Preston_11.qxd 5/9/06 7:51 PM Page 183 Preston and Ouderkirk ⫹ 1-4020-4877-7 ⫹ Proof1 ⫹ 10 May 2006 CLARE PALMER 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Please note that we have deleted the number in level head to make it consistent across the book. If your chapter had numbered level head and if you have crossreferred in your chpater (e.g. see section 1.1 etc) kindly re-phrase those areas. (For example if your level head was 1. Introduction and if you have referred in your chapter see section 1) 11. RETHINKING ANIMAL ETHICS IN APPROPRIATE CONTEXT: HOW ROLSTON’S WORK CAN HELP If your chapter has the usage of paper instead of chapter.Could we please change the usage of paper to chapter INTRODUCTION Holmes Rolston has long been regarded as a leading figure both in environmental philosophy and in science and religion. In this paper, though, I argue that Rolston’s work also paves the way towards rethinking animal ethics. Given the well-known hostility between many forms of environmental philosophy and animal ethics, to turn to Rolston—a notorious champion of the former field—in order to advance work in the latter field, may seem singularly perverse. But, I will maintain, Rolston’s arguments— whilst undeveloped and in some respects problematic—provide a better basis for advancing work in animal ethics than the advocacy or rejection of utilitarian or rights positions that have dominated animal ethics for several decades. In particular, I will suggest, Rolston’s work provides tools for thinking through the complicated location of domesticated animals both conceptually and ethically. So, at the end of the chapter, I make some initial moves in outlining how Rolston’s position might be developed to contribute to new thinking with respect to animals and ethics. In focusing on the place of animals in Rolston’s environmental ethics, I will touch only obliquely on the area of his work that has caused most controversy: his endorsement of a theory of objective intrinsic value in nature. That topic has been exhaustively discussed elsewhere, and I will not revisit that discussion here. The place of animals in Rolston’s work is not, though, entirely virgin territory either. Some attention has been paid to it before (notably by Peter Wenz [1989], to which Rolston responded, and later by Ned Hettinger [1994] and Woods and Moriarty [1997]). But my interest takes a somewhat different trajectory to that of existing debates, concentrating on Rolston’s understanding of nature and culture as part of the architecture of a context-oriented approach to animal ethics. In order to develop this argument, I will begin by outlining—in a very basic way—what seem to be some central problems in what we might call “philosophical animal liberation” approaches to animal ethics. Then I will move on to draw out key aspects of Rolston’s understanding of “nature” and “culture”. I will consider how animals are located within these categories, and then make some suggestions as to how Rolston’s position might contribute to a more contextual approach to animal ethics. Two further initial comments should be made for clarification. First, in using the term “animals”, I intend to confine my discussion to non-human mammals and birds. Second, I will be assuming—as does Rolston—that, on grounds of sentience at least, it makes sense to talk about these animals as being morally considerable (no stronger claim, such as that animals have rights, is intended). I will not be putting forward arguments to defend the moral considerability of these animals here. 183 C. Preston and W. Ouderkirk (eds.), Nature, Value, Duty, 183–202. © 2006 Springer. Printed in the Netherlands. AQ:Woods and Moriarty (1997) is not listed in References. Please check Preston_11.qxd 184 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 184 CLARE PALMER PHILOSOPHICAL ANIMAL LIBERATION: OUTLINING A PROBLEM Animal liberation is the common—although inaccurate and unhelpful—catch-all term for a range of different kinds of arguments concerning human ethical obligations towards animals. These arguments arise within both utilitarian and rights traditions and are usually associated in the case of utilitarianism with the work of Peter Singer (1979, 1983), and in the case of rights theory with Tom Regan (1984, 2001). What binds these different arguments together as advocating “animal liberation” is that they hold in common the view that animals make strong moral claims on humans; often so strong as to maintain that the interests of humans and animals should be taken equally into consideration in moral decision-making. Exactly what such claims entail in practice varies in different versions of utilitarian and rights theories, but it is commonly argued that some ways of treating animals—such as eating them—widely thought to be morally acceptable are, rather, morally unacceptable. (It is not necessary for me to pursue these arguments in more detail here.) There are a number of standard objections to philosophical animal liberation arguments. Some are based on the view that the conception of animal capacities and cognitive abilities usually adopted in animal liberation positions is too high-level, sophisticated and anthropomorphic. Others focus on problems in the relevant ethical theory (see Lockwood [1979],Frey [1983], Leahy[1993]). Interesting though these objections are, I will not linger on them here. I want, rather, to consider objections that focus on what we might call the “class system” manifest in philosophical animal liberation. Here, I use “class” in two senses. In the first sense, class is about procedures of classification. Animals are grouped on the basis of their membership of a putative class; the class is defined by the possession of a particular keystone capacity or ability, or cluster of keystone capacities and abilities. The second sense of class concerns the hierarchical value of these classes. The selected characteristics that divide are also value-bestowing. Indeed, the keystone capacity or ability does not just pick out the animals in the class as being morally considerable. That certain basic capacities (such as the ability to feel pain) are central in determining a baseline of moral considerability does not seem especially problematic. But in philosophical animal liberation, such capacities solely and equally determine the degree and kind of moral significance of all animals in the relevant class. The central notion, then, of philosophical animal liberation seems to be that animals possess certain innate capacities, and that these determine—and solely determine—moral significance. This thesis, though, is a vulnerable one in several ways. The way that interests me here stems from the kinds of criticisms made of philosophical animal liberation within environmental ethics. Examples of these criticisms are, for instance, that philosophical animal liberation has no grounds for thinking that a member of an endangered species should be of any more value than one of a well-represented species; and, depending on what “class” the rare individual is in, may be worth less. Another is that philosophical animal liberation positions seem to have a problem with predation that might mandate heavy human management of the wild for animal welfare reasons. Preston_11.qxd 5/9/06 7:51 PM Page 185 RETHINKING ANIMAL ETHICS 185 But, because of the “class system” I have already outlined, there is no easy solution for animal liberation to these kinds of criticisms. An endangered wild cat, a domestic house cat and a non-native feral cat all have (roughly) the same innate capacities. There are no grounds for treating wild animals one way while regarding domestic animals with similar innate capacities differently. This is because contextual differences of this kind are not of ethical significance to philosophical animal liberation. Yet it seems implausible that a thoroughgoing approach to animal ethics can be developed without attention to