Download Understanding Pain and Pleasure in Animals: A Philosophical Perspective and more Study notes History in PDF only on Docsity!
Animal Rights: A Non-‐Consequentialist Approach
Uriah Kriegel Arizona Abstract/Introduction It is a curious fact about mainstream discussions of animal rights that they are dominated by consequentialist defenses thereof, when consequentialism in general has been on the wane in other areas of moral philosophy. In this paper, I describe an alternative, non-‐consequentialist ethical framework (combining Kantian and virtue-‐ ethical elements) and argue that it grants (conscious) animals more expansive rights than consequentialist proponents of animal rights typically grant. The cornerstone of this non-‐consequentialist framework is the thought that the virtuous agent is s/he who has the stable and dominating disposition to treat all conscious animals, including non-‐human conscious animals, as ends and not mere means.
1. The Consequentialist Case for Animal Rights Perhaps the most influential case for animal rights (see, most notably, Singer 1975) is consequentialist in spirit and relies on two premises. The first is philosophical: the right moral action in every circumstance is that which maximizes pleasure and minimizes pain. The second premise is empirical: that many animals – certainly most animals we have dealings with in our everyday life, including animals used in
the food and clothing industries – can experience pleasure and pain. Let us state the relevant consequentialist thesis as follows: (CT) In deciding what actions to perform, one ought always to attempt to maximize the pleasures and minimize the pains consequent upon one’s actions. For the purposes of a consequentialist defense of animal rights a weaker thesis – taking pain and pleasure to be sufficient conditions for moral relevance, say – may be enough. But (CT) will do for the sake of illustration. As just formulated, this consequentialist case for animal rights is couched in terms of pain and pleasure. One often finds the case put in those terms, but sometimes it is also put in terms of suffering and joy (or enjoyment). There are two ways to understand the relationship between pain and suffering, pleasure and joy. One construes suffering as a deeper, more involved, perhaps more intellectual kind of pain, and correlatively joy as a deeper and more cognitive form of pleasure. The other focuses not on depth but length: suffering is a prolonged, systematic, and stable exposure to pain, joy a prolonged and stable kind of pleasure. In any case, it appears that suffering and joy can be analyzed in one way or another in terms of pain and pleasure. For this reason, I will focus henceforth on the latter. My contention is that the notions of pain and pleasure is ambiguous, and it would follow from this that so are the notions of suffering and joy. The ambiguity can be appreciated through discussion of David Chalmers’ (1996 Ch.1) claim that mental terms typically lead a ‘double life’ – they have a psychological life and a phenomenological life: they can be taken to denote mental states conceived of in terms of their psychological role, or mental states conceived of in terms of their phenomenal character. To conceive of a mental state in terms of its psychological role is to conceive of it as essentially the kind of state that has a certain functional role within the subject’s overall psychology, that is, the kind of state that has certain typical causes and effects. These causes and effects can be restricted to sensory inputs, behavioral outputs, and other mental states, but can
organization of their psychology would be identical to ours. These zombies would have internal states with the same functional role as our pleasures and pains, but would never consciously experience pleasure or pain. In other words, they would undergo psychological pleasures and pains but not phenomenological pleasures and pains. On the face of it, there is no reason to think that causing these zombies the internal states corresponding to our pleasure would be worth the effort and investment, or that causing them the internal states corresponding to our pain experiences would be worth the effort of avoiding. If the zombies do not experience their pleasures and pains, but simply host , so to speak, internal states with the pleasure/pain functional profile, their internal states do not have any moral weight. The point can be brought out in a different way. Consider that it is only when a pain is experienced that it hurts the person undergoing it – that it is only when a pain is experienced that the person is pained by the pain; and likewise, that only when a pleasure is experienced that it pleases its subject. If there are internal states that do not pain or please the persons undergoing them, but which deserve the names ‘pain’ and ‘pleasure’ for some other reason (e.g., due to their functional role), those states do not carry any moral weight. For it is the pains and pleasures of persons , or subjects , that interest us.^3 From this perspective, the key moral difference between phenomenal pain/pleasure and psychological pain/pleasure is that only the former are inherently such as to pain/please their subject. A psychological pain is often associated, as a matter of contingent fact, with a phenomenal pain, and may to that extent can be said to pain its subject. But only phenomenal pain is such as to pain its subject in and of itself. Notice now that since we found psychological pains and pleasures to be morally insignificant, it would seem that the second interpretation of CT is preferable not only over the first one, but also over the third one, that which recommends maximizing both phenomenological and psychological pleasure and minimizing both phenomenological and psychological pain. That is, the moral weight of CT is better captured through the following principle:
(CT+) In deciding what actions to perform, one ought always to attempt to maximize the phenomenological pleasures and minimize the phenomenological pains consequent upon one’s actions. I take this point to be in line with the standard consequentialist defense of animal rights, as can be seen by its common appeal to the notion of sentience (Singer 1975).^4 The term ‘sentience’ is hard to associate with merely psychological or functional states; rather, it intimates a phenomenal, experienced state. Animals may indeed have rights insofar as they are sentient, but sentience is probably associated only with the capacity to be in internal states with a phenomenal character, an experiential dimension – not with the capacity to be in internal states with a functional role.
