Categorical Desires and the Badness of Animal Death

Granted, there might be other reasons not to slaughter animals—e.g., environmental considerations for domestic animals, and worries about animal families for wild ones—but such concerns won’t occupy us here.

The parenthetical qualification is designed to distinguish those cases that do and do not give us reason to prevent death. So, for example it might well be the case that death is bad for plants, but presumably it isn’t bad in the way that matters, which is to say that we have no intrinsic reason to prevent the deaths of plants. Should they serve some agricultural, aesthetic, or similarly important end, then saving plants might make perfect sense. The point is just that, apart from such reasons, their deaths may well be bad for them, but not in a way that provides us with reason to act.

For instance, it’s no good to criticize Belshaw, as Bradley (2016) does, based on the odd consequences of a categorical desire account of the badness of death. Belshaw has received this criticism before, and now accepts the consequences that Bradley regards as counterexamples to the view—e.g., that it isn’t bad for babies to die. “I claim it is not bad, in the way that matters, for babies to die. Even when, predictably, their future life will be good and, again predictably, they will let some time want very much to continue with it, this life isn’t something in which they have any interest now” (2015, p. 14).

See Varner (2012) for discussion. Note, however, that in his treatment of animal cognition, Varner makes a significant leap in ascending from more to less basic forms of animal cognition, skipping from sentience right over to autonoetic mental states, and not placing much emphasis on desires.

See note 3.

Belshaw borrows the example from Bradley (2016), who thinks we ought to say the cow’s desire is about its future, but doesn’t explain why.

Thanks to an anonymous reviewer for helping us see the connection between these desires and what makes a chicken’s life worth living.

Note that teleosemantics is not the only forward-looking or benefit-based theory of mental content. The pragmatist approach to semantics championed by C.S. Peirce and William James is another version of the idea. Very roughly, and borrowing James’ words, the pragmatist position is that “to develop a thought’s meaning, we need only determine what conduct it is fitted to produce” (James [1907] 2000, p. 25). There is also the success semantics of Whyte, according to which “[t]he truth of beliefs explains the success of the actions they cause” (Whyte 1990, p. 149). You can take these views either as competing with (Blackburn 2005) or complementing (Papineau 1993) Millikan-style teleosemantics. You might prefer those approaches, depending on your philosophical sensibilities. If the appeal to biological function in teleosemantics seems dubious to you, or if you’re skeptical of its reductionist aspirations, you might opt for a version of our story retold in terms of these theories. We see no reason why they would fail to apply just as well to the cases of animal behavior we’ve discussed. Bermudez (2003), to cite a prominent example, accounts for mental content in non-human animals with a version of success semantics. We should also note that if you are skeptical about attributing mental content to non-human animals because you think content only comes packaged together with language, our story can be re-fashioned to accommodate this. Intentional states like belief can be decoupled from propositional (or conceptual) content and still plausibly be attributed to animals (Glock 2013). Moreover, Daniel Hutto (2007) and Erik Myin (Hutto and Myin 2013) defend a Millikan-esque “teleosemiotics” that could easily fill the role that teleosemantics plays in our story. It would explain the intentionality of mental states, but without construing them as having content (i.e., accuracy conditions) of any kind.

To take this line is to endorse a version of the precautionary principle. Some versions of the precautionary principle can get us into serious trouble: if we start attributing categorical desires to insects, for example, we will probably need to radically revise our ideas about what we ought to do—see Fischer (2016) and Sebo (ms). However, since we rely on precautionary reasoning in so many contexts, we think that it’s safe to appeal to it here, and we’ll have to bank on their being some way of drawing a line between reasonable and excessive levels of moral caution. Moreover, note that our argument here only commits us to saying the deaths of animals can be bad for them—not that any particular practical conclusion follows. As an instance of precautionary reasoning, ours is a fairly modest one.