May's (Reference May2018) Regard for Reason in the Moral Mind is a novel and important defense of the view that reason guides moral thought and motivation in human beings. Although rationalist views of moral psychology have many defenders, few engage in as much detail as May with empirical arguments from situationists, egoists, and Humeans. If rationalists avoid these challenges, they face the criticism that although their theories describe a possible moral psychology, it is not the one that human beings have. May is not afraid to get his hands dirty with the empirical data, and much of his book responds in detail to his opponents’ empirical arguments.

Although I doubt situationism and reject egoism, I defend the Humean theory of motivation. May and I agree on how the Humean theory should be formulated: It includes commitments both to the necessity of desire for motivation, and to the impossibility of generating new desires by reasoning from beliefs alone. Desire, then, is not merely an immediate motivator of action that reason can summon up on command. It is the fundamental source of all motivation, and new motivation cannot be generated without it.

Formulated this strongly, the Humean theory is incompatible with the view that moral judgments are beliefs with intrinsic motivational (or desire-generating) force. Creatures with Humean psychologies cannot make moral judgments that fit this cognitivist and internalist model. May regards moral judgments as beliefs that can generate desires this way, and therefore must reject the Humean theory.

In Humean Nature (Sinhababu Reference Sinhababu2017), I argue that the Humean theory is part of the best explanation of how we think, feel, and act. Desire does not just motivate action. It causes pleasant and unpleasant feelings when we have various sorts of thoughts about its object, and it directs our attention toward its object in various ways. Because of its emotional and attentional effects, desire is well-suited to explaining the thoughts and feelings that arise in practical deliberation and various other phenomena like procrastination and daydreaming.

May responds to my arguments at length after describing me as “the best ‘philosophical nemesis’ one could ask for” (p. xii) in the preface. Here I will try to live up to his praise by defending the Humean theory against three different lines of argument he makes against it.

First, May argues that the Humean theory runs against commonsense explanations of human motivation that we often rely on. He cites examples of people who describe their own moral motivation as the result of a belief that something is the right thing to do. Then he argues that “We often describe one another, and ourselves, this way – as ultimately motivated by beliefs with normative or evaluative content” (p. 180). On May's view, the content of such beliefs enables them to generate new desires to act accordingly, violating the Humean theory. I agree with May that it would be a problem for the Humean theory if our intuitive folk-psychological theory was committed to the possibility of this kind of fundamentally belief-driven moral motivation. Folk psychology may not always be right, but it works well enough that there is a cost to denying its core commitments. And if people who explain their moral motivation in terms of their moral beliefs really were insisting on an anti-Humean explanation of their motivation, Humeans would have a problem.

Fortunately for Humeans, there is no reason to see “I did it because I believed it was the right thing to do” as an anti-Humean explanation of motivation. The belief may have played its motivational role only by combining with a desire to do the right thing. People often explain things by pointing out one particularly salient explanatory factor, and in doing so they are not denying the presence of other explanatory factors. If someone explains that she did not eat the mushroom because she believed it was poisonous, she is not committing herself to an anti-Humean psychology where beliefs about poison have intrinsic motivational force. It is perfectly consistent with her explanation that a pre-existing desire not to be poisoned combined with her belief and motivated her not to eat the mushroom. When we explain things to each other in ordinary conversation, we do not usually name all the causal factors – that would take too long and bore our audience. We name some, and let our audience infer the others. If that is all people are doing when they mention only their moral beliefs in explaining motivation, such explanations leave plenty of room for desires to do the right thing, and thus provide no evidence against the Humean theory.

Second, May argues that many mental states other than desire can produce the phenomena that I credit the Humean theory with explaining. I take desire to motivate action, cause pleasant and unpleasant feelings, and direct attention, with all these effects taking greater magnitude when the desire's object is vividly represented. My argument for the Humean theory is that desire, so conceived, provides the best explanation of a variety of psychological phenomena. May responds that many other psychological states can do these things as well (pp. 192–95). He argues that habits can cause behavior without amounting to desires as I conceive them, sensory stimulation that does not involve desire can generate pleasant or unpleasant feelings, attention can be directed by psychological associations that are not grounded in desire, and vivid representations of obscenities can make a Tourette's sufferer more likely to tic by saying the obscenities. I accept much of this. If other mental states have many of the effects that desire does, why are Humean explanations of our action, thought, and feeling superior to anti-Humean explanations?

