Carruthers and the Argument from Marginal Cases

Transcription

1 Journal of Applied Philosophy, Vol. Carruthers 18, No. and 2, the 2001 Argument from Marginal Cases 135 Carruthers and the Argument from Marginal Cases SCOTT WILSON ABSTRACT Peter Carruthers has argued that the present popular concern with animal rights is a sign of moral decadence. I argue against this conclusion by focusing on how Carruthers handles the notorious argument from marginal cases. I first explain what the argument from marginal cases is and how a theory can run foul of that argument. I then show that Carruthers must face this argument since he concludes that the marginal cases have, while animals lack, direct moral status. I then introduce a two-tiered moral theory as a way of accommodating the concept of direct moral status within Carruthers s framework. Although the concept of direct moral status that results from such a two-tiered theory has important differences from the standard conception, it does justify Carruthers s claim that the marginal cases have direct moral status. However, I argue that animals will likewise have direct moral status with this new conception, thus demonstrating that, even given Carruthers s theory, a concern for the rights of animals is no more a sign of moral decadence than a concern for the rights of the marginal cases is. I In his recent work, Peter Carruthers challenges many of the predominant views on the moral status of animals. In fact, Carruthers writes: I regard the present popular concern with animal rights in our culture as a reflection of moral decadence. Just as Nero fiddled while Rome burned, many in the West agonise over the fate of seal pups and cormorants while human beings elsewhere starve or are enslaved. [1] Carruthers reaches this conclusion due in part to his adherence to a contractualist moral theory. According to such a theory, Carruthers argues, animals will not have direct moral status. This is presumably why extending moral concern to animals while other humans live in such a deplorable state is a sign of moral decadence, for if animals have direct moral status, extending moral concern towards them would be morally required in many instances. I will argue that Carruthers is mistaken in concluding that extending moral concern to animals is a sign of moral decadence. My strategy will be to focus on how Carruthers handles the argument from marginal cases. I intend to show that if his response justifies attributing direct moral status to the marginal cases, it will likewise justify attributing that status to animals. In other words, Carruthers has not provided an adequate response to the argument from marginal cases, but has rather run headlong into it. I will conclude by considering the practical implications that follow from this argument., Blackwell Publishers, 108 Cowley Road, Oxford, OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

2 136 S. Wilson II I take the argument from marginal cases to be an argument schema that can be used in a variety of circumstances [2]. The form of that schema is as follows: (1) If we are justified in attributing moral property P to such marginal cases as the senile, the severely mentally handicapped, infants, etc., then we are likewise justified in attributing moral property P to animals. (2) We are justified in attributing moral property P to the marginal cases. (3) Therefore, we are justified in attributing moral property P to animals. This schema must be filled out in two ways. First, the property P in question must be specified. Second, the justification of the premises based on the property chosen must be offered. For example, suppose we use for P the property of being a rights-holder. In that case, someone that uses the above schema will need to explain why we are justified in attributing the property of being a rights-holder to animals if we are justified in attributing this property to the marginal cases. This explanation may or may not be the same explanation that would be used if P were something else, such as the property of having direct moral status. However, all versions of the argument from marginal cases have this same basic form. They are an appeal for consistency in our moral thinking, demanding that the attributions of moral properties to entities be made in a rationally consistent manner. There are two ways that an opponent of a particular version of this argument might respond. First, they could reject the claim that our being justified in attributing moral property P to the marginal cases implies that we are justified in attributing moral property P to animals [3]. Second, they might reject the claim that we are justified in attributing moral property P to the marginal cases [4]. A particular theory will run foul of the argument from marginal cases if and only if the theory claims that we are justified in attributing moral property P to the marginal cases, are not justified in attributing moral property P to animals, and cannot provide an adequate explanation for this disparity. III Carruthers must respond to the argument from marginal cases since he claims that we are justified in attributing some property to the marginal cases which we are not justified in attributing to animals. That property is the property of having direct moral status. Carruthers does not tell us in any detail what he takes this property to be, however, and I think that it is worth pausing for a moment to examine it. I will begin by examining the concept of direct moral status as it commonly occurs in the literature. I will then show how Carruthers s use of this concept is related to, but slightly different from, the concept as it is normally used. The differences will become especially apparent when we see how he argues for the claim that the marginal cases will have direct moral status on his theory. Warren offers a clear account of the concept of direct moral status as it is commonly used:

