Hume on Animal Reason

Hume Studies Volume 29, Number 1, April 2003, pp. 3-28 Hume on Animal Reason DEBORAH BOYLE 1. Introduction Hume's writings contain frequent references to non-human animals; indeed, in both the Treatise and the first Enquiry, he devotes a section to a discussion of animal reason, offering an argument from analogy in which he compares how humans make causal inferences with the way that animals make such inferences.1 With such a comparison, he brings humans into the realm of nature, dubbing their reason a kind of "Instinct," in stark contrast to the Cartesian view of humans as possessing a unique ability to reason that animals lack. To remove humans so far from non-human animals, Hume implies, conflicts with the precept that reasoning concerning matters of fact are (and should be) founded on analogy (EHU 9.1; SBN 104); that is, to suggest that humans are, in their mental processes, radically different from animals is not good reasoning when there are such evident similarities in the ways that animals and humans act.2 Still, although Hume thinks the moral reasoning of humans and animals is similar, his texts do suggest some differences in their reasoning abilities. One such difference is that animals, but not humans, possess particular cognitive instincts designed to help them survive and reproduce; another is that animals do not engage in demonstrative reason; and a third is that humans, but not animals, can improve in their reasoning. In this paper, I discuss Deborah Boyle is Assistant Professor of Philosophy, College of Charleston, 66 George St., Charleston, SC 29424-0001, USA. e-mail: boyled@cofc.edu 4 Deborah Boyle Hume's argument from analogy, and then examine how Hume can hold that there are differences in the reasoning abilities of humans and animals without having to attribute to either a special capacity that the other lacks. As we shall see, Hume holds that good reasoning is a virtue in humans. One further question we can ask is whether he would make the same claim for animals. A number of commentators have maintained that Hume wanted to exclude animals from the sphere of morality.31 will argue that while Hume may think animals lack the ability to make moral judgments, they can still be the subjects of our moral evaluations. Thus Hume can claim that both human reasoning and animal reasoning can be virtuous. Indeed, I will argue that Hume's claims about sympathy and his methodological commitment to empirical observation give him good reason to hold that in animals just as in humans, reason is a virtue. 2. Hume on Human Reasoning We should get clear, first, about what Hume means by "reasoning." David Owen has argued, convincingly to my mind, that by "reasoning" Hume means the imagination's ability to link and relate ideas in various ways that produce either belief (in instances of causal reasoning) or knowledge (when the reasoning is demonstrative).4 Hume says in the Treatise that "reason is nothing but a wonderful and unintelligible instinct in our souls, which carries us along a certain train of ideas, and endows them with particular qualities, according to their particular situations and relations" (T 1.3.16.9; SBN 179). I will return later to the sense in which Hume thinks reason is an instinct; at this point what is important is Hume's characterization of reason as a way of carrying a thinker through a train of ideas. What Hume is rejecting, Owen emphasizes, is the traditional, Cartesian conception of reason as a distinct faculty which operates independently of imagination.5 In both the Treatise and the first Enquiry, Hume identifies two sorts of reasoning, although he does not use the same terms in the two works. In the Treatise, he distinguishes between two groups of relations between ideas, and his division of reasoning depends on what kind of relation of ideas that reasoning concerns. Some relations "depend entirely on the ideas, which we compare together" (T 1.3.1.1; SBN 69). These include the relations of resemblance , contrariety, degrees in quality, and proportions in quantity or number. In such cases, if the relation changes, then the ideas involved must have...