, than A2. CSSALSC is therefore false.

This conclusion regarding CSSALSC hinges crucially upon what it means for an alternative to be ‘worse for’ some agent. Alternatives A3 and A1, in the example considered, are apparently worse for Jack, according to Bradley, because they do not make things as good for Jack as some other alternative would. But here is where pressure can once again be applied to Bradley's interpretation of the particular version of satisficing consequentialism under consideration. It seems that there is a more plausible interpretation of what it is for an alternative to be ‘worse for’ an agent that both better captures the intuitive data considered above and makes it the case that CSSALSC does not succumb to this counterexample.

The kind of self-sacrifice that is taken to be relevant in the preceding considerations is the kind that results in an agent's being less well off than he was prior to the sacrifice. In one of the examples, the sacrifice involves becoming destitute and diseased. In another, the sacrifice requires devoting all of one's income and time to others. Turri's defense of a version of self-sacrifice satisficing consequentialism is motivated in part by an example where an agent, in order to perform the maximizing act, must go from being a healthy human being to being entirely paralyzed.Footnote ⁴⁶ All of this suggests that Bradley's interpretation of the phrase ‘worse for’ in CSSALSC fails to result in a theory that best fits the intuitive data or that best fits what self-sacrifice satisficers appear to have in mind. Rather, the general idea seems to be that an alternative is ‘worse for’ an agent if and only if that alternative makes the agent less well-off than he was prior to acting. An alternative is worse for an agent if and only if the alternative has consequences with net disutility for the agent relative to the agent's prior level of welfare – in this case, things are worse for the agent after enacting the alternative than they were prior to doing so. In short, saying that an alternative is ‘worse for’ some agent really amounts to saying that it makes things worse for him.

This interpretation of ‘worse for’ can be explicitly incorporated into a version of satisficing consequentialism. Consider CSSALSC2:

CSSALSC2 does not succumb to the counterexample to CSSALSC. A2 has a utility of at least n (20), but it is not the case that any better alternative is such that, were that alternative to be enacted instead of A2, the agent's resultant personal welfare level after the enaction would be less than it was prior to the enaction. Indeed, of alternatives A1, A2, and A3, CSSALSC2 says that only A1 is morally permissible. This is exactly in line with our intuitions regarding the case. This is reason to think that CSSALSC2 is a more plausible view than CSSALSC.

Even so, CSSALSC2 is still not fit to accommodate all of the intuitive data. Suppose that in alternative A1, Jack's foot falls asleep as he continues sitting on the couch, and so enacting A1 results in −1 utility for Jack. With the utilities thus modified, CSSALSC2 delivers the result that it is permissible for Jack to intercept the envelope on its way to charity and give it to Will. This still does not seem right.Footnote ⁴⁷

The reason it does not seem right, I think, is that the sacrifice is not significant enough. The sacrifices considered above are not sacrifices that involve some small amount of disutility. Rather, they are sacrifices that involve things becoming significantly worse off for the agent. These are the sorts of sacrifices that make it the case that it is permissible for an agent to bring about some lesser alternative. Turri, in a previous citation, explicitly uses the language of ‘appreciable importance’, saying that sacrifices requiring the agent to sacrifice something of such importance are the ones that (can) make it the case that it is permissible for the agent to bring about consequences that are less than the best.

These points can also be incorporated into a version of satisficing consequentialism, though something more should first be said about what it is for a sacrifice to be ‘significant’ or ‘of appreciable importance’. What I will say is the following. It seems to me that these notions involve a certain amount of ‘relativity’, as the following example shows. Suppose that an agent is faced with the alternatives of either wrapping the gift which he has purchased for his wife or not doing so.Footnote ⁴⁸ Suppose that if the agent wraps the gift, his wife will be quite pleased with it, but he will endure a stinging paper cut (and not enough pleasure to ‘make up for it’ on his end). Suppose that if he does not wrap the gift, his wife will still be rather highly pleased with it, and he will endure no paper cut. Even though the former alternative will (suppose) result in slightly higher overall utility, it seems that the latter option is at least morally permissible, given that the agent is, in that alternative, able to avoid suffering the paper cut, and his wife is still highly pleased. Perhaps this means that, given the only minor difference in resultant overall utilities between the two alternatives (the agent's wife is only slightly happier in one as compared with the other), the sacrifice of enduring the paper cut is, relative to that very minor difference, significant enough to be relevant to moral permissibility. This suggests, perhaps, that the level of significance of a sacrifice required for the sacrifice to be morally relevant (i.e. of ‘appreciable importance’) is determined by considering the disutility of the sacrifice for the agent relative to the difference in overall value between the available alternatives. If the sacrifice is very small relative to the difference in overall utility, then it will not be of appreciable importance. But if, relative to the difference in overall utility, the sacrifice is not comparatively minor, it will be morally relevant. I will call a sacrifice that is significant in this latter sense a comparatively significant sacrifice.

