I recently had the pleasure of reading an interesting paper by Neil Sinhababu, a friend I met while I was a visiting scholar at the University of Texas. Neil’s paper, thoughtfully posted online, is entitled The Epistemic Argument for Univesal Hedonism. In it, Neil makes an argument for a strong and controversial position that I’ve flirted with before myself: that pleasure and displeasure are the only things intrinsically good or bad in any ethical sense.
Neil’s argument proceeds roughly as follows (and this summary, qua summary, must necessarily leave out some of the detail and precision of his argument): Ethical judgement all derive from one of two sources: emotional perception and phenomenal introspection. The source of most of our commonsense judgements about morality is emotional perception: a process by which we react emotionally to states of affairs in the world, form moral judgements in connection with these emotional reactions, and thereby perceive the states of the world as having objective moral qualities. Neil draws on Jonathan Haidt’s empirical research to support this point.
Neil goes further, however, in arguing that we are wrong to make moral judgements on the basis of emotional perception, thus rejecting Mencius’s metaethics as well as those of the moral sense theorists. Emotional perception, he claims, is inherently unreliable. Here he draws on the work of Todd Stewart, in order to argue roughly as follows: There is widespread disagreement about the morality of many behaviours (sexual practices, intoxicants, gender roles, vengeance, etc.) These divergent views have generally been arrived at through emotional perception. If emotional perception were a reliable way to achieve knowledge about morality, then we would expect that each of these divergent views would be justified. But on an issue where many positions contradict each other (as many of these moral issues seem to be), only one can be correct; the others must be false. And if emotional perception produces this many false beliefs, it cannot be genuinely reliable.
By contrast, phenomenal introspection – “a process of looking inward at one’s subjective experience, or phenomenology, and determining what it is like” (21) – is generally reliable. While it is occasionally fallible, it is only so in the way that vision is occasionally fallible; like vision, and unlike emotional perception, its conclusions do not diverge so widely as to make it unreliable. Conclusions about colour (such as “yellow is bright”) are one example of these conclusions. But the conclusions significant to the argument are: pleasure is good, and displeasure is bad.
I wonder if there’s a certain amount of tautology in Neil’s phenomenal account: he identifies pleasure as “felt goodness,” which suggests that the goodness of pleasure is analytic rather than synthetic, providing us only with a definition and not a fact. This definition seems to need work. But I’d consider that point a relatively minor objection.
I find a potentially bigger problem in the distinction between emotional perception and phenomenal introspection. I’m not sure how much these really differ from each other. What are pleasure and displeasure, first of all, if not emotions? Neil doesn’t define “emotion” in the paper, so the paper has no way of distinguishing the two. If pleasure is “felt goodness,” it sure sounds like one of our emotions, our “feelings.” Then while phenomenal conclusions about colour are not necessarily accompanied by emotion, phenomenal conclusions about pleasure are – because they derive from emotional states (namely pleasure and pain) that we experience. So when the paper says that phenomenal introspection tells us that pleasure is good “independently of any emotional responses we might have to our pleasure” (21) – it might be technically true in that it’s independent of emotional responses to our pleasure, but that doesn’t seem to matter if it’s dependent on the emotional response that is our pleasure. Some clarification on this matter would help the paper a lot.
But let us suppose that the distinction between emotional perception and phenomenal introspection does hold up. Where Neil’s argument most goes astray, I think, is in identifying emotional perception and phenomenal introspection as the only sources of knowledge about ethical facts. He acknowledges at the end (41-2) that there might theoretically be other such sources, noting correctly that Kant saw his own moral philosophy as deriving from something other than these two. But since he believes that Kant’s arguments for such a source fail, he doesn’t see Kant’s philosophy as any reason to identify another such source.
I also disagree with Kant’s moral philosophy; but there’s a weaker form of Kant’s arguments (found to some extent in Plato and Augustine) which I think does hold true, and does give us ground to believe in an alternative source of ethical facts. Kant aims to derive his moral philosophy entirely from the logical requirements of reason; it is simply illogical to follow anything other than the categorical imperative. He would claim that “pure practical reason” is a reliable source of ethical knowledge (and that neither emotional perception nor phenomenal introspection is reliable in this way). While I don’t buy the whole structure Kant puts up, I agree with him that there’s at least one kind of ethical fact we can derive from pure practical reason in this sense: truth is an intrinsic good, a good in itself, independent of pleasure (or happiness).
My argument here follows the same foundationalist line of reasoning I referred to in the post on certain knowledge. Thought must be possible, for to doubt its existence is self-contradictory. So likewise, truth must be intrinsically good. To make any claim at all – about pleasure, ethics or anything else – is to make a claim to its truth, and also a claim that one’s audience should believe the claim. In order to deny that truth is a good, one must assume that truth is a good; if one doesn’t assume that truth is a good, one has no reason to be convinced by that denial or by any other claim.
Now it’s a trickier matter to say what follows from the goodness of truth, what kind of concrete ethical view is implied by this claim. It would suggest that lying and promise-breaking are bad at least if they don’t have some other good as an end – even if they have no net effect on our pleasure or displeasure. Beyond that, it doesn’t say anything about how we rank-order the goods of truth and pleasure. But sticking for the moment to the context of Neil’s paper, we don’t yet need to go there. The point is, there’s more to life (and ethical goodness) than just pleasure.
This point is especially important with respect to the section of Neil’s paper that deals with asceticism (33-4). Contrary to the paper, ascetic suspicion towards pleasure doesn’t have to arise merely from emotional perception. On Augustine’s account, God himself is truth; pleasure is problematic because it pulls us away from this truth. I don’t buy Augustine’s arguments here, but as with Kant, the claim for such an asceticism rests on something beyond either emotional perception or phenomenal introspection.