context. Thousands of different species, some not yet discovered by humans and many with members that will never encounter humans are involved. And, in contrast, also to be taken into account are animals shaped and formed by human actions in terms of genetic make-up, susceptibility to disease, reproductive capacity, bodily form, temperament and cognitive abilities. Focusing on capacities such as the ability to feel pain alone cannot, in particular, capture anything about the dramatic transformation in animals wrought by domestication. From a philosophical animal liberation perspective there is nothing of direct ethical interest to be said either about human intervention in the processes of bringing into being, selectively breeding and shaping the natures of domesticated animals; or about the human independent, ecological embeddedness of wild animals. It is in recognizing the significance of context and relationships such as these to humans that, it seems to me, Holmes Rolston’s work can assist in the development of animal ethics. His account—while I think it is problematic in some respects—provides important building blocks for a more contextual and relational animal ethics. In order to consider his contextual position more fully, though, it is essential to have some understanding of the place the contexts of nature and culture play in Rolston’s work. ROLSTON ON NATURE AND CULTURE Understanding Rolston’s approach to ethics in general, and to environmental and animal ethics in particular, requires an overview of his interpretation(s) of nature, and how nature may be distinguished from culture.1 This is by no means straightforward. In the first chapter of Environmental Ethics, and in earlier papers on the subject, Rolston outlines a number of different ways in which humans might be said to “follow nature.” This inevitably involves Rolston in offering a range of different possible interpretations of the meaning of “nature” and “natural” behavior. However, Rolston’s discussion does not map very well onto his own usage of the terms “nature” and “the natural.” It is the ways in which these terms operationally contribute to Rolston’s arguments that are significant here in trying to understand Rolston’s position. I will begin, then, by outlining different ways in which an analysis of Rolston’s writing suggests that he is using the terms “nature” or “natural”.2 (N0) That which most broadly “obeys natural laws” (Rolston 1986, 31). Although this may be interpreted to include “astronomical nature”, Rolston prefers to restrict this to a global sense, and in particular to the physical and biological processes that form the system which “gives birth to life.” 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 186 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 186 CLARE PALMER (N1) That which is not of human origin; the wild understood as “nature outside human control”(Rolston, 2003, 108); processes that are spontaneous, non-reflective and independent of humans, e.g. evolution and speciation. (N2) Behaviors and practices that are common across species (including humans), for instance behavior that enables survival and reproduction e.g. sleeping, grooming, eating. (N3) The expression of distinctive (not shared) species-specific characteristics and behaviors manifested by “natural kinds” e.g. coyote nature, warbler nature, human nature; all species, Rolston maintains “live in the world with some degree of uniqueness.” (Rolston 1994, 1) Complex relationships obviously exist between these senses of “nature” and “natural”. N0 is the broadest sense; the N1 and N2 senses of nature are wholly subsets of the N0 sense. Some N3 behaviors and practices are also subsets of N0 and N1; but N2 and N3 behaviors are exclusive of one another. An example might help to explain these distinctions. Suppose a wild rabbit is eating grass on a mountain-side. In what senses is this natural behavior? It is natural in the N0 sense—part of the biological systems of earth; it is natural in the N1 sense—spontaneous and independent of humans; it is natural in the N2 sense—eating is a common cross-species behavior. It is not natural in the N3 sense, as expressing the specific, exclusive nature of “rabbitness” (unless the eating was being described in a particular, fine-grained way to emphasize its rabbit-distinctiveness as opposed to, say, the eating habits of prairie dogs). These distinctions may seem tiresome, but they are necessary in order to understand where Rolston locates human beings. Humans are, inevitably, included in the N0 sense of nature; they emerged as part of the physical and biological systems that produced life on earth and are still dependent on these systems. But Rolston seems to have a more particular interpretation of what part of human life falls under the description of N0 nature—that is, those processes which happen to humans and that are (currently) beyond human control; so for instance, blinking when grit gets in an eye or sneezing at too much chili. Humans are, in contrast, by definition excluded from the N1 sense of nature where “any deliberated human agency, however well intended, is intention nonetheless and interrupts these spontaneous processes” (Rolston 1991, 371). Humans can, though, be included under both the N2 and N3 senses of nature. In the N2 sense, humans do, of course, display behaviors that when broadly described, are cross-species behaviors (such as sleeping and eating). In the N3 sense, humans display distinctive species-specific characteristics that are exclusively human and do not fall under the N2 sense of nature. These human-exclusive characteristics correspond closely to Rolston’s use of the term culture. There are, presumably, some exclusively human characteristics, products of N0 nature, that do not fall into the category of culture (perhaps no other species sneezes at too much chili) but Rolston is not particularly interested in these. What he does argue is that culture falls wholly within the human sense of N3 nature; culture is exclusively human and species-specific. Preston_11.qxd 5/9/06 7:51 PM Page 187 RETHINKING ANIMAL ETHICS 187 What, then, does Rolston understand by culture? He does not precisely define cultural behaviors and characteristics, but stresses intention and deliberation (as opposed to the spontaneous action of nature in N1); the dramatic and deliberate rebuilding of the environment (Rolston 1988, 1, 5; 1989: 132) and “moral and spiritual sensibilities” including value invention and recognition (Rolston 1988, 40). In later work, Rolston repeatedly emphasizes that culture is distinguished by the transferal of information neurally rather than genetically, resulting in human education and knowledge-accumulation, and by the deliberation and reflectiveness involved in the inevitable process of environmental transformation. (Rolston 1994, 2–3) Rolston’s view of what qualifies as culture, then, is quite high-level, given the multiplicity of ways in which culture can be understood. He requires such an idea because he identifies culture with that which is species-specific to humans and is not shared by non-human animals.3 Although culture is dependent on N0 nature in the sense that it still “requires nature in life support” (1988, 40), Rolston insists that culture is radically discontinuous with it (1989, 32); it “did evolve out of nature, but has evolved out of it”(1998, 3). Although some human behaviors (blinking, sneezing) remain N0 natural, human culture, according to Rolston, no longer falls into this category. One way Rolston conceptualizes this is to say that nature (understood in an N0, N1, N2 and non-cultural N3 sense) is primarily about causes (chemical and genetic for instance) while culture, as an expression of human-unique nature, is about reasons.4 However, we need to be careful here, for there is yet one other way in which Rolston deploys the term nature: (N4) Cultural, deliberate human choices to behave “more