3 Carruthers and the Argument from Marginal Cases 137 To have [direct] moral status is to be morally considerable, or to have moral standing. It is to be an entity towards which moral agents have, or can have, moral obligations. If an entity has [direct] moral status, then we may not treat it in just any way we please; we are morally obliged to give weight in our deliberations to its needs, interests, or well-being. Furthermore, we are morally obliged to do this not merely because protecting it may benefit ourselves or other persons, but because its needs have moral importance in their own right. [5] Two important points about this concept are these: (1) If a being has direct moral status, then it is the kind of being towards which agents can have duties, and (2) if a being has direct moral status, at least some of the duties that agents can have to that being arise owing to the intrinsic properties [6] of that being [7]. We can understand this concept better by way of example. Consider first a case in which a man uses a sledge hammer to destroy another person s car. Such an action has a wrong-making feature. However, few of us would be inclined to judge that the car thereby has direct moral status. The reason for this is that we do not find the destruction of the car itself to constitute the reason why the action has a wrong-making feature. Rather, it is the effect on the person who owns the car that gives rise to this kind of reason. Compare this to a case in which a man takes a sledge hammer to another person, and then destroys (i.e., kills) that other person. Here we have a vivid illustration of the concept of direct moral status. For in a case like this, we do not need to see if the damage to the person constitutes a harm to someone else (although it most likely will do that as well), but we will rather conclude that the action has a wrong-making feature primarily due to the harm caused to the person destroyed. We can thus divide entities into at least two kinds: those with direct moral status and those with indirect moral status. The principle that separates these two groups seems to be a principle that focuses on the kinds of considerations that give rise to moral reasons. That is, a being belongs to the class of beings with direct moral status if and only if we can have at least some moral reasons to treat that being in a certain way due to the nature or intrinsic properties of the being itself. A being belongs to the class of beings with indirect moral status if and only if all of the moral reasons we have to treat that being in a certain way arise due to the relational properties of that being. IV In this section I will discuss how Carruthers uses the idea of direct moral status. In order to do this, we must first understand his basic argument for contractualism, and what he believes follows from accepting that theory. According to Carruthers, the common element in all versions of contractualism is that they deem morality to be a human construction, created by human beings to govern their relationships with one another in society. [8] This picture of morality is favoured by Carruthers because it avoids commitment to any kind of objectionable intuitionism about morality while at the same time it avoids the counter-intuitive results that we get from consequentialist theories [9]. He therefore accepts contractualism largely because it is the theory that best survives the process of reflective equilibrium [10].

4 138 S. Wilson He accepts a version of Rawls s contractualism [11] in which morality is seen as the system of rules that rational and self-interested agents would choose from behind the veil of ignorance [12]. Such agents know general facts about psychology, economics, sociology, etc., but do not know any particular facts about themselves, such as their age, sex, race, intelligence, etc. They are to choose rules to govern their behaviour in society from purely self-interested motives. However, since they will not know their race, sex, age, etc., they will not choose rules that give special consideration to any particular race, sex, age, etc. Rather, they will choose rules that protect, first and foremost, the autonomy of rational agents. They will choose these sorts of rules because they are self-interested, and the only thing they can be sure of about themselves is that they are rational and autonomous. We might then conclude that Carruthers s theory gives us an account of direct moral status that is similar to that described above. It would seem as if the properties that give one direct moral status are the properties of being rational and autonomous. Since this is so, then all and only rational agents will have direct moral status on Carruthers s account. However, things change when we consider how Carruthers handles the marginal cases. He recognizes that a theory that accords direct moral status to all and only rational agents will not be able to withstand the process of reflective equilibrium. That is because most of us think that the marginal cases have, at the very least, direct moral status. Any moral theory that cannot justify this thought may be too revisionary to be acceptable. So Carruthers is faced with the task of explaining why the marginal cases do have direct moral status on his account. He offers two reasons for why the marginal cases do have direct moral status according to his theory, but we only need to consider the first since they both have the same basic structure. Carruthers argues that if the contractors failed to accord the marginal cases direct moral status, then they will take their first steps on a slippery slope in which some beings that are rational and autonomous will be harmed [13]. The reason for this is that there is no sharp boundary between humans that are rational and autonomous and humans that are not rational and autonomous. Without such a sharp boundary, some people will inevitably judge that humans that are in fact rational and autonomous are not so, and then do things to those humans which would violate their rights as rational and autonomous beings. Since this fact is a general fact about human psychology, the contractors would know that it obtains. Since they do not know if they are among the rational and autonomous humans that would be judged to lack these qualities, they will adopt rules that grant protection to all humans, regardless of their rationality, etc. This reasoning is open to direct challenges. Someone might want to argue that it relies on an empirical claim which has been shown to be false, namely the claim that judging that some humans lack rationality will inevitably lead to the abuse of rational humans [14]. This is an interesting issue and the responses to be found in Carruthers to versions of these objections do not really go far enough [15]. However, I want to pursue another line of objection here. For I think that it is still not clear how the marginal cases have direct moral status on this account. In the next section, therefore, I will attempt to provide a way of justifying Carruthers s claim that the above line of reasoning will give the marginal cases direct moral status. In order to do this, I will be forced to modify the account of direct moral status slightly in order to

