[1.1] The subject of environmental ethics has rekindled the debates about the scope of morality, the basis of moral considerability, and the location of intrinsic value. Sometimes it has done so by serving as a challenge to its opponents. Thus in her recent essay "A Refutation of Environmental Ethics," which rejects environmental ethics (in any values-in-nature version) as nonviable, Janna Thompson was constrained to suggest at the same time a more viable account of ethics and of the good life (Thompson 1990). But I have discussed that essay elsewhere, and do not want to traverse old ground here. 1 More positively, environmental ethics in the person of Kenneth Goodpaster has pointed out that, granted the centrality of beneficence and non-maleficence in ethics, widely recognised both by environmental ethicists and others, it becomes important to identify which entities have interests, and a good of their own; for entities with a good of their own or with interests will plausibly comprise the class of possible beneficiaries, and thus the class of things which ought to receive moral consideration. 2 At the same time such non-instrumental benefits as they can receive are strong candidates for recognition as bearers of intrinsic value.
Abstract The view is defended that, unlike artefacts, living creatures have a good of their own and are morally considerable. In response to Frederik Kaufman's claim that the capacity for preferences is necessary for having interests and for moral considerability, it is argued that in these regards this capacity is sufficient but unnecessary, and that the capacity for health is also sufficient. This has the important implication that there is more to human good than mental states and their objects.
[1.2] These issues are newly raised in Frederik Kaufman's challenging essay "Machines, Sentience and the Scope of Morality" (Kaufman 1994). Kaufman argues there that if nonsentient creatures such as trees have interests and qualify for moral consideration, then so do machines; and proceeds to maintain that having preferences (or being capable of having them) is necessary for having interests, and thus for moral considerability. I shall argue, in reply, that having preferences is sufficient but not necessary for having interests, that living creatures have a good of their own in a sense which does not apply to artefacts, that mentality is not necessary for moral considerability, and that our understanding of intrinsic value in the lives of humans is enriched once such value is seen not to be confined to preferences or their satisfaction. Thus not only is the scope of moral considerability at stake, but also the content of value-theory in general.
[2.1] As Kaufman says, much depends on how we understand what it means to have an interest (Kaufman: 59-61). If, like Kaufman, "we understand interests in the sense of what someone or something desires (or would desire under idealized conditions)," then moral concern will only be appropriate towards beings capable of desires. But if, as a number of environmental ethicists maintain, having interests is "to be construed in the sense of having a good or a well-being," then many nonconscious things will warrant moral recognition.
[2.2] Kaufman proceeds to quote a number of writers, including myself, who accept the moral considerability of living creatures but deny such considerability to machines; and maintains that "The assumption at work in the responses of these philosophers seems to be that only beings with ends of their own can have interests (and hence moral standing); and since machines do not have ends of their own, they have no interests" (Kaufman: 63). He then suggests that the difference between having ends of one's own and having derivative ends cannot be crucial; and thus that machines, which themselves appear to have ends, cannot on this basis be denied interests or thus moral considerability (Kaufman: 63-64).
[2.3] Kaufman's account of what is involved in saying that machines or natural systems have ends is that this is a convenient "but in principle eliminable" way of describing their behaviour, one which, however, can be abandoned in favour of purely mechanical explanations. This abandonment occurs much more readily in the case of machines, whose builders tend to understand the mechanisms involved, and which no one thinks really to have ends, and much less readily with natural systems, where our ignorance makes talk of ends remain irresistible (Kaufman: 64-65). But there is no real difference, for, Kaufman asserts, "plants and other simple forms or life (as well as ecosystems, and even people on some views), are mechanisms too...," for all that people including environmental philosophers ascribe ends to them (Kaufman: 64). Indeed he claims that "The goal-directed behavior of living things... results from the ordinary causal interaction of various constituent subsystems..." (Kaufman: 64).