naturally” or to use Rolston’s term, “follow nature”. This understanding of “natural” is rather different from N0–N3; the N4 sense of nature is not about origin. In terms of origin, N4 behavior is not natural in any of Rolston’s senses; it is cultural; it flows from culture and involves reason and deliberation. However, what is pursued in this deliberation, the object of the reasoning, is nature, in one or other of Rolston’s senses. A deliberate choice might be made to act sparingly, hold back, withdraw or leave alone to allow N0 processes to operate (for instance, to use one of Rolston’s examples “natural childbirth” is a decision to refrain from a series of alternative medical practices). Deliberate attempts to imitate, parallel, fit in with or approximate more closely some perception of wild, spontaneous N1 nature, perhaps by restoring land to look like wilderness or by growing crops to fit local soil and climate are natural in this sense. Rolston also suggests that humans may act naturally (in an N4 sense) by deliberately choosing to do what some other species do unreflectively—that is, by culturally imitating either N2 behavior or N3 behavior specific to another species. ANIMALS, NATURE AND CULTURE IN ROLSTON’S WORK Sorting out different senses in which Rolston uses “nature,” “natural” and “culture” allows us to consider where Rolston locates animals. Most fundamentally, Rolston 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 188 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 188 CLARE PALMER rests on distinctions between N1 nature and culture. Animals are divided into two groups. One group falls wholly into the category of N1 nature (and also, by definition, N0 nature; but the former is what is significant here). Members of this group are part of, and indeed partly comprise, wild, spontaneous nature. They exist independently of humans, though they may encounter humans when humans move or act in wild nature (by hiking, or constructing roads or pipelines, for instance); and they may be unintentionally affected by other human actions (in terms of, for instance, pollution or the introduction of exotic species). Both the N2 and the N3 behaviors of these animals are subsets of N1. The second group of animals Rolston considers are those that have been, as he puts it, “captured for food, domestication, research or other utility”. He does not explicitly distinguish between different kinds of “captured” animals (for instance between zoo animals and household pets). His (relatively brief) discussion of these animals focuses on domesticates, in particular domesticates kept for food. He locates domesticates in the following way in relation to his nature/culture distinctions: (a) Domesticates are outside N1 nature: Domestic animals are not part of spontaneous wild nature; they are “living artifacts”; they have been “transformed by culture” (1988, 79); they require human action to survive and sometimes to reproduce. (b) Domesticates are subject to N0 natural processes: Even though their minds and bodies may have been transformed by human culture, they themselves are entirely subject to “natural laws” in Rolston’s sense (to disease, injury and so on); (c) Domesticates are not natural kinds and so have no N3 behavior: Because humans have selectively bred domestic animals, they are now breeds and not natural kinds. Thus domestic animals cannot behave naturally in an N3 sense; N3 behavior can only be manifested by natural kinds. (d) Domesticates are not part of culture: Culture is the manifestation of deliberated N3 species-specific human behavior. Domesticates, not being human, cannot manifest these characteristics and so are outside culture. (e) Domesticates can behave naturally in an N2 sense: Domesticates are not part of wild N1 nature; they are not natural kinds, and so cannot act naturally in an N3 sense; they are not human so cannot behave naturally in a deliberative N4 sense. They are, however, part of N0 nature in the sense that they are subject to natural processes; and they can manifest N2 natural characteristics in the limited sense of sharing in common cross-species survival behaviors. Rolston maintains that both pets and agricultural animals have been removed from N1 nature, but that neither can be part of human culture. Agricultural animals, Rolston suggests, inhabit (literally) the no-man’s land of the “peripheral rural world”. (1988, 79) These categorizations are central to Rolston’s discussion of human ethical responsibilities towards animals. ROLSTON, ANIMALS AND ETHICS Rolston argues forcefully in Environmental Ethics that all living beings carry intrinsic value. (As I have already said, this argument has been exhaustively discussed Preston_11.qxd 5/9/06 7:51 PM Page 189 RETHINKING ANIMAL ETHICS 189 elsewhere, and I will avoid further comment on it here.) He also accepts (rather than argues for) the view that higher animals are sentient and that this sentience is of significance in ethical decision-making.5 But although sentience is significant in ethical decision-making it is, crucially, context that gives guidance on duties towards animals (1988, 61). Duties towards similarly sentient animals vary according to how animals are placed in relation to N1 nature and culture, and what kind of human actions are, or have been, involved in their situation. Scrutiny of Rolston’s work seems to suggest that three key contextual factors need to be taken into account when considering ethical obligations towards an animal: 1. How the animal is placed in relation to N1 nature and human culture; 2. Whether previous human actions have had an impact on the animal’s current situation; 3. Whether humans are acting naturally towards the animal (and if so, in which sense). Of particular concern for Rolston is human treatment of wild animals—that is, sentient animals wholly located within N1 nature. Such animals come into being and live independently of humans. Whilst he objects to gratuitously harming wild animals, he also argues against assisting them by, for instance, attempting to reduce their pain. Human intervention both changes N1 patterns of evolution and speciation (perhaps preserving individuals with weaknesses that otherwise would not have survived and allowing these weaknesses to be genetically inherited) and humanizes wild N1 nature. “Pain in nature is situated, instrumental pain; it is not pointless in the system, even after it becomes no longer in the interests of the pained individual” (Rolston 1988, 60). That animal pain can be outweighed by other, ecological and evolutionary, values relating to the systems in which wild animals are embedded is one sense in which context might be important in making ethical decisions about animals. But Rolston hints at another sense. Humans have, he says “no obligations to help wild animals; we are obliged to leave them alone.”