My answer is that Humean explanations provide a better-integrated explanation of motivation, feeling, attention-direction, and the effects of vividness than anti-Humean explanations do. Suppose we are trying to explain why a very hungry person was pleased to be told he would soon be served a delicious meal, why his hunger prevented him from paying close attention to the boring dinner-table conversation around him, why he became especially excited to eat when the food was brought out before him, and why he ate with enthusiasm. Desire for food has the hedonic, attentional, vividness-related, and motivational effects to explain all of this at once. The other sources of these effects that May discusses would not fit into such a well-integrated explanation. Should we posit a sudden pleasant sensory experience when he was offered food, a set of psychological associations distracting him from the dinner-table conversation, something like the ticcing of a Tourette's sufferer when the food is brought out, and a habit of eating whatever is on a plate before him? Often we will not have any reason to posit such a disconnected hodgepodge of psychological factors. The fact that the things we are motivated to pursue attract our attention and cause pleasant and unpleasant feelings, and that all of these effects are amplified by vivid representations, is best explained in terms of a unified psychological state with all these effects. We cannot count on the other factors May cites to come together frequently enough to explain the phenomena.

Third, May criticizes my frequent appeals to parsimony. I claim that the Humean theory is more parsimonious in treating instrumental reasoning, where a desire for an end combines with a means-end belief to generate a desire for the means, as the one and only way that reasoning can generate a desire. Continuing a debate that we have had in previous work, May argues that the Humean theory may not actually be more parsimonious. As he notes, I am individuating processes by their relata. This makes instrumental reasoning (where the relata are a desire for an end, a means-end belief, and a new desire for the means) a different process from anti-Humean desire generation (where the relata are a normative belief and a new desire for the normatively favored course of action). He argues that we do not always individuate processes this way: “we don't posit two kinds of baking or two kinds of corrosion just because the relationship can hold between different entities. A human or a robot can bake a cake (or a quiche); water or acid can corrode a pipe (or a rock) […] We needn't posit two kinds of motivational process just because one is initiated by a desire while the other is initiated by a belief” (p. 197). The upshot is that Humean theory is not more parsimonious than opposing views, as instrumental desire-creation and anti-Humean desire generation can be treated as the same process. We just have to give up the assumption that processes are individuated by their relata.

Here I will defend the idea that processes are to be individuated by relata, and that May's baking and corrosion examples do not provide good analogies to the psychological issues at hand, because the differing relata he mentions are not essential to characterizing the processes. The reason we might not divide up baking into separate processes depending on whether the baker is a human or a robot, or the specific type of food, is that these are not essential to characterizing baking. What makes something an instance of baking (rather than say, frying or applying no heat) are a general way of applying heat and general sorts of effects on the food, not the identity of the baker or the precise nature of the dish. Similarly, what makes something an instance of corrosion is the interaction between metallic particles and ions, not the specific nature of the substance that contains the metallic particles or the liquid that contains the ions. Once we are sufficiently precise about the nature of the processes, we see that we do individuate them by their relata.

Maybe there is some more general level of explanation on which we could likewise treat instrumental reasoning and anti-Humean desire generation as instances of the same general process of reasoning, with the relata being very general – perhaps, “some antecedent psychological states” and “a new desire.” But admitting these general relata takes us away from the level on which my psychological debate with May is being conducted. We are advocating psychological explanations that invoke different intentional states, and the fact that the differing relata we invoke could be lumped together at some other level than psychology that invokes specific intentional states is neither here nor there. To illustrate the point: Is reasoning the same process as telepathy, because both are ways for psychological states to affect other psychological states? Maybe one could find an explanatory level where generalizing the relata and lumping these processes together makes sense. But at the level of psychological explanation, we need to distinguish these processes, and admit reasoning while rejecting telepathy. Similarly, psychology needs to distinguish between instrumental reasoning and anti-Humean desire generation. Whether we should believe only in the former, or also admit the latter, is at the heart of the debate between Humean Nature and Regard for Reason in the Moral Mind. To have that debate, we need to recognize the differences between these processes, rather than treating them as the same.