5 Carruthers and the Argument from Marginal Cases 139 accommodate Carruthers s argument. If I am correct about what this modified version of direct moral status is, then animals will likewise have that status. In other words, I hope to show that Carruthers has not given us any reason to believe that the marginal cases have direct moral status while animals do not. V The obvious objection to the above argument is that it does not really give the marginal cases direct moral status at all. In order for something to have direct moral status, it must have intrinsic properties which make it morally considerable in its own right. Carruthers argues that we must extend moral concern to the marginal cases in order to avoid unwanted consequences for the treatment of rational humans. This certainly does not sound as if he has attributed direct moral status to the marginal cases; it appears that he has attributed a complicated kind of indirect moral status to them instead. In this section I will propose a solution to this problem that will allow Carruthers to claim that his reasoning gives the marginal cases direct moral status. In doing this, I will make use of a two-tiered moral theory. I will then show how the concept of direct moral status can be modified to fit into such a theory, and will show how this new concept applies to the marginal cases. Two-tiered theories divide issues into first-order issues and second-order issues. The second-order moral issues are typically issues that centre on what the goal or underlying purpose of moral action is to be, what the proper objects of moral evaluation are to be, and what the criteria of moral evauluation are to be. This level of moral thinking is not tied directly to action. What I mean by this is that the reason we engage with second level issues is not to decide what to do at a given moment, but is rather to allow us to gain a better appreciation of what we are to aim at when it comes time to act. The general idea here is that before we can be good moral agents, we need to first know what it would take in order to be a good moral agent: we must define our target before we can possibly hit it. In order to do this kind of thinking, then, we need the time to carefully examine different possible theories and examples in order to gain a clearer insight into the goal of moral action. This is why this kind of thinking is not tied directly to action. We do not engage in this kind of thinking when confronted with a moral problem in our lives, for we would not have the time to do it properly in such a situation. Rather, we engage in this second level thinking when we have the luxury of time and the commitment to think things through to the end. That is not to say, however, that the second level thinking is severed completely from action. Rather, we use the results of our second-order moral thinking in order to inform the content of our first-order moral thinking. This first level can be roughly described as the practical level of moral thinking. The idea here is that once we have got clear on what we should be aiming at in moral action, we need to know the best way of hitting our target. When confronted with a real-life moral problem, we cannot spend the time necessary to think things through to the end, but we must rather have a set of rules and principles that have been derived from the second level of moral thinking. The rules and principles are guides to action, and will be relatively easy to apply in various situations, thus ensuring that we are better able to do the right thing more often than not.