(Rolston 1989, 134). Wild animals are self-sufficient; they can and do provide for themselves. There is no reason why they have any claim on humans, and to assist them is to create a relationship that did not previously exist. The basic argument here is, then, one I shall call the no relation/ no obligation argument. That humans have no relations to wild animals means that they also have no obligations to them. It is only where human actions have already created some kind of effect or relationship—where the suffering, even though it may be part of nature in an N0 sense, is not genuinely N1 in origin—any ethical obligations come into play. In this regard, Rolston comments: “If human intervention, and not just the forces of natural selection are causing the deaths, that does seem to make a difference to the welfare claim” (Rolston 1988, 56). He approves, for instance, of a case where a rancher was required to flatten portions of extensive new fencing in order to allow starving antelope to reach their traditional winter grazing areas. Again, although Rolston does not make this explicit, two kinds of reasons seem to hold here. Human actions have already created some kind of relation or interaction with the antelope which means they are no longer independent and self-sufficient; and wild ecological processes have already been disturbed by the separation of the antelope from the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 190 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 190 CLARE PALMER grazing; no further disturbance, and some diminution of its effects would result from removing the fencing. There is, however, one context where, even without the justification of prior human activity, Rolston maintains that humans can act on sentient animals in N1 nature. This is when humans are acting naturally in an N2 sense—that is, when they are acting in ways common across species for survival. Some forms of hunting , Rolston maintains, fall into this category. When hunting is carried out in an N2 sense Rolston views it as a form of human predation. In this category he is inclined to include all hunting where the hunted is eaten or used in some other significant way (e.g. for fur, if the fur is for warmth rather than fashion), even if the food or fur is not hunted as a necessity for the hunter (as we would expect in the case of subsistence hunters). Trophy hunting, on the other hand, does not fall into this category. I will discuss some problems concerning this classification of hunting later. We can, then, extrapolate something like the following principle from Rolston’s work: (P1) Humans should not harm, but have no duty to assist, wild animals in N1 nature unless the action is in response to some prior human activity or humans are behaving naturally in an N2 sense. Based on the trumping significance of other ecological values, and the no relation/no obligation argument, Rolston creates a contextual principle, P1, of non-intervention in N1 nature for animal welfare reasons. This principle certainly avoids some of the difficulties arising out of philosophical animal liberation—such as intervention in predation.6 But it’s a contextual principle that solely concerns animals with which humans do not interact. What about ethical obligations towards animals with which humans do interact—the domesticates who are not part of N1 nature, but also, according to Rolston, not part of human culture either? Rolston argues that domesticated animals should properly be considered in ethical terms against the background of the N1 nature from which they originally came. Unlike the philosophical animal liberation positions outlined earlier, the comparison class is not human beings who share the same valuable capacities as animals, but rather wild animals for whom humans should not relieve suffering. The origin of domestic animals in wild nature, for Rolston, still determines appropriate ethical treatment. This grounds what Rolston calls the “homologous principle”: (P2): Humans should not cause inordinate suffering to domestic animals, beyond those orders of nature from which the animals were taken (1988:61). Rolston calls this a principle of “non-addition”: Humans should not add to the suffering an animal would have endured in the wild, but, Rolston points out, non-addition does not mean subtraction from suffering; there is no obligation to cause or allow only lesser suffering. Elsewhere, he comments “where culture captures value in nature, there is only a weak duty to subtract from the pain” (Rolston 1988, 61). (He does not explain, though, what distinguishes a strong duty from a weak duty, nor what kind of Preston_11.qxd 5/9/06 7:51 PM Page 191 RETHINKING ANIMAL ETHICS 191 obligation this entails). To this, however, Rolston adds another principle: (P3): The suffering of domestic animals should be removed, as far as it can, where it has become pointless because the animals are no longer in the environment of natural selection, though this is “commended rather than required” and is “benevolence not justice.” (1988, 61) The homologous principle, P2, then, supposes that wild animals are the appropriate comparison class for domestic animals. If we accept the homologous principle, domestic animals should not be caused more suffering than occurs in the wild. But we already know that any amount of suffering should be permitted among wild animals in N1 nature, since we should not assist wild animals by relieving suffering. So P2 alone, as a principle, potentially permits humans to cause high levels of suffering in domestic animals, since high levels of suffering may occur without relief in wild animals. But P3 changes this situation significantly, because P3 introduces a strong contextual concern. It is the case, on Rolston’s principles, that high levels of suffering are acceptable without relief amongst wild animals. But one of the reasons for accepting this is because wild animals are part of valuable N1 evolutionary processes. P3 notes that at least some domestic animals are outside the environment of natural selection, and that therefore a good reason for accepting suffering in a wild context does not exist in the domestic context. Therefore suffering should be removed as far as it can, though this is benevolence not justice. Indeed, once P3 has been taken into account, it isn’t clear why P2 is necessary, since there are no domestic animals that fall outside the scope of P3, but to whom P2 applies. No domestic animal is in the environment of natural selection, so (according to P3) the suffering of all domestic animals should be relieved, as far as it can, as benevolence not justice. It’s also worth noting that what is entailed in P3 is somewhat stronger than that which is entailed in P2. P2 concerns human actions as the cause of suffering; P3 concerns human actions as removing suffering (which presumably incorporates not only the concern in P2 not to cause it, but also to relieve it when it arises from non-human N0 sources such as disease or accident). A minor reformulation, P3* should serve instead of P2 and P3: P3* The suffering of all domestic animals should be removed, as far as it can, since it has become pointless because the animals are no longer in the environment of natural selection, though this is “commended rather than required” and is “benevolence not justice.” (1988, 61) In dismissing P2 and moving to reformulate P3, though, I perhaps moved a little too swiftly. Rolston does suggest another argument that might be used in support of P2. Even though domestic animals are not part of N1 nature, nonetheless, he argues, the process of domestication in agriculture parallels the actions of wild ecosystemic processes; agricultural animals are in quasi-ecosystems. The killing of animals for food then, parallels predation. So the same rules about suffering are appropriately applied to both wild and domestic animals. This seems to be based on a broader view that agriculture is a way in which “humans step back from culture into the wild” 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 192 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 192 CLARE PALMER (Rolston 1988, 79) and that agricultural systems should be ethically appraised just as wild ecosystems are ethically appraised.7 This suggestion puts domestic animals into a quasi-N1 category. But the argument presents problems in a number of ways. First, it is an argument that could apply only to some domestic animals, primarily agricultural animals (not to pets, nor to experimental animals, for instance). Second, we need to get clear about what Rolston means by “quasi-ecosystems”. Is the suggestion a literal one, that domestic food animals, in roaming on pastures and mountains, become quasi-ecosystemic members? This is, I suppose, an empirical question (which raises questions about Rolston’s division between N1 nature and other environments). Of course, this reading of quasiecosystems could only apply to a minority of agricultural animals in any case; animals intensively farmed, in particular