Here is the resultant view:

This view appears to avoid the counterexamples to CSSALSC presented by Bradley and myself, and to accommodate all of the aforementioned intuitions about self-sacrifice. Returning to the modified example of Jack and the letter to charity, for instance (i.e. the version of this example wherein Jack's foot falls asleep if he takes alternative A1), CSSALSC3 delivers the verdict that only option A1 is morally permissible. This is because neither of the non-maximizing alternatives, A2 or A3, is such that any better alternative (namely, A1) involves a comparatively significant self-sacrifice for Jack. Relative to the overall differences in utility between the better alternative, A1, and those alternatives not involving any sacrifice on Jack's part – a difference of 79 in the case of A2 and a difference of 49 in the case of A3 – Jack's sacrifice with a disutility of −1 is not comparatively significant.Footnote ⁴⁹ In Turri's language, it is not of ‘appreciable importance’. The verdict that CSSALSC3 delivers here is exactly in accord with our intuitions about the case.Footnote ⁵⁰ As such, CSSALSC3 appears to have substantial promise.

A FINAL, HYBRID VIEW

CSSALSC3 is a version of satisficing consequentialism that avoids Bradley-style objections. Showing this is already showing something significant; it is showing that there is at least one version of satisficing consequentialism that meets his condition (ii), that of not allowing the gratuitous (or inappropriate) prevention of goodness. We have also seen some reason, mentioned above, to think that CSSALSC3 accommodates the intuitive data suggesting that some acts are supererogatory, which intuitive data versions of maximizing consequentialism appear to have difficulty accommodating. This being so, CSSALSC3 is a version of SC that meets the whole of Bradley's challenge. But I wish ultimately to arrive at what I think is the most plausible version of satisficing consequentialism vis-à-vis maximizing consequentialism, and I do not think that CSSALSC3 is it. It succumbs to an objection unlike those considered so far.

The objection is this: despite the fact that CSSALSC3 is a satisficing view, it is still ‘too demanding’.Footnote ⁵¹ More specifically, the view asks too much in certain cases where the situation is already ‘good enough’. Suppose, for example,Footnote ⁵² that Trent has just eaten a quite respectable lunch and finds himself no longer hungry, though also not completely stuffed. If Trent were to eat a candy bar, he would enjoy it. Further, there is (suppose) just such a candy bar at the snack stand nearby. It is available for free, as part of a promotion running in Trent's area. In this situation, Trent has at least two alternatives available, one being that of remaining on the park bench where he has just eaten lunch, and the other being that of walking to the snack stand, getting the candy bar, eating it, and then returning to the bench. The latter alternative, suppose, is the alternative that maximizes utility. Taking that alternative requires no sacrifice on Trent's part; things will not become worse for Trent should he consume the candy bar. Taking the other alternative, though, is still pretty good; it results, suppose, in a situation with a value greater than the relevant threshold. However, because taking the maximizing alternative requires no comparatively significant sacrifice on Trent's part, CSSALSC3 delivers the result that taking this lesser alternative is not morally permissible.Footnote ⁵³ And that does not seem right.Footnote ⁵⁴ Surely it is permissible for Trent to relax on the park bench rather than eat a candy bar, even if it is not best, and even if the best requires no significant self-sacrifice. This seems to be a kind of satisficing that is an intuitively morally permissible alternative to maximizing.Footnote ⁵⁵

We saw earlier that SSC did a decent job of accommodating these latter intuitions. Its flaw was that it allowed the gratuitous prevention of the best results. It was for precisely this reason that we moved to a consideration of views which allowed the prevention of best results only when significant self-sacrifice would be required to achieve them, and arrived at CSSALSC3. This suggests that SSC and CSSALSC3 each face a particular difficulty that the other avoids. CSSALSC3 is too demanding; SSC is not. SSC allows the gratuitous prevention of the best results; CSSALSC3 does not. The situation is such that not attempting a combination of at least some elements of the two views (in order to avoid both problems) would be negligent.

Consider the following attempt, which I will call ‘situational absolute-level self-sacrifice satisficing consequentialism’, or SASSC. SASSC incorporates, from CSSALSC3, the notion that comparatively significant self-sacrifice is morally relevant, in the sense that it can play a role in making it the case that some act with consequences less than the best is morally permissible. It also incorporates, from SSC, the notion that the value of the situation after an act, relative to the value of some other situation(s),Footnote ⁵⁶ can play a role in making it the case that not maximizing is morally permissible.