6 140 S. Wilson The need for a two-tiered theory often arises from adherence to a consequentialist theory in which rightness is defined in terms of goodness. These kinds of theories often have counter-intuitive results, and one way of avoiding those results is by adopting a two-tiered moral theory. Critics have charged, however, that such a move is too ad hoc to be acceptable. However, R. M. Hare has shown, I think, that the need for a two-tiered system can be demonstrated independently of adherence to any form of consequentialism [16]. Indeed, the need for such a theory arises from the nature of the practical problems that we as moral agents situated in the world face. When we are faced with a moral problem, and it is time to decide how to act, we need a decision procedure that allows us to make decisions that have the highest possible chance of success (however success is to be defined here). We must often make these decisions without the luxury of time to examine all of the relevant features of the situation and determine all of the possible outcomes and alternatives open to us. Instead, we are faced with salient features of a situation, which we must then use to determine the course of action we should take. We also face an epistemological problem of determining what consequences each of the available acts open to us will have. Since we can never be certain of what exactly will happen if we are to do one thing rather than another, we must rely on rules that take probabilities of outcomes into consideration. Furthermore, we find ourselves tempted by such distorting influences as bias and selfinterest which can lead us to distort the way in which we represent the situation to ourselves. The first level of a moral theory should provide us with general rules for determining which of these salient features are the morally important ones and which ones take precedence over others, and its relative simplicity should provide the means to help us avoid being swayed by distorting influences. However, these rules and rankings cannot be simply given without any form of justification. The second level is used for explaining why certain features take precedence over others, and why certain features are relevant while others are unimportant. Such a two-tiered system helps to accommodate the two different positions we occupy: the first is a practical stance in which we must act, and the second is a more reflective stance in which we need to justify our practical decision making apparatus. Some comments of Carruthers lead me to believe that this division of moral issues into two tiers would be acceptable to him. He writes: There may be a variety of different levels to moral thinking. On the one hand there is the level of thought that manifests our settled moral dispositions and attitudes... but on the other hand there is level of theoretical reflection upon those dispositions and attitudes, asking how they may be justified by an acceptable moral theory. [17] Carruthers is echoing the sentiments that lead many to believe that we need a twotiered moral theory in which one level is meant to provide a set of rules and principles that are easily applied, and another level is needed to justify the selection of that set of rules and principles over others. These considerations suggest that Carruthers s theory fits nicely into such a twotiered theory. We can thus view the rules and principles that would be adopted by the contractors as constituting the first level of morality, while the second level of morality consists of the principles and considerations which lead Carruthers to accept

7 Carruthers and the Argument from Marginal Cases 141 contractualism, along with the reasoning that the contractors would use in selecting their set of rules. We can now slightly amend the understanding of direct moral status that was given above so that it can be applied to a two-tiered theory. Recall that we had claimed that a being has direct moral status if and only if it is the kind of beings towards which we can have moral duties, and the reason why we have at least some of those duties to that being arise due to the intrinsic properties of that being itself. Now that we have a twotiered system in place, we can change this account to the following: a being has direct moral status if and only if it is the kind of being towards which we can have moral duties, and the first-order reason why we have at least some of those duties to that being arises due to the intrinsic properties of that being. In other words, a being has direct moral status on this account if they have intrinsic properties that, at the first level of morality, are morally relevant in their own right. We can also develop a modified version of indirect moral status. A being has indirect moral status on a two-tiered moral theory if and only if agents can have duties to that being, but the first-order reason for each of those duties arising is that the being stands in some relation to a being with direct moral status. In other words, a being has indirect moral status on this account if our moral concern for them is mediated at the first level of morality. We can use these modified versions of direct moral status and indirect moral status in order to understand and justify Carruthers s argument concerning the marginal cases. Carruthers s argument can be seen as an argument for making the property of being human sufficient for being morally considerable at the first level of moral thinking. If it were not, and our first-order moral reasoning allowed us to determine if a particular marginal case were appropriately related to rational humans, we would end up abusing rational humans. The contractors would want to avoid this kind of abuse, and would therefore adopt a system in which the first-order reason we ought not to harm humans is simply that they are human. However, just because a being has intrinsic properties that are relevant at the first level of moral thinking does not imply that those same properties will be relevant at the second level. This is apparent if we consider Carruthers s argument for making the property of being human relevant at the first level of morality. He does not claim that the property of being human just is morally important and that the goal of morality therefore is to protect all humans. Rather, he maintains his original position that morality is best seen as the system of rules that rational, self-interested individuals would choose from behind a veil of ignorance. At the second level of morality, then, the relevant properties that make one morally considerable are the properties of being rational and autonomous. It just so happens that the best practical system to adopt in order to protect rational and autonomous beings is a system in which the property of being human is directly relevant to the assessment of actions. This account of direct moral status is close to the account I have called the standard conception. However, it differs in an important way. On the standard account a being must contain properties that are relevant at the second level of morality in order to have direct moral status. On the account sketched above, however, direct moral status can be obtained when a being has intrinsic properties that are relevant on the first level of morality, but are not relevant at the second level. What is intended in claiming that certain properties confer moral importance in their own right is that certain properties