those kept indoors, could hardly be described as members of ecosystems in ways comparable to wild animals. I shall argue below, however, that such a comparison—even of a minority of domestic animals—reveals a problematic understanding of domestication. However, it may be that Rolston does not intend “quasi-ecosystems” to be taken in such a literal sense; rather he means that just as wild animals prey on other animals in ecosystems, so humans prey on animals in agriculture. The processes are parallel and therefore domestic animals should be regarded, ethically, in the same way as wild animals. Rolston certainly does seem to suggest that agricultural meat-eating is a form of predation and is thus natural in an N2 sense. But the objection to this argument must be that agricultural animals are selectively bred, controlled, sustained and killed as part of a deliberated cultural practice. Since this is not N2 human behavior, it cannot count as one of the P1 exceptions, even though it could be argued that animal agriculture is natural in one form of the N4 sense (that is, where culture mimics nature). This makes the continued maintenance of P2 problematic. The elimination of P2, of course, does not mean the rejection of animal agriculture, nor entail vegetarianism. All that is established is that a good reason for accepting suffering in the wild context, which informed P1, is not a good reason for accepting suffering in the domestic context, as maintained in P3 (and P3*). It would be perfectly possible to accept P1 and P3* and to maintain animal agriculture. Indeed, P3* is not a strong principle. If P3* is taken to be compatible with intensive farming, removing suffering “as far as it can” may still leave a considerable amount of suffering (and the removal of the suffering is, anyway, not required by justice, only by benevolence). If P3* is interpreted as being incompatible with intensive farming, it is still compatible with other forms of animal agriculture and with meat-eating; since, as Ned Hettinger rightly points out elsewhere, Rolston offers no principle here that taking an animal life is itself bad, if no suffering is involved.8 This discussion has taken no account, however, of the no relation/no obligation argument that appeared to be part of Rolston’s case for non-interference with wild pain. Introducing it here, though, would not change anything. Since domestic animals are not only outside the environment of natural selection, but also outside the context of no relation, neither of the reasons Rolston suggests for accepting wild pain apply in the domestic case. Domestic animals are, plainly, in a situation of relation Preston_11.qxd 5/9/06 7:51 PM Page 193 RETHINKING ANIMAL ETHICS 193 with human beings. That this is a situation of relation rather than no relation, does not of course in itself entail any obligation. But Rolston does make a gesture towards affirming a possible form of a relation/obligation argument (and later in this paper I will also be attempting to construct such an argument). Rolston comments of domestic animals “In taking an interest in them [relation] humans have assumed a responsibility for them [obligation]” (1988, 79). But his work makes little attempt to flesh this claim out, either in explanatory or practical terms. Primarily interested in the context of the wild, concern for the context of domestication is slight and engagement with it minimal (and there is a persistent sense that the domesticated and rural, tainted by human contact, are inferior contexts to the wild, as are the animals within them). Indeed, Rolston simultaneously exaggerates some senses of the difference between domesticates and wild animals, whilst underplaying other senses. In Rolston’s scheme, domestic animals are artifacts of human culture. They are, he maintains, breeds, not natural kinds, and so do not have species natures (we can, for instance, talk about wolf nature, but we cannot talk about dog nature; we can only talk about airedales or poodles). Yet there clearly are characteristics that all dogs have in common (for instance, that they are social animals with dominance hierarchies; that they have a highly sensitive sense of smell, and so on). Many of these characteristics are shared with wolves—albeit in neotonized form. Dogs are not fully artifactual; certainly they bear the imprint of human influence in their bodies and temperaments; but also they carry many characteristics of their wild forebears. Here, Rolston seems to take the transformations effected by domestication too far. But, on the other hand, Rolston does not seem to take other effects of domestication far enough. Artifactual though domesticates may be, as we have seen with the original P2, he nonetheless considers that their comparison class is the wild, and that we should think about “what would have been their lot in the wild, on average, adjusting for their modified capacities to care for themselves” (1988, 79). But how should this be interpreted? The domesticates concerned would not have existed if humans had not bred them. Without humans there would be no life at all, not an alternative possible wild life. And what sort of “adjustment” does Rolston mean? Does he mean us to imagine (say) that a poodle has the capacities to survive in a wild that a wolf has, and to compare suffering in the poodle’s life with suffering in the life of an (average) wolf? Or to imagine the kind of life this particular poodle would have if released into the wild, adding a few extra claws and some extra aggression as an adjustment? That Rolston thinks it possible merely to “adjust for modified capacities to care for themselves” is an indication that he takes the changes of domestication insufficiently seriously. Many domestic animals would not survive in the wild at all. Some would survive as scavengers. A few might “resettle homeostatically into environmental niches” (1988, 78–9)—though as animal breeding and genetic modification of animals progress, this possibility becomes increasingly unlikely. It is, in general terms, the loss of the ability to survive well, or to survive at all, independently that significantly separates domestic from wild animals; indeed the creation of domesticates is the deliberate creation of dependence.9 Dependence, 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 194 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 194 CLARE PALMER I will suggest, is an important contextual characteristic (amongst others) in reconsidering human ethical relationships with animals. AN ASIDE: THE MISCLASSIFICATION OF HUNTING Before moving on to consider how Rolston’s animal ethics may be developed, I want to make one further point concerning an apparent misclassification in his work. Rolston maintains that some kinds of human practices (as we have seen with animal agriculture) fall into the category N2 natural—shared across species. Where human behavior towards animals is concerned, one main instance of this he cites is hunting. This claim that is attacked both by Wenz (1989) and by Moriarty and Woods (1997). All three question whether human hunting practices can be conceived of as natural (though Wenz allows that hunting by indigenous peoples might be thought so). Moriarty and Woods emphasize that hunting, in the US at least, even where the kill is consumed, is characterized by hunting codes, technological gear and equipment and reliant on automobile use. Therefore, they argue, it is a cultural practice misclassed by Rolston (and once it is classed in culture rather than nature, the reclassification has consequences for its ethical acceptability). These criticisms can be explained in a more careful way using the nature/culture distinctions I have outlined. Rolston suggests that humans can behave naturally in different ways: in N0 ways when acted upon by the “laws of nature”; in N2 ways when manifesting shared, cross-species behavior; in N3 species-specific cultural behavior; and