[3.1] Here it may well seem that even to contemplate the very possibility that human beings are mechanisms begs large questions, since on many views freely chosen actions are precisely those which are not in all respects mechanically determined, and the very concept of mechanism makes sense in contrast with behaviour which is not, without remainder, mechanically explicable. If, further, all goal-directed behaviour is in every respect causally determined, then the contrast which gives rise to the concept of mechanism has disappeared. And if talk of ends is to be abandoned once the underlying mechanisms are understood, and humans are ultimately to be understood as mechanisms, then humans must lack ends; and if having interests depends on having ends, then humans too can actually be held to lack interests, and presumably moral considerability therewith.
[3.2] But I shall not pursue these points here, as there seems in any case to be a confusion affecting the sense of `ends.' For, while one sense of `ends' requires talk of ends to be dropped once a mechanical explanation is available, this is not in any case the sense which usually inclines people to speak of machines having purposes or serving ends. Kaufman's point concerns not so much the apparent special purposiveness of computers, as the ends of machines in general. But machines, as artefacts, essentially have the purposes or ends for which their designers made them, and there is no need to abandon talk of such purposes or ends once their mechanisms are understood. This sense of `purposes' or `ends' does not so much as begin to involve any fiction about machines being conscious; it concerns, ultimately, purposes inbuilt in machines and deriving from the conscious purposes of their designers or users, not purposes held before them by the machines themselves.
[3.3] This being the sense in which machines have purposes, there is a strong discontinuity between their purposes and the purposes of living systems, which do not, in general, derive from intelligent human design. (Genetically engineered organisms, which behave in part in accordance with the purposes of their devisers, are a partial exception.) Indeed, if machines are to be explained by the purposes of conscious entities, then there must ultimately be conscious entities which are not machines, some of them, plausibly, being human beings or other living creatures. Indeed, in view of the discontinuities between artefacts and other entities said to have ends, rather little light is thrown on the good of living creatures, and its supposed relation to ends, by Kaufman's discussion of the ends of machines.
[4.1] But what should be said of the assumption, ascribed to several philosophers by Kaufman, that only beings with ends of their own can have interests (and hence moral standing)? While some of the philosophers cited use the language of "having ends," "being ends in themselves" or "having a telos" when explicating the good of living creatures, including nonsentient ones, not all of them do; and none of the passages cited actually says that living creatures in general have ends of their own (although one passage of Val Plumwood implicitly concerns them having goals of their own) (1991: 146). 3 This, I suggest, is because to have ends of one's own is to have conscious purposes, and because none of these writers is committed to anything of the sort with respect to living creatures in general. Further, while few would dispute that to have conscious purposes is sufficient to have a good of one's own, it cannot be assumed that it is a necessary condition of having a good of one's own. (I shall later be arguing that it is not a necessary condition at all.)
[4.2] In any case, the concept of the flourishing of a creature can be explicated without reference to ends, as opposed to the development or fulfillment of native capacities or tendencies. There is no need to say, as Kaufman thinks many people do in the context of living creatures having ends, that "natural entities such as plants and other living things, and ecosystems, `try' to maintain themselves," or that a thing's "good or well-being consists in those conditions which are conducive to achieving its ends." Indeed if having ends involves talk of achieving those ends, then it is best to reject all talk of nonsentient living creatures having ends altogether. Capacities for growth, self-maintenance and reproduction can be realised without such purposiveness. In other words, Kaufman's account of the common assumption of believers in the moral considerability of nonsentient creatures and the intrinsic value of some of their states miscarries. Such philosophers need not hold that living creatures have ends of their own, or that having ends of one's own is necessary for having interests, as they need not hold (and certainly do not all hold) that such creatures have ends at all.
[5.1] Nevertheless, living creatures inescapably have a good of their own, and thus interests of their own. To flourish as a living creature is, I suggest, to develop most, if not all, of the inherited capacities characteristic of one's kind; and to be capable of such development is to be capable of well-being, and of being benefitted or harmed with respect to the conditions of such development. 4 If so, all such creatures have interests, which (except in the cases of some of the interests of artificially bred and of genetically engineered organisms) are typically independent of the interests of other creatures. Incidentally, as for some creatures these capacities do not include capacities for having preferences, it would already follow that the capacity to have preferences is not a necessary condition of having interests, of bearing intrinsic value, or of being morally considerable.