2. A Non Consequentialist Framework The immediate alternative to consequentialism is of course deontologism, of which the Kantian variety is the best known. For my part, I find the second formulation of the categorical imperative, the ‘humanity formula,’ the most compelling: one should always act in such a way that one treats humanity, whether others’ or one’s own, as an end in itself and not merely as a means to some other end. Perhaps because this formulation of the categorical imperative, and Kant’s moral philosophy as a whole, restrict moral worth to the sphere of humanity, Kantian defenses of animal rights are hard to come by. But it should be straightforward for a proponent of animal rights with Kantian predilections to modify Kant’s dictum to encompass the sphere of the sentient or conscious. Let us call the following the consciousness formula : (CF) One should always act in such a way that one treats conscious creatures as ends in themselves and not merely as means to other ends.
come to mind is the so-‐called simple conditional analysis (Quine 1960): x has the disposition to exhibit manifestation M in circumstances C iff: if C were the case, x would exhibit M. Many philosophers today reject the simple conditional analysis, in favor of either a complicated conditional analysis (Lewis 1997), a non-‐conditional analysis (Fara 2005), or a primitivist account (Molnar 1999). At the same time, the simple conditional analysis has been defended recently as adequate after all (see, e.g., Choi 2008). We need not enter this fray here. Different accounts of disposition will result in different understandings of the disposition to treat conscious creatures as ends and not mere means. It is easiest, for purposes of exposition, to illustrate matters with the simple conditional analysis. According to it, a person has the disposition to treat conscious creatures as ends and not mere means iff the following conditional is true of her: in most circumstances, she would treat conscious creatures as ends and not mere means. So if we use the simple conditional analysis, what (CF+) demands from us that is that we treat conscious creatures as ends and not mere means in most circumstances. Note that this conditional requires that the disposition manifest itself in most circumstances. This is what the qualifier ‘stable and dominating’ is supposed to capture: the greater the variety of types of circumstance in which the disposition is manifested, the stabler the disposition, and the greater the number of times the agent manifests the disposition in each type of circumstance, the more dominating the disposition.^5 Clearly, on the whole the more stable and dominating the disposition, the more commendable the agent, given the kind of sensibilities catered to by (CF+). But it is an interesting question whether our ideal should be that the disposition be totally stable and dominating, that is, manifest itself in every circumstance. On the one hand, the phronimos , or the ideally virtuous agent, surely treats conscious creatures as ends in themselves always and everywhere. At the same time, it appears to be a psychologically contingent fact about sub-‐optimal moral agents such as us that absolute adherence to moral dicta tends to make us dour and intolerant: we run the danger, in our puritanism, of becoming deeply disappointed with humanity (for failing to live up to the moral standards we live up
to) and ultimately adopting a derisive and uncompassionate attitude towards our peers. To avoid these pitfalls, it may be wise for us to demand of ourselves something less than a perfectly stable and dominating disposition.^6 The last clarification we ought to offer is of the phrase ‘ends and not mere means.’ Here too, the existing literature is vast and there is no need, nor hope, for us to make an original and plausible contribution to it here. It bears stressing, however, that not treating someone as a mere means does not entail not treating them as a means. If A treats B both as a means and as an end, A does not treat B as a mere means, in the sense that A does not treat B merely as a means. This is why asking someone for the time does not necessarily involve treating them as mere means, although it may, namely, if the agent does not treat the patient also as an end. This raises the thorny issue of what is involved in treating someone as an end (since it cannot be analyzed in terms of not treating them as a means, given that one can treat a person both as an end and as a means). Presumably, it involves the agent’s keeping in mind, in her dealings with the patient, the right psychological attitude (respect, empathy, love, or what have you) toward the patient. For my part, I am tempted by the view that there is a sui generis phenomenology of treating someone as an end, and that the treating of someone as an end is to be analyzed in terms of the presence of this phenomenology. Another tempting view is that there is a special kind of affective state, which may be called with some qualification love, that is involved in this.^7 Again, different positions on this matter will result in different versions of (CF+). (CF+), so understood, diverges from (CT+) in its practical prescriptions in a number of ways. Perhaps the most striking of these is the fact that, if a conscious animal could be killed without causing pain to it (or anybody else), it is not clear that (CT+) would prohibit the killing. As a matter of fact, conscious animals raised today in the mainstream food industry typically lead miserable lives of ceaseless suffering. However, it is conceivable that some conscious creatures be brought up in quite pleasant circumstances, albeit for the purpose of eventual – but painless – slaughter. (Indeed, it is not implausible that this in fact takes place in some less
defending humane slave owners on the grounds that they offered their slaves a comparatively pleasant life. Even if such slaves worked regular hours and were generously compensated, indeed worked less and were better compensated than they would be as free men and women, the very fact that they would be considered the possession of, and a resource for, their owner is what we would find so morally objectionable. By the same token, the very fact that some conscious creatures are treated as a resource for other conscious creatures, I contend, should raise in us moral indignation, indeed possibly the very same moral indignation. This is the deep motivation for defending animal rights.