in an N4 way by deliberated cultural behavior that in some sense follows nature. Rolston maintains that hunting is natural in the N2 sense. But for this to be the case he would have to argue that hunting was a non-reflective, non-deliberate practice, since it is reflection and deliberation that, for him, are central in distinguishing N2 natural behavior from N3 cultural behavior. But studies of hunting throughout human cultures indicate that this is not the case. All forms of human hunting are deliberate in this sense. Some refinement to Rolston’s spheres of nature and culture might, however, allow him to accommodate hunting as natural in an N2 sense. Suppose one took a controversially high-level view of what went on in some forms of hunting by other animals; that is, that something analogous to deliberation or planning took place. Then the deliberation and reflection involved in human hunting would look more like a shared, cross-species behavior. As currently constituted, though, attributing these higher-level abilities to non-human hunters would lead to more permeability at the boundary of animals and culture than Rolston would want to accept. It might be possible, though, for a finer definition of culture to be constructed such that animals are excluded even if they are able to plan or deliberate at some level; but more work would need to be done before this revision could stand. While it may be problematic for Rolston to suggest that hunting is natural in an N2 sense, it is entirely plausible, though, that, in many Western non-indigenous cultures at least, hunting is frequently viewed as deliberately natural in an N4 sense. Hunting codes designed to give animals an opportunity at escape may be a deliberate attempt to hold back and not to use all the technological possibilities at hand (one possible Preston_11.qxd 5/9/06 7:51 PM Page 195 RETHINKING ANIMAL ETHICS 195 interpretation of N4 naturalness). Further, the commonly-made case that hunting emulates or reminds of simpler, more traditional ways of living, or provides ways of obtaining food that are closer to the wild than those of mechanized farming also fit this idea of N4 naturalness. Rolston’s account, suggesting that hunting is “harmonious with nature” seems to echo this view. He may also be thinking of hunting as humans imitating what animals do instinctively; (Rolston 1988: 92, 94) but he never suggests that it is not deliberate or intended. In all these cases, naturalness is naturalness in an N4 sense, a cultural, deliberate attempt to follow nature. On Rolston’s current account, then, all human hunting is misclassified as N2 behavior, since it is always deliberate (this is not to say that there are no N2 human activities, but rather that hunting is not one of them). This conclusion undermines the ethical case Rolston makes for hunting, by maintaining that Rolston has misclassified hunting, and that the change in classification means it can no longer be seen as a permitted exception under P1. (It is not intended to imply that therefore there is no ethical case to be made for any form of hunting, however; merely that this case is not it). ROLSTON, CONTEXT AND ANIMAL ETHICS As we have already seen, Rolston maintains that not only capacity (primarily sentience) but also context is important in ethical decision-making about animals. In understanding context, it is useful to distinguish between the levels of origin and prior contact, process and place. For instance, in terms of origin N1 nature is that which does not originate from humans and has not been altered by human prior contact; in terms of process it is all those processes (evolutionary, ecological) that operate independently of humans; in terms of place it is the land that issues from this origin and these processes—that is, the wilderness. Although Rolston does not make these distinctions explicitly in the context of animals, nonetheless applying them to animals allows for the possibility of some finer discriminations. For instance, the separation of the context of origin from the context of place allows for distinctions between animals constitutionally independent of humans (that is, not having been bred by humans) and animals that live independently of humans (whether or not they are constitutionally independent). Questions about prior contact allow for the discussion of special claims. It might be argued that domestication itself provides grounds for a form of special claim. But alongside this, there are also more specific special claims, where an individual human or groups of individual humans have chosen to establish relationships with particular individual animals that otherwise would not have existed. Rolston’s own account does not develop such distinctions. Indeed, they set up some difficulty for his principles P1–P3. For instance, human commensals or scavengers may be wild in their context of origin but not in their context of place, since they may live alongside humans and rely on them for provision; do they fall under P1 or P2? Feral animals, in contrast, may not be from N1 nature in their context of origin—they have been domesticated—and they may affect N1 processes; but on the other hand, they may be independent of human provision and largely out of human 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 5/9/06 7:51 PM 196 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Page 196 CLARE PALMER control (which makes applying P2 and P3 awkward). Rolston’s categorizations are problematic inasmuch as they fail to fit the more specific contexts in which many animals are located and therefore provide us with no guide as to how to how such animals should be treated ethically. A more discriminating categorization than Rolston offers—though using the kind of distinctions that can be drawn from his work—is required to lay out fully the different contexts in which animals are located. Such a categorization (though certainly not exhaustive) might go something like this: 1. A wild context, where animals originate, subsist and are located independently of humans, though sometimes encountering humans (N1 nature in Rolston’s sense). 2. A domestic context, of animals that have been, at some time, bred by humans and are now dependent on them for food and shelter, where this has been deliberately encouraged by humans. This may be a relation of positive affect (e.g. pet animals). 3. A feral context of animals that have been, at some time, bred by humans and are now dependent on them, but as scavengers rather than as intentionally encouraged by humans (e.g. urban feral cat colonies). 4. An exotic context of animals, released by humans into an area, and now living independently of humans (e.g. mink released from fur farms in the UK). 5. A scavenging context of animals that are constitutionally wild (have never been bred by humans) but that live alongside humans, manifesting various degrees of dependence on humans for food or shelter. This dependence has not been deliberately intended by humans and in the case of animals viewed as pests, may be positively discouraged and be a relation of negative affect (e.g. rats, raccoons, pigeons). 6. A commensal context of animals that are constitutionally wild but that live some or all of the time alongside humans, and are partly or wholly dependent on humans. This dependence is intentionally encouraged by (some) humans, and the animals are regarded positively though rarely with individual affect (e.g. wild birds fed at bird tables). 7. An agricultural context of animals highly bred by humans, kept for a functional role, highly dependent on humans, and deliberately created by humans in this way (experimental animals fall into a similar category). 