Before presenting my own rejoinder in Section 4, I will now examine two possible but unsuccessful replies to this challenge.
A Non-Debunking Explanation
To be fair, LRS had anticipated Kahane’s objection. In the piece to which he refers, they acknowledge that UB “needs a theory of well-being, or else it is empty of content” (LRS 2012: 27). Still, they don’t take the issue very seriously and are satisfied with “pointing out that if no theory of well-being … were immune to a debunking explanation, this would show only that no theory could be preferred over others on the ground that it alone cannot be debunked” (LRS 2012: 28). But, as we just saw, we are not in such a stalemate situation, since some of our prudential beliefs were not shaped by evolution.
More recently, LRS have granted the issue more attention. They now deny that our belief that pleasure is good and pain is bad was selected for:
To be motivated to seek pleasure and avoid pain, it is not necessary that we have the normative belief that pleasure is good and pain is bad. The way pleasure and pain feel is already sufficiently motivating. Hence, there is no reproductive advantage in our holding that belief. In this respect, it contrasts with, for instance, the normative belief that incest is wrong, for some people are motivated towards having sex with members of their family and the belief, especially when socially reinforced, will help to combat that motivation. (LRS 2014: 268–9)
But this reply is not fully satisfactory, for the case of pain and pleasure seems to be analogous to that of incest in this respect. Although most of us tend to avoid pain to the extent that we can, some—namely, masochists—don’t. And although most of us seek pleasure, some—namely, ascetics—don’t. This attests that, by themselves, the phenomenologies of pain and pleasure do not always suffice to have us avoid pain and seek pleasure. In light of this, there is no clear-cut contrast between the case of pain and pleasure, on the one hand, and that of incest, on the other. Undeniably, some people are tempted by incestuous intercourse (just as some are attracted by pain or uninterested in pleasure), so that the belief that incest is wrong may have helped them avoid incestuous intercourses. Still, most people feel a strong disgust at the mere thought of such behavior (just as we typically avoid pain and seek pleasure), and this basic repulsion is motivating to quite an extent. LRS’s rejoinder therefore tends to overgeneralize: if our belief that pleasure is good was not shaped by evolution, then neither was our belief that incest is wrong.
Moreover, in focusing on the specific claim that our belief that pleasure is good and pain is bad was selected for—and ignoring the vaguer contention that it was somehow influenced by evolutionary forces—, LRS fail to address the more fundamental point of Kahane’s objection. As far as their rejoinder goes, our belief that pleasure is good and pain is bad could have been shaped by evolution indirectly, without being itself an adaptation. For instance, it might be that evolution selected for a positive attitude to pleasure and a negative attitude to pain, and that our belief that pleasure is good and pain is bad would result causally from these attitudes. This belief would then be an evolutionary by-product rather than an adaptation. Yet, the worry would remain: this would amply suffice to ground a debunking argument, for we would still have acquired this belief via an off-track process—our positive attitude to pleasure and our negative attitude to pain were not selected for because pleasure is good and pain is bad (Kahane 2011: 111–2). Barring evidence that this belief was in no way subject to evolutionary influences, Kahane’s challenge is still awaiting for a solution.
Going Subjectivist
One might think that this leaves another option to utilitarians: going subjectivist. That way, they could escape both the ethical and prudential dilemmas. Indeed, subjectivists can easily accommodate evolution’s influence on our normative beliefs (Street 2006). Naïve subjectivism is a perfect illustration. Assuming that our normative beliefs are beliefs about our positive and negative attitudes, it is very plausible that they are true although they were shaped by evolution. These attitudes are evolutionary adaptations, and our normative beliefs represent them accurately. For instance, our belief that incest is wrong is correct because disapproval of incest was selected for and we know what we disapprove of. If evolution had selected for approval of incest instead, then we would approve of incest. Since we would know what we approve of, we would believe that we approve of incest or, in other words, that incest is right. And this belief would be true.
Kahane is quite pessimistic about this strategy, though. On his view, while UB is “logically compatible with antirealist views, [it is] extremely hard to defend once we accept a broadly Humean antirealist metaethics … on which moral truth is constructed out of our subjective attitudes and sentiments, including, presumably, the sentiments and intuitions underlying much opposition to utilitarianism” (2014: 339). Considering that our attitudes are far from being impartial, if they were to ground moral truths, they would hardly ground a principle as impartial as UB. Subjectivism is much more congruent with first-order moral theories that leave room for special duties than with strictly agent-neutral theories such as utilitarianism (Williams 1981).
Kahane even turns this critique of LRS’s argument for utilitarianism into an objection to utilitarianism. At the end of the day, we have to choose between two options neither of which is compatible with utilitarianism. We can go objectivist about normativity, and we will get UB (which survives debunking because the corresponding belief wasn’t shaped by evolution), but then we will be deprived of pleasure’s goodness (which does not survive debunking because the corresponding belief was shaped by evolution). Or we can go subjectivist, and we will get pleasure’s goodness (which is compatible with prudential truths being grounded in our attitudes), but then we will be deprived of UB (which is incompatible with ethical truths being grounded in our attitudes) (Kahane 2014: 339). The problem is that both UB and pleasure’s goodness are essential components of utilitarianism. So, it would seem that utilitarianism is no longer an option either way.
Although utilitarianism is admittedly incompatible with crude variants of subjectivism, one might reply that it is clearly compatible with more sophisticated versions. Ideal observer theories immediately come to mind. Suppose moral judgments were judgments about an ideal observer’s states of approval and disapproval. On the reasonable assumption that such an observer would approve of actions that maximize well-being and only of those actions, this would mean that an action is right if and only if it maximizes well-being. This seems to disprove Kahane’s claim that utilitarianism is in tension with subjectivism. However, the above assumption is reasonable only if the ideal observer is construed as impartial. Ideally informed and rational or not, a partial observer would not have perfectly utilitarian attitudes. But then, the worry is that impartial observer theories aren’t subjective enough to escape the ethical dilemma the way naïve subjectivism does. Naïve subjectivism can accommodate evolution’s influence on our moral beliefs because it takes these beliefs to be about our attitudes, which are evolutionary adaptations. But this explanation fails assuming that our moral beliefs are about what an impartial observer would approve of. Such an observer’s attitudes weren’t selected for by our evolution, and it is unclear how our belief that what she would approve of is right could have resulted from our own attitudes. In other words, there is a gap between our actual attitudes and those of an impartial observer, and this gap is precisely impartiality. In a sense, impartiality makes impartialist subjectivism too objectivist to escape the ethical dilemma.