3. The Question of Animal Consciousness The non-‐consequentialist framework I have articulated leaves open what is perhaps the most important question from the standpoint of practical ethics: which animals should be treated as ends and not mere means? The reason this is left open is that the above non-‐consequentialist framework is silent on the question of which animals are conscious or sentient. The question is of course empirical, and so the role of the philosopher in answering it is limited. One role s/he can adopt, however, is methodological: to offer the general form of a procedure for generating a reliable conjecture concerning the extent of consciousness within the animal kingdom. The procedure I would like to propose is fairly straightforward: we should seek the neural correlates of consciousness/sentience in creatures that are unquestionably conscious, then attempt to establish which other creatures, if any, exhibit neural structures relevantly similar to these. Several clarifications are in order. First, which animals are ‘unquestionably’ conscious depends on what we take ‘unquestionably’ to mean. I suggest that the best understanding of ‘unquestionably’ is as follows: it is unquestionably the case that p just in case any claim that ~ p would not be taken seriously within the relevant community of inquiry (where this could
be operationalized in terms of publications in peer-‐reviewed journals, presentations in academic gatherings, or some such indicators). In the present case, then, a creature is unquestionably conscious just in case the claim that it is not conscious would not be taken seriously within the cognitive-‐scientific community. Since as a matter of fact even Daniel Povinelli’s claim that chimpanzees are unconscious (see Povinelli et al. 1994) has been taken seriously, but no claim that normal adult humans are unconscious has ever been taken seriously, I suggest that we consider normal adult humans as our paradigmatically unquestionable instance of a conscious creature.^10 Secondly, what makes a brain structure ‘relevantly similar’ to another is also problematic, especially in the context of cross-‐species comparison. Neuroanatomical comparison, in particular, may be neither reliable nor feasible. Better to focus on comparisons of structures as functionally individuated. Thus, one brain region may be different in terms of its spatial location from another and yet qualify as a similar brain structure , namely, if the two regions perform the same neural function. Thirdly, the fact that a brain structure is relevantly similar to one that underlies consciousness does not guarantee that it too underlies consciousness, and likewise the fact that it is relevantly dis similar to one that underlies consciousness does not guarantee that it does not underlie consciousness. However, such relations of relevant similarity and dissimilarity do offer defeasible evidence of underlying consciousness (or lack thereof), by supporting an inference to the best explanation to that effect. Such an inference makes its conclusion probable , though it does not guarantee its truth. Thus if brain structure S 1 underlies consciousness in creature C 1 , where C 1 is unquestionably conscious, and structure S 2 is relevantly similar to S 1 , then it is probable that S 2 underlies consciousness in C 2 ; and if S 2 is relevantly dissimilar to all brain structures of all unquestionably conscious creatures, then S 2 probably does not underlie consciousness in C 2.^11 Incorporating these clarifications into our understanding of the empirical procedure I described leads to the following methodological principle:
argument that a scientific consensus emerged that only mammals are conscious. Our credence in this scientific proposition would not reach 1 at that point. Suppose further that, being of a critical bent, and keeping in mind all the hidden social, methodological, and philosophical assumptions that go into constructing a consensus, our credence in the scientific consensus being strictly true is .55. It follows that our credence in any implication of the scientific consensus, including for animal rights, cannot exceed .55. This is one way in which a more attenuated picture is forced on us by consideration of probability. However, the objection before us pushes beyond this. The objector claims rather that if we have a credence of .05 that oysters are conscious, we ought to have credence of 1 that oysters ought to be treated as ends and not mere means. Put this way, I think the objector’s claim is too strong to be plausible. According to panexperientialists, everything in the universe has a degree of phenomenal consciousness, if only a very small one. Since their position is coherent, and some of the arguments for it are not altogether ludicrous (Chalmers 1996), I have .01 credence that vegetables are phenomenally conscious as well. Surely, however, I ought not to treat vegetables as ends in themselves on the grounds that panexperientialism is coherent and non-‐ludicrous. I take this to be a reductio ad absurdum of the strong claim that any non-‐negligible probability of consciousness entails treating as an end: the result of adopting that claim is to treat everything as an end. It would, of course, be entirely arbitrary to declare that although a non-‐ negligible probability of consciousness does not secure rights, a probability of .08, or .12, or some such figure, does. More generally, there is something wrong about the idea that there is a cut-‐off point beyond which credence in an animal’s consciousness should entail a credence of 1 in the animal having the rights of conscious animals. A more sensible model would embrace a non-‐trivial function from credence in consciousness to credence in rights. The dogmatic model we have worked with thus far matched .12 credence in an animal’s being conscious to. credence in it deserving the rights of conscious animals. The objector’s model