8. A captive context of animals that are constitutionally wild, but that are kept confined by humans for a variety of purposes (often for display) and are dependent on humans inasmuch as they are captive and outside their native habitat. If these contexts are tabulated, differences can be clearly highlighted: Wild Pet Urban Feral Invasive Exotic Pest/scavenger Commensal Agricultural Captive Wild Wild Wild Living Dependent on Constitution (Origin) (Process and place Humans Intended/Encouraged by Humans √ X X X √ √ X √ X √ X X X √ √ √ √ X X √ X (√) X X X √ (√) X (√) (√) √ √ Preston_11.qxd 5/9/06 7:51 PM Page 197 RETHINKING ANIMAL ETHICS 197 This account is, of course, not exhaustive, the boundaries here are fluid, and the categories rather generalized. Some scavenging and commensal animals, for instance, may only be partially dependent on humans; some pets or agricultural animals might manage to survive independently of humans (though they would then begin to move into the feral category). Some people may encourage animals that others despise; other animals, introduced at a period in the past and encouraged at that time—such as red foxes in Australia—later become widely regarded as pests to be destroyed. These complications would have to be taken into account in considering any particular case, and in any more detailed examination of animal contexts. However, this categorization suffices to pick out particular kinds of relations between humans and animals. But why are these kinds of relations—developed from Rolston’s discussion—to be regarded as significant? Let’s think a bit further about just one of these as an example: domestication, and in particular, the dependence it entails, in the context of a possible relation/obligation argument. ONE VERSION OF A RELATION/OBLIGATION ARGUMENT Rolston suggests that “taking an interest” in animals by domestication generates some kind of moral responsibility for them that does not apply to wild animals. He does not, however, fill out this claim at all, nor indeed does his account of actual obligations to domesticated animals (not to make their lives worse than those of wild animals) seem to support it. While the claim is intuitively attractive, much more work would need to be done to make it convincing; I will attempt here only to make some very preliminary moves. First, we should distinguish between two different kinds of dependence, which I will call external and internal dependence. Captive wild animals, for instance, might be able to provide for themselves in their native species habitat but in confinement are circumstantially dependent on humans to provide food and shelter. Their dependence is an external, humanly-imposed effect of captivity. Domestication, though, is a process that usually produces internal dependence, dependence that is constitutional and permanent. Domesticated animals are, usually, dependent on humans to provide for their vital needs in terms of food and shelter (at least). It is this internal dependence of domestication in which I am interested here, and it throws up a number of questions that turn out to be really rather complicated. Humans are, in some sense, deliberately responsible for the existence of dependent domesticated animals. Whatever one might think about the possibility of animal “collusion” in the initial relations that led to domestication, it is clear that humans actively and deliberately engaged and still do engage in ever more specialized forms of selective breeding and more recently genetic modification. Animals are thus brought, by humans, through deliberate human choices, into dependent relations. The resulting loss of animal independence (with respect at least, to wild forebears) while sometimes a side effect of selective breeding for other ends (e.g. the breeding of cats without claws) is also deliberate, since an ability to live independently may jeopardize the plasticity of domestic animals to human intentions. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 198 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5/9/06 7:51 PM Page 198 CLARE PALMER But pursuing this notion of responsibility more closely raises a number of standard philosophical difficulties about responsibility. Who is causally responsible, and in what ways, for the creation of dependent domesticated animals? Was it the original domesticators? Those individuals (farmers, breeders, experimental scientists) who deliberately act to produce domesticated animals in the present? Or those who voluntarily demand the products of domestication, from pets to pastrami? How are individual and collective responsibilities to be understood, in such a complicated and ongoing situation as this? 10 These questions are interesting in themselves, but their significance is increased if, by the attribution of causal responsibility, some sort of moral responsibility is thought to come into play. Of course, that someone or some group has played a part in creating x, with the deliberate intention of creating a dependent x, does not in itself show that they have any duties to take x into account in moral decision-making. X must have some further claim to moral consideration. But the assumption that animals are morally considerable was made at the beginning of the paper, and is not at issue here. Given this, there does, as I have already suggested, seem to be a plausible argument along the line that the deliberate creation of a dependent morally considerable being brings obligations to provide for that being. This would be one particular form of a relation/obligation argument where importance is based both on deliberate creation of the relation and the kind of relation [dependence for vital needs] as the basis for obligation. Some of those philosophers who have considered relations not dissimilar to these but in the human situation have, however, found this kind of analysis unsatisfactory. Goodin (1985) maintains that moral obligations in situations that involve one party’s vulnerability to another—such as children to parents—are not related to previous actions voluntarily undertaken and self-assumed agreements. So, he insists, such obligations solely derive from whether the individual concerned is depending on us, is particularly vulnerable to our actions and choices (Goodin 1985, 11) and/or whether we are the best or most obvious person to meet their need or protect them in their vulnerability. In some cases this would not result in any difference in terms of who has the duty; but the reason for the duty is different. As Goodin himself accepts, these two kinds of explanation for relational obligations in the context of vulnerability and dependence are not necessarily exclusive, although he argues that many relational duties to the vulnerable cannot properly be subsumed into the voluntaristic model, since they were never voluntarily assumed by anyone. This may be so, but equally some of the issues discussed here may raise questions over the broadness of application of his thesis. Goodin (1985, 181) maintains that humans have special duties to animals on the grounds that animals possess morally significant interests and that they are vulnerable to humans, both individually and collectively. But if the presence of vulnerability and being the most able person to assist are sufficient to generate moral obligations, then (for instance) the park-keepers in Yosemite National Park should treat the bighorn sheep’s natural diseases. Of course, Goodin might respond that his position could be overridden by other environmental values. But he would not be able to accept Rolston’s sense of no relation/no obligation because, in Goodin’s scheme, encounter is enough to generate Preston_11.qxd 5/9/06 7:51 PM Page 199 RETHINKING ANIMAL ETHICS 199 the relation. If the vulnerability presents itself, and you are in the best position to relieve it, then the relation comes into being and along with it, the obligation. That you had no active involvement or causal role in the situation is irrelevant. As with utilitarianism and rights approaches to wild animals, though, this position seems not to work well with wild animals, even if it were thought to work well in the case of people. CONCLUSION