[5.2] There is a further good which can supervene on having inherited capacities, namely health. However the concept of health should be expounded, it manifestly does not depend across the board on having preferences or on any conscious states at all, despite the fact that apathy is a mark of ill-health for some species. Thus health can wax or wane during sleep and even during a coma; and for nonsentient creatures there is no question of it turning on the creature's preferences. (Indeed, at least for wild nonsentient creatures, if not for domesticated ones, no one would dream of suggesting that what it is to be healthy turns on any preferences whatever.) The capacity for states of health and ill-health also suffices to show that its bearer can be harmed and/or benefitted; this capacity thus comprises a further sufficient condition of moral considerability, and shows once again that the flourishing of creatures cannot simply be a function of their preferences. Nor, indeed, can their interests be.
[5.3] Do artefacts have a good or interests? Some of them (like lawnmowers) have capacities, e.g. to perform functions, while some of them (like refrigerators) are or involve systems, capable of states such as homeostatic equilibrium. Good ones perform their function well; and, while it does not follow that they have a good, the possibility of damaging them strongly suggests that this is the case. Yet we speak of damaging them rather than harming or injuring them, and of repairing or mending them, rather than of curing or healing; and all this reflects the fact that their good is entirely dependent on human purposes. Further, except when they have some sort of aesthetic value, their value is entirely instrumental and derivative; it is not, in any case, intrinsic value. None of this is changed by the fact that, as long as they remain functional, they have inbuilt purposes or "ends," a phenomenon which actually serves to differentiate them both from purposive and from most other living creatures, and thus from creatures capable of well-being.
[5.4] Thus both the good and the value which artefacts sometimes have is derivative. Their good, if they have one, depends wholly on human contrivance or the contrivance of other intelligent species (such as beavers), whether or not it guides their movements; and their value, if any, is entirely dependent on valuable states of other entities (living entities, I should claim). By contrast living creatures, which often have both a good and a value of these kinds, have also a good and a value deriving from nothing but their own nature, and as such can be beneficiaries. Non-living entities, however, whether or not they are artefacts, cannot properly speaking be beneficiaries; and, while some of them have a good, it would be a misuse of language to say that they have interests.
[6.1] It might be objected, however, that to stress the inherited status of capacities as the basis of having a good of one's own is arbitrary. In a related context, Kaufman remarks that to require that to have interests a creature arise by natural selection is arbitrary, since this would imply that an organism arising by spontaneous generation would lack interests (Kaufman: 63). Indeed if we imagine two organisms with exactly similar structure and behaviour, but one of which was spontaneously generated, it would be indefensible to deny interests to the one just because it lacked the evolutionary origin of the other. As Kaufman says, we could still recognise what would be good or bad for an organism even if we knew nothing about its origins (Kaufman: 63-64). But what enables us to recognise this? An understanding, I suggest, that it is endowed with a certain kind of nature, a nature, that is, which involves capacities for growth, self-maintenance and reproduction, and, relatedly, a capacity for health. The phrase `inherited capacities' is merely a phrase for conveying that the organism has received such a nature, whether from evolutionary ancestors or not.
[6.2] If a creature has such a nature, then it is capable of developing these inherited capacities (plus maybe others), and thus of flourishing after its kind, unlike a nonliving artefact. Even if a creature has such a nature in part, modified by genetic engineering, it still remains capable of flourishing, albeit not fully in accordance with the flourishing of other members of its kind or former kind, rather as cultivated apple trees (the products of selective breeding) are still capable of flourishing in something like the manner of crab apples. Nonliving entities, by contrast, lack such capacities, and the capacity for health, and thus a good of their own, whatever may be the functional capacities with which, as artefacts, they may have been endowed.