8 142 S. Wilson a being has make that being morally important regardless of any relations that it has to other beings. This idea has not been captured in the account of direct moral status sketched above, and this is why this new account is still importantly different from the standard account. VI A natural objection to Carruthers s theory arises if we accept his conclusion that animals do not have direct moral status. If animals do not have direct moral status, but are more like inanimate objects, then it would seem perfectly legitimate for someone to torture a cat merely to receive pleasure from such an act. We think it is permissible for someone to destroy their property if they get a sense of satisfaction in doing so; if animals have the same moral status as inanimate objects, then why should we not say the same about them? Carruthers correctly believes that he must face this objection, for if he does not his theory could not survive the process of reflective equilibrium. Furthermore, he recognizes that he must provide a response that does not merely claim that these sorts of acts are wrong only if they either (a) destroy someone else s property or (b) cause some animal lover distress. What Carruthers s theory must accommodate in order to be successful is the intuition that it is wrong to torture animals for fun even if no one else owns the animal, and even if no one ever finds out about the act of torture. Carruthers responds to this objection as follows: Such acts are wrong because they are cruel. They betray an indifference to suffering that may manifest itself... in that person s dealings with other rational agents. [18] The idea that Carruthers is expressing here seems to be this. If I am the kind of person that enjoys causing a sentient being pain, then I have a morally defective character [19]. This morally defective character is dangerous for it may lead me to inflict suffering on rational humans. Since I have a duty to refrain from causing pain to rational humans, I likewise have a duty to alter this disposition and to refrain from performing acts that would reinforce such a disposition. In other words, the contractors are going to adopt rules that prohibit the senseless harming of sentient beings because a failure to do so will allow the wrong kind of dispositions to be cultivated in their society. There will be, then, a first-order rule that claims that it is wrong to torture sentient beings for fun. We now have to ask whether the property of being sentient is to be directly relevant at the first level of morality or if concern for sentient beings is to be mediated at that level. It is plausible that sentience should be directly relevant at the first level of morality. If it were not, there could be many situations in which people thought that they could harm just one animal without untoward consequences for rational humans. In other words, a kind of self-deception concerning character development would creep into first-order moral thinking. First-order moral thinking is meant to avoid just these kinds of problems. Its point is to guide people in such a way that they become better able to conform to the goal of morality. A system that allowed one to debate whether a particular act of harming an animal would have bad consequences for rational humans