In this paper, I’ve suggested that Rolston’s work provides a way forward in addressing some, at least, of the difficulties posed to philosophical animal liberation by some forms of environmental ethics. Rolston’s contextual emphasis (although problematic as it stands in his own account) alongside his acknowledgment of the value of animals as sentient individuals, begins to suggest some new approaches to animal ethics. In particular, his contextual account opens up new ways of thinking about domestication. Domestication can be seen as a process such that, in creating relationships that close down domestic animals’ abilities to live independent lives, creates special human responsibilities to provide for them.11 Indeed, it seems likely that further development of this position would deliver some, at least, of the protection for domesticated animals that philosophical animal liberation advocates set out to achieve in the first place. And this can be achieved without implausible consequences in the wild to which environmental ethicists so strongly object. This approach to animal ethics, then, may be able to bring environmental and animal ethics closer together, whilst at the same time providing tools to develop a more complex animal ethics, sensitive to human-animal relations and to the contexts in which different animals are located. ACKNOWLEDGMENTS I am grateful to Erica Fudge, Emily Brady and Jose Luis Bermudez for reading and commenting on earlier versions of this manuscript. Helpful comments were also made when versions of this paper were given at Washington University in St Louis in February 2004 and at an APA-Pacific session honoring Homes Rolston in March 2004. I am particularly in debt to Andrew Brennan and Yeuk-Sze Lo who, inter alia, argued that the N0 category was needed to fully work through Rolston’s position. NOTES 1 Indeed, in his response to Wenz, Rolston explicitly says that the chapter of Environmental Ethics on animals and ethics should be read in the light of his earlier chapter on following nature. 2 It would be spatially extravagant to explain how I have drawn these different uses of nature from Rolston’s work; so for now the accuracy of these uses will have to be taken on trust. 3 Questions about individual humans who do not manifest these distinctive species-specific human characteristics are sidestepped by Rolston’s focus on the norm for the species. Some humans may not manifest these distinctive characteristics; but they do not represent what is normal for the species and are not species-representative. The emphasis on species norms is also intended to protect Rolston against versions of the Argument from Marginal Cases. 4 Andrew Brennan (personal comm. 2004) drew my attention to this distinction in Rolston’s work—see Rolston (1988: 34). 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 200 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X 5 6 7 8 9 10 11 5/9/06 7:51 PM Page 200 CLARE PALMER I am not going to raise questions here about Rolston’s take on sentience and moral considerability, nor am I going to quibble at his use of the term “suffering” where one might prefer to use pain. This issue has been widely discussed elsewhere, and is not especially interesting in Rolston’s account. Hettinger rightly argues that, as constructed, Rolston’s principles here favor plants over animals, since plants, unlike animals, are governed by a principle of “non loss of goods”. Hettinger’s account here is interesting and useful; but I want to take a different direction in this paper, so I will not be pursuing Hettinger’s particular arguments more closely. P1 is not a principle without difficulties, though it is less problematic than P2 or P3. One legitimate worry (expressed in questions on each occasion when I have given this paper) might be as to whether there really is any N1 nature in existence; without N1 nature, P1 has no application. Rolston (1994: 72) develops the idea of “agricultural integrity,” where agricultural areas are “managed sustainably” such that their operation does not disrupt the surrounding natural systems” and they are “enveloped by natural systems”. Nonetheless, nothing about this account suggests that he would evaluate agriculture by any other standards than those he adopts for wild nature. Hettinger rightly notes that this seems to privilege plants over animals, and suggests a revision to Rolston’s principle by extending his Principle of Non Loss of Goods to animals from plants. See Hettinger (1994) for the detail of this argument. By indicating that humans deliberately created dependence, I am not meaning to rule out the thesis, most prominently argued by Budiansky, that (some) animals in some sense “colluded” with domestication. See Budiansky (1992) Brennan (personal communication 2004) quite rightly raises the question how far I am responsible for things that are problematic because of the behavior of other members of my species, not of me. I began to consider some issues of this kind in Palmer (2003) with respect to the feeding of feral pigeons in Trafalgar Square; but plainly this is a major area for further research. Of course, it might be argued that the moral problem lies in creating dependent animals in the first place, rather than in the ways they are treated once in existence, but I do not intend to propose this argument here. REFERENCES Budiansky, Stephen. 1992. The Covenant of the Wild. New York: Weidenfeld and Nicholson. Frey, R.G. 1983. Rights, Killing and Suffering. Oxford: Blackwell. Goodin, Robert. 1985 Protecting the Vulnerable. Chicago: University of Chicago Press. Hettinger, Ned. 1994. “Valuing Predation in Rolston’s Environmental Ethics.” Environmental Ethics 16:3–20. Jamieson, Dale. 1990. “Rights, Justice and Duties to provide Assistance.” Ethics 100/2: 349–362. Leahy, Michael. 1993. Against Liberation: Putting Animals In Perspective (Revised). London: Taylor & Francis. Lockwood, Michael. 1979. “Killing and the preference for life.” Inquiry 22: 157–190 Moriarty, Paul Veatch, and Mark Woods. 1997. “Hunting is not Predation.” Environmental Ethics 19: 391–404. Palmer, Clare. 2003 “Placing Animals in Urban Environmental Ethics.” Journal of Social Philosophy 34/1: 64–78. Regan, Tom. 1984. The Case for Animal Rights. London: Routledge. Regan, Tom. 2001. Defending Animal Rights. University of Illinois Press. Rolston, Holmes. 1986 Philosophy gone Wild. New York: Prometheus. Rolston, Holmes. 1987. “Beauty and the Beast: Aesthetic Experience and Wildlife.” in Decker and Goff, eds., Valuing Wildlife: Ecological and Social Perspectives. boulder CO: Westview. Rolston, Holmes. 1988. Environmental Ethics: Duties to and Values In the Natural World. Temple University Press. Rolston, Holmes. 1989. “Treating Animals Naturally?” Between the Species 5/3: 131–137. Rolston, Holmes. 1991. “The Wilderness Idea Reaffirmed.” The Environmental Professional 12: 370–377 Jamieson (1990) is not cited in text. Please provide citation. Vande Veer (1979) is not cited in text. Please provide citation Preston_11.qxd 5/9/06 7:51 PM Page 201 RETHINKING ANIMAL ETHICS 201 Rolston, Holmes. 1994. Conserving Natural Value. New York: University of Columbia Press. Rolston, Holmes. 1998. “Technology versus Nature: What is Natural?” Ends and Means 2/2 online at http://www.abdn.ac.uk/philosophy/endsandmeans/vol2no2/index.shtml Rolston, Holmes. 2003. “Life and the Nature of Life—in Parks.” In Harmon, David and Putney, Allen, eds., The Full Value of Parks: From the Economic to the Intangible. Lanham MD: Rowman and Littlefield, pp. 103–113. Singer, Peter. 1975; 1983 ed. Animal Liberation. Wellingborough: Thorsons. Singer, Peter. 1979. Practical Ethics. London: Cambridge University Press. VandeVeer, Donald. 1979. “Interspecific Justice” Inquiry 22/1: 55–70. Wenz, Peter. 1989. “Treating Animals Naturally.” Between the Species 5/1: 1–10. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44X Preston_11.qxd 5/9/06 7:51 PM Page 202