Reason and Feeling in Hume's Action Theory and Moral Philosophy Daniel Shaw 1. The Slave Metaphor The statement in the Treatise which follows Hume's arguments about why reason alone can neither produce nor prevent action is surely one ofthe most famous or, I should say, notorious opinions to be found in Hume: Thusitappears, thatthe principle, which opposesourpassion, cannot be the same with reason, and is only call'd so in an improper sense. We speak not strictly and philosophically when we talk ofthe combat ofpassion and ofreason. Reason is, and oughtonly tobe the slave ofthe passions, and can never pretend to any other office than to serve and obey them.1 That really does demote reason to a very subservient position in ethics. It is in this passage that Hume overstates his case against ethical rationaUsm. For this final claim about reason being the slave ofthe passions goes well beyond what the preceding arguments, even ifsuccessful, have established. Those arguments, ifsound, would show that reason alone is not sufficient to produce or prevent action; that over and above reasoning, desire ofsome kind is also necessary. But in arguing for that conclusion Hume nowhere denies that reason is also necessary to produce or prevent action; that is, he gives us no grounds for thinking that desire without reason could produce or prevent action any more than could reason without desire. Indeed, from what he has said about the roles ofa priori and causal reasoning in deliberation, it seems hardfor him to escape the conclusion thatreason is as necessary as is desire for motivating action. How could someone ever act in order to achieve some purpose if he never engaged in causal reasoning concerning what action on his part would result in the achievement of his desired end? Or, to take one of Hume's examples, how could a merchanteverpayhis debtsifhe never engagedin a priori arithmetical reasoning to calculate his debts? On Hume's own story, and in light of Hume's own examples, reason appears just as necessary for action as does desire. What right has Hume to relegate it to the role of slave of Volume XVIII Number 2 349 DANIEL SHAW passion? Why not simply treat it as of equal importance alongside desire in producing action? One reason Hume treats desire as more important than reason is that he sees it as the original source of our motivation, whose original impulses are only subsequently directed by a priori and causal reasoning. On Hume's model, the merchant begins with a desire to settle his debts. Only afterwards does he employ a priori arithmetical reasoning to work outhow to do it. So, too, the agent who reasons about cause and effect begins with the desire for some end and only afterwards goes on to reason about the means. That then is one point that I think Hume has in mind when he treats desire as offirst priority; desire comes first in time. The secondreason thatHume treats desire asmore important than reason is as follows: since, on Hume's model of motivation, the desire does come first in time, it does not depend on any prior reasoning process. By contrast, the reasoning which follows after the desire not only follows after, but also depends upon the prior desire; for example, had the merchant never desired to pay his debts in the first place, he would never have gone on to do his arithmetic at all—that is, the a priori reasoning depends for its very existence on the prior existence of his previous desire. Had the agent never conceived a desire for his objective, the subsequent causal reasoning would never have come into being. These two claims then, the claim about the temporal priority ofthe desire to the reasoning, and the claim about the dependency of the reasoning on the desire, are surely the main points of Hume's slave metaphor, ofHume's claim that reason is the slave ofthe passions. Thus understood, is that claim true? Does causal reasoning about means to an end always follow after and depend upon the prior existence of a desire for that end and never the other way round? Imagine the case...