[7.1] Kaufman further argues, against Peter Singer, that sentience is not sufficient for having preferences or for having interests (Kaufman: 67-69). His objection does not turn on strong cognitive requirements for having preferences, as he is prepared to allow that appropriate behaviour is a sufficient indicator of preferences. His objection rests, rather, on the possibility of creatures which can believe and know and reason (and which are thus aware) but which do not care about anything, and thus lack preferences. Such creatures, however, would seem to be conscious rather than sentient, for, despite their cognitive capacities, they do not feel emotions. Thus this argument shows nothing about sentient creatures. Perhaps it should be added that the creatures depicted would probably have interests, even if they took no interest in them, as for them it would be better to hold true beliefs and sound conclusions rather than false beliefs and unsound conclusions. As other creatures might deceive them on occasion about these matters, their interests could also be an object of moral concern for other agents.
[7.2] In a related argument, Kaufman maintains the possibility of being able to feel pain (and thus be sentient) but not to have preferences (Kaufman: 68-69). While I have some inclination to say that pain is necessarily unpleasant, I do not wish to argue the point here. Instead I want to point out that to be sentient is to be capable of suffering, which is necessarily undesirable, and thus to have interests in the avoidance of suffering; and thus that sentience remains in any case a sufficient condition of moral considerability.
[7.3] Kaufman, however, concludes that sentience is insufficient for having interests or for moral considerability (Kaufman: 69-70). For these, the capacity to have preferences is necessary. That it is necessary for having interests is supposed to follow from the earlier discussion of alternative accounts, which supposedly locate both interests and moral considerability far too widely; hence having desires and/or preferences is to be required. It will be clear from the above brief discussion of the interests of cognitive beings that this is by no means an obvious conclusion; even for cognitive beings there seem to be interests which do not turn on desires. Further, the earlier discussion above on the relation of interests to capacities like the capacity for health serves to show that for human beings as well as for nonsentient creatures there is a further range of interests which do not turn on the feelings, preferences or attitudes of their holders.
[8.1] Indeed an account of interests in terms of preferences is liable not only to cast the net of moral considerability too narrowly; it is likely to supply too restricted an account of the interests of humans and, come to that, of other conscious and sentient creatures. The capacity for preferences is certainly sufficient for having interests, granted the stunting which occurs if it is not developed and the frustration which becomes possible once it is exercised. But the point about stunting applies equally to the capacities of other living creatures, and is by no means restricted to the capacity for preferences or even to conscious capacities only (or, relatedly, to having a point of view). For this and other reasons already given, these creatures have a good of their own, unlike machines, and unlike them are morally considerable.
[8.2] Not only is mentality unnecessary for having interests and for moral considerability; there are many interests, of which the interest in health is one, which do not turn on preferences or even on mentality at all, but which must not be forgotten if we are to have an adequate rather than a threadbare account of human wellbeing and interests, and thus of what is of value in human life as well as in nonsentient life. There is intrinsic value, I have argued elsewhere, in the development of each of a person's faculties (Attfield 1987: chapter 3); if it takes a consideration of the interests of nonsentient creatures to remind us of this, that serves as a further reminder of the value of living amongst a kaleidoscopic range of creatures, some of them with no cares, but still worthy of consideration and care on our part.
Attfield, Robin (1987). A Theory of Value and Obligation. London: Croom Helm.
---- (1991). The Ethics of Environmental Concern (second edition). Athens and London: University of Georgia Press.
Kaufman, Frederik (1994). "Machines, Sentience and the Scope of Morality." Environmental Ethics, (16) 1: 57-70.
Plumwood, Val (1991). "Ethics and Instrumentalism." Environmental Ethics, (13) 2.
Thompson, Janna (1990). "A Refutation of Environmental Ethics." Environmental Ethics, (12) 2: 147-60.
1 A brief reply to Thompson can be found at (Attfield 1991: 205-07). Return.
2 That it is plausible that possible beneficiaries have moral standing is not here an issue, and is taken on trust; arguments for this claim are to be found in (Attfield 1991), chapter 8, and in (Attfield 1987), chapter 3. Return.
3 Cited by (Kaufman 1994) at page 62. Return.
4 I have discussed these matters in (Attfield 1987), chapters 3 and 4. Return.