9 Carruthers and the Argument from Marginal Cases 143 is a system that is more dangerous to rational humans than is a system that does not allow this. I thus conclude that sentience is sufficient for being morally considerable at the first level of morality. The second-order reason for this is that a first-order moral theory with this feature is able to protect rational humans better than a theory without such a rule. Furthermore, Carruthers seems to agree with this general line of thinking: It is important to stress that right action with respect to animals, on this character-expressive account, will generally be non-calculative. People who act out of sympathy for the suffering of an animal do not do so because they calculate that they will become better persons as a result. Rather, their actions manifest an immediate sympathetic response, and are undertaken for the sake of the animal in question. For this is what having the right kind of sympathetic virtue consists in. [20] However, if we accept the account of direct moral status according to which a being has that status if they have intrinsic properties that are relevant at the first level of morality, then it follows that animals will have direct moral status. Carruthers believes that this account does not yield direct moral status to animals because the concern for the animals is mediated ultimately by a concern for rational humans [21]. However, if what I have argued for is correct, this situation is structurally identical to the situation of the marginal cases. The rules that protect the marginal cases are accepted ultimately because those rules serve to protect the interests of rational humans. If this kind of mediation strips animals of direct moral status, then it will likewise strip the marginal cases of that status. Carruthers cannot have it both ways: if direct moral status requires rules that protect the interests of a being simply due to the intrinsic properties of that being without any form of mediation at all, then neither the marginal cases nor animals will have direct moral status. If, on the other hand, direct moral status is a matter of being protected by rules that are unmediated merely on the first-level of morality, then both animals and the marginal cases will have direct moral status. Carruthers has not provided us with any reason to believe that the marginal cases have direct moral status while animals do not. VII In this section I will argue that the line of reasoning that allows Carruthers to conclude that it is wrong to harm animals for trivial reasons will do more than imply that the contractors will adopt a rule to prohibit cruelty to animals. Rather, I think that that line of reasoning will lead to the conclusion that the contractors will adopt a variety of rules that require them to foster certain virtues and avoid certain vices. If I am right about that, then the implications for the treatment of animals under a contractualist moral theory are very different from what Carruthers thinks they are. It is important to remember why the contractors would adopt a rule that prohibits cruelty. It is tempting to claim that they would adopt such a rule because it is obvious that it is wrong to be cruel. However, such a claim is not something that a contractualist can use. The main virtue of contractualist moral theories is that they are supposed to provide us with an account of morality that begins with no preconceived ideas concerning

10 144 S. Wilson our moral duties. They are intended to show that rational contractors would adopt a set of rules to regulate their behaviour given certain initial conditions, and this is supposed to show that these rules are something that we are committed to as rational agents. For this reason, a contractualist cannot begin with the assumption that it is wrong to be cruel, and that this is why the contractors would adopt a rule forbidding such behaviour. Rather, the contractors would adopt such a rule because a system that contains such a rule would benefit them more than a system that did not. What this means is that there is nothing morally special about cruelty itself. What is relevant about cruelty is simply that a system that allowes people to act cruelly is a more dangerous system than a system that disallows it. If this is correct, then it is plausible to assume that the contractors will adopt a variety of rules that centre on the virtues and vices. In particular, it would seem as if the contractors would adopt rules that would require them to foster such virtues as beneficence, generosity, and the like. The reason for this is simply that a system that contains such rules would benefit the contractors more than a system that did not. Consider Kant s example of the duty to help people in need [22]. He argues that it is not possible to consistently will that no one will ever help anybody that needs help, while at the same time willing that you are able to achieve the means necessary to achieve your ends [23]. The reason for this is simply that there will inevitably arise a time in which it is necessary for others to help you in order for you to achieve your ends. Since this is so, it follows that acting on the maxim to never help anyone that needs help is practically inconsistent. The contractors would be aware of these considerations. They will certainly know that it is reasonable to believe that almost everyone will, at some point or other, need the help of other people. Since they know that this is true, they will want to adopt a system of rules that will enable them to receive such help when they need it. What is likely to get people to help you out in times of need is for the people that are in such a position to be sympathetic, beneficent, and virtuous people. Since this is so, the contractors would adopt a system of rules in which there is a moral imperative to foster these other-regarding virtues, and not merely an imperative to avoid such otherdisregarding vices as cruelty. In the next section, I will show how this result affects Carruthers s position regarding the treatment of animals. VIII The practical implications of my argument are limited, but significant. Let me first explain the limitations. My argument does not imply that contractualists are committed to claiming animals and humans share the same basic rights and protections, or that we should give the same weight to the interests of animals as we do to humans. Rather, the conclusion of my argument is simply this: if Carruthers is going to claim that the marginal cases have direct moral status, then he should likewise claim that animals have that same status as well. My position is compatible with seeing our duties and obligations being divided into spheres of concern, to use an image borrowed from Mary Midgley [24]. To borrow a phrase from Midgley, my main concern has been to prove that if that the marginal cases matter to Carruthers, then so should animals [25].