matches .12 credence in the animal’s being conscious to a credence of 1 in it deserving the rights. Neither is plausible. It is much more reasonable, it seems to me, to adopt a function that, say, matches .12 credence in consciousness to. credence in rights. Obviously, we cannot expect that there be objective facts about the ‘correct’ function. But I hope that this illustration makes clear what I think is correct about the spirit of the objection before us. What is correct about it is that the dogmatic function from credence in consciousness to credence in rights is wrongheaded: the moral cost of making a moral error is such that a more ‘proactive’ function should be embraced. One way to think of the ‘correct’ function issue is in terms of the stability and dominance of one’s disposition to treat animals as ends and not mere means. As noted above, the more stable and dominant her disposition, the more morally commendable the agent is (modulo the mentioned worries about puritanism). We can incorporate the considerations we have been discussing in the present section into this framework: mutatis mutandis , an agent who treats an animal in whose consciousness she has .12 credence as an end and not mere means 40% of time is more commendable than an agent who treats such an animal as an end 30% of the time and less commendable than one who treats it as an end 50% of the time. Likewise, an agent who treats an animal in whose consciousness she has. credence as an end and not mere means 40% of time is mutatis mutandis more commendable than an agent who treats an animal in whose consciousness she has .17 credence as an end and not mere means 40% of the time, and less commendable than one who treats in such a way an animal in whose consciousness she has. credence. This is because of variation in the stability and dominance of the key disposition in these various agents. As warned above, this discussion makes free use of oversimplifying toy examples. The lived reality of moral conduct is not as clean-‐cut as the assignment of precise figures in the above discussion suggests. I have adopted this way of discussing the objection merely as an expository device, by way of showing what I think is right about the objection and what I think is wrong. It is right that the mere
Hempel, C.G. 1945. “Studies in the Logic of Confirmation.” Mind 54: 1-‐26. Kriegel, U. 2009. Subjective Consciousness: A Self-Representational Theory. Oxford: Oxford UP. Lewis, D.K. 1997. “Finkish Dispositions.” Philosophical Quarterly 47: 143–158. Molnar, G. 1999. “Are Dispositions Reducible?” Philosophical Quarterly 49: 1–17. Povinelli D.J., A.B. Rulf, and D.T. Bierschwale. 1994. “Absence of knowledge attribution and self-‐ recognition in young chimpanzees ( Pan troglodytes ).” Journal of Comparative Psychology 108: 74 – 80. Quine, W.V.O. 1960. Word and Object. Cambridge MA: MIT Press. Shapiro, L. 2008. “How to Test for Multiple Realization.” Philosophy of Science 75: 514-‐525. Singer, P. 1975. Animal Liberation. New York: Random House. Tiberius, V. 2008. The Reflective Life: Living Wisely with Our Limits. Oxford and New York: Oxford University Press. (^1) At the background is a general distinction between two conceptions of mind. The psychological conception of mind characterizes mental phenomena third-‐personally in terms of their causal relations to each other and to the environment; the phenomenological conception characterizes them first-‐personally in terms of their phenomenal or subjective feel. The former focuses on the mechanical dimension of mental life, the latter on its experiential dimension. Both dimensions are real and central to the nature of mind, so there is no substantive question here of which conception is ‘more accurate’ – both are important. (^2) These classification criteria not only are distinct, but also do not coextend. Suppose a subject experiences a toothache, but upon noticing a fire in the kitchen starts attending to the fire and is thus too distracted to experience the toothache, which unfortunately reappears as soon as the kitchen drama is over. Should we say that this subject had a single continuous toothache, which went unconscious at some point and was then consciously experienced again, or that the subject had two distinct pains, separated by a painless interval? The answer, of course, is that we should say both, but in different senses of ‘pain.’ In the psychological sense, there is only one pain. This is because there is unconscious pain – a mental state which is not experienced by its subject but which nonetheless plays the pain role in the subject’s psychology. In the phenomenological sense, by contrast, there are two pains, and there is no such thing as non-‐experienced pain – as soon as the early pain is phenomenally extinguished, it goes out of existence, and a new one reemerges when the phenomenology of pain reappears later. (^3) Sometimes the term ‘person’ is used to refer to any subject of conscious experience, or any subject of moral worth, whereas on other occasions it is used to refer only to human subjects, or subjects with relatively sophisticated inner life. I use it here in the former sense, and add ‘or at least of subjects’ for the benefit of those who insist on using it the latter way. This is just a terminological matter. (^4) This is not to say that Singer is committed to the entire picture of the nature of pain and pleasure that we have developed thus far. (^5) For a related distinction between a trait being broad and a trait being stable, see Tiberius 2008.