SASSC is our most complex version of satisficing consequentialism yet. Some of its elements require explanation. I will provide these explanations and then go on to argue that the theory avoids all of the problems that the previous views faced.

It is first worth stating SASSC in somewhat more colloquial terms, so as to better understand exactly what it claims. SASSC says that there is some absolute threshold, n, that the value of a given situation must be at or above in order for satisficing to be permissible; otherwise, maximizing is required. When the value of the situation is above the threshold, satisficing is permissible only if a number of additional conditions are met. First, it must be that the satisficing action results in a situation with a value at least as high as the value of the situation preceding the action; no permissible satisficing action can make a situation worse.Footnote ⁵⁹ Second, it must be that any alternative action that results in greater overall utility than the satisficing action satisfies at least one of the following conditions: either (i) enacting the alternative instead of the satisficing action would require a comparatively significant self-sacrifice (i.e. a sacrifice of ‘appreciable importance’) for the agent, or (ii) the overall utility that would result from the enaction of the alternative is not appreciably greater than the value of the situation that results from the enaction of the satisficing alternative. When all of the preceding conditions are satisfied, SASSC says that the act is morally permissible.

Stating SASSC in both its formal and colloquial terms has required appeal to the (new) notion of a situation's value being ‘appreciably greater’ than the value of some other situation. This needs explanation. As I am thinking of things here, whether or not the value of one situation is appreciably greater than the value of another is determined by comparing the difference in value between the two situations with the cost to the agent of bringing about the situation with the greater value as opposed to the one with the lesser.Footnote ⁶⁰ If, for example, the difference between the values of the situations is quite minimal, and bringing about the one with the greater value has a large cost to (i.e. requires a large amount of exertion or energy from) the agent, then the value of the better situation is not appreciably greater than the value of the lesser. If one situation is somewhat greater in value than another and bringing it about exacts quite minimal cost, however, then the value of the situation is indeed appreciably greater. Here, it is important to note that the ‘cost to the agent’ is not to be understood in terms of sacrifice (as a loss of personal welfare) for the agent. Rather, ‘cost to the agent’ is intended as a measure of the amount of energy, work, or exertion required on the agent's part to bring about some alternative. Such work or exertion need not involve a sacrificial loss of personal welfare on the agent's part.

Given this understanding of one situation's value being ‘appreciably greater’ than the value of another, SASSC is not too demanding in the way previously considered. It does not say that it is morally impermissible for Trent to relax on the bench rather than eat a candy bar. At least, it does not say this given certain assumptions. Assuming that the value of the situation in which Trent eats the candy bar is not appreciably greater than the situation in which he remains sitting on the bench – which seems plausible: the transition to the former situation exacts some cost (Trent's leaving his bench, walking to the snack bar, requesting the chocolate, etc., when he feels no real desire to do so) in exchange for extremely minimal value – then it is morally permissible for Trent to remain sitting on the bench.Footnote ⁶¹ The alternative in which Trent eats the candy bar instantiates a situation with a value not appreciably greater than the value of the situation that obtains when Trent remains on the bench. In summary, the value of the situation preceding Trent's act of remaining on the bench is, we are supposing, above the absolute threshold (it is ‘good enough’), his act does not make the situation any worse, and the overall better alternative, were it to be enacted, would not result in a situation of appreciably greater value than the one that results from the act under consideration. So, according to SASSC, Trent's act of remaining on the bench is morally permissible.

SASSC also does not fail in the opposite direction, by allowing the inappropriate or gratuitous prevention of goodness. Consider again the case of Jack and the letter to charity, for example. Alternative A1 comes out as morally permissible, according to SASSC, and alternatives A2 and A3 come out as impermissible, which is just the right result. This is because alternatives A2 and A3 have a better alternative, A1, which is such that (i) its enaction does not require significant self-sacrifice and (ii) its value is appreciably greater than the value of the situations that obtain in the other alternatives (A2 and A3). This shows that SASSC avoids the problems that remained for CSSALSC3 and for SSC. That is an advantage of the view.

The view's only obvious disadvantage – one that I have already noted – is its complexity. But given the theory's demonstrable merits, I do not regard this as a serious disadvantage. In fact, I do not regard it as a disadvantage at all; given the apparent complexity of moral matters, the theory does not seem more complex than should be expected. This being the case, I recommend it for further consideration. SASSC is a version of satisficing consequentialism that (i) does not allow for the gratuitous or inappropriate prevention of goodness and that (ii) solves the problem of accommodating the intuitive data regarding both supererogation and demandingness that other versions of consequentialism fail to accommodate. It thus meets Bradley's challenge and stands as a version of satisficing consequentialism worth considering.