11 Carruthers and the Argument from Marginal Cases 145 Such a conclusion necessarily has certain limitations in practical application. For instance, my conclusion will not allow us to definitively settle debates about when, if ever, it is permissible to experiment on living animal subjects. However, my conclusion does provide a criterion of adequacy for anyone who wants to try to settle such a debate. For anyone who attempts to settle the debate without giving any weight to the interests of the animals to be used has made a mistake. They have overlooked an important piece of data in their calculations, and so their conclusions cannot be complete. This brings us to what I think is significant to my conclusion. Carruthers believes that his position is consistent with claiming that we are permitted to raise animals on factory farms, use them for experimental subjects, place them in zoos and rodeos, without ever considering how these actions affect their welfare. However, if what I have argued for above is correct, then Carruthers is quite mistaken. The reason for this is simple. The virtues of sympathy, generosity, and beneficence are incompatible with a practical attitude in which the interests of an entire segment of the world s population are simply ignored. It is not possible to be a sympathetic person while at the same time neglecting to ever consider how certain actions affect others around you. It may be objected at this point that it is possible to be sympathetic to humans while being entirely unsympathetic to animals [26]. However, I have serious doubts about such a possibility. My doubts arise primarily owing to the way in which we come to believe that any kind of being at all, whether it be animal or human, has conscious mental experiences [27]. When we come to have these sorts of beliefs, we reason by way of analogy. The closer a being is to us (e.g., behaviourally, anatomically, and evolutionarily), the more we believe that the being has conscious mental states. One of the things we do when we reason this way is imaginatively identify with the being we are thinking about. For example, one of the reasons I believe that a dog can feel pain while a rock cannot is that I can imagine what it would be like for a dog to be hit with a heavy object, but cannot imagine what it would be like to be a rock in the same position. This process is the same regardless of what kind of being we are reasoning about. What this means is that if we find that the interests of an animal leave us completely cold, then we are the kind of person who has trouble identifying with others interests. In order to be sympathetic, we must be able to identify with the interests of others, regardless of their species. I think that this kind of reasoning is the best way to understand what Kant meant when he said that we can judge the heart of a man by his treatment of animals. [28] Human virtues are not divided according to how they apply to species, but are rather more unified than this objection under consideration assumes. Whether this means that a contractualist must conclude that practices such as factory farming or the use of animals for experimental purposes are wrong is not something that I am in a position to conclude right now. I hope to have shown, however, that these are open questions that contractualists must face more directly and with more attention to the particular facts about these practices than people such as Carruthers have done in the past. I have been mainly concerned with disarming a very powerful and common objection aimed at those of us that do care about the welfare of animals and believe that others should care as well. If animals are not the kinds of beings that are warranted such concern, then people such as myself will have been making a moral mistake of our own in insisting that others should care about the welfare of animals [29].

12 146 S. Wilson By showing that even contractualism, the theory that is perhaps the most unfriendly theory to animals, requires us to care for the welfare of animals, I have cleared the way for the more specific debates concerning the proper treatment of animals to proceed without the fear of such a charge. Scott Wilson, University of California, Santa Barbara, Department of Philosophy, Santa Barbara, CA wilson@umail.ucsb.edu NOTES [1] CARRUTHERS, P. (1992) The Animals Issue (New York, Cambridge University Press), p. xi. [2] For a book-length discussion of the Argument from Marginal Cases, see DOMBROWSKI, D. (1997) Babies and Beasts: The argument from marginal cases (Chicago, The University of Illinois Press). [3] As we will see, this is the strategy favoured by Carruthers. [4] This strategy is favoured by R. G. Frey. See FREY, R. (1980) Interests and Rights: The case against animals (Oxford, The Clarendon Press), chapter III. [5] WARREN, M. (1997) Moral Status (Oxford, Clarendon Press), p. 3. [6] I understand an intrinsic property of x to be a property that x (logically) could have if x were the only thing that existed. [7] Another phrase commonly used is moral patient. I prefer direct moral status because it allows us to construct the useful contrasting phrase indirect moral status. For another discussion of the idea of direct moral status see, DEGRAZIA, D. (1996) Taking Animals Seriously: Mental life and moral status (New York, Cambridge University Press). [8] Carruthers, P. (1992), p. 36. [9] Carruthers takes particular aim at the theories of Singer and Regan. For Singer s views, see SINGER, P. (1990) Animal Liberation, New Revised Edition (New York, Avon Books) and SINGER, P. (1994) Practical Ethics, Second Edition (New York