(^6) It is an intriguing question what this would imply in practice. One possibility is to force ourselves to treat conscious creatures as mere means on occasion. Just how often ‘on occasion’ should be ought to be determined by psychological research on the forming of the dour and intolerant attitude of which I spoke. It may well be that such research will discover that my observation is simply wrong (perhaps it is based on the wrong anecdotes), or that it can be avoided by simpler means. Whatever such research turns up would have to be integrated into the Kantian virtue-‐ethical stance I describe here. (^7) These two views are, moreover, clearly compatible. Unfortunately, and somewhat surprisingly, there is relatively little work on the analysis of treating someone as an end, even among Kant scholars. Moreover, it would seem that Kant himself had little to offer on this matter, so that many interpretations are in principle open (see Griffin 1986). But the Kantian proponent of animal rights could just suggest here that whenever some consensus emerges regarding the correct analysis of this matter, it could be plugged into (CF+). (^8) At the same time, note that (CF+), like (CT+), does not prohibit eating, say, cows altogether: a cow dead of natural causes, since it is no longer conscious, has no rights, and can be eaten. (This is, by the way, the current practice in Bhutan.) (^9) Thus, if combined with the view that all animals are conscious, (CF+) would prescribe not only vegetarianism but veganism. For reasons explained earlier, I think it should only be combined with the claim that all mammals are conscious, and so veganism does not follow: we are most certainly justified in consuming honey, for instance, since bee brains most certainly do not engage in higher-‐ order monitoring (and if I am right, the same holds also for the products of fish and chickens). (A qualification is necessary here, however. My claim is that using the products of these unconscious animals is not intrinsically wrong. But there may be ways in which it is instrumentally wrong. For example, gratuitous torturing of an unconscious animal that elicits in us the intuitive ascription of consciousness is sure to do harm to our character, and therefore such torturing is morally wrong, though only instrumentally. Perhaps the treatment of geese in the context of foie gras production is a good example: it is not intrinsically wrong, on my view, since geese are probably unconscious, but it is instrumentally wrong, in that it desensitizes us to what would be cruel treatment if geese were conscious.) At the same time, according to (CF+), commercial products involving the products of mammals are morally tainted one and all: leather jackets, shoes, and belts, fur coats, ivory sculptures, etc. are all to be avoided. (^10) The term ‘seriously’ is problematic, inasmuch being taken seriously comes in degrees. Some philosophers – eliminativists – have argued that nobody is conscious, not even humans (Churchland 1984). This has always been a marginal view, however, in a way Povinelli’s view has not. I recognize, however, that there may be an element of artificiality in deciding what we should take as unquestionable cases of consciousness. Perhaps all apes could be taken as unquestionably conscious. (^11) It might be objected that while the presence of the right brain structure clearly provides evidence for the presence of consciousness, the absence of the former does not provide evidence for the absence of the latter. As it is sometimes put, absence of evidence does not provide evidence of absence. My response to this is twofold. First, for reasons I cannot go into here, and which have to do with Hempel’s (1945) diagnosis of the so-‐called Raven Paradox, I think that absence of evidence is evidence of absence, albeit in some circumstances comparatively weak evidence. Secondly, whether the absence of the right brain structure provides strong evidence for the absence of consciousness is crucially tied up with the question of whether consciousness is multiply realized across the animal kingdom. The question here is not the philosophical question of whether consciousness is multiply realiz able , but the empirical question of whether it is really multiply realiz ed in the actual world. Following a number of philosophers, while I am sympathetic to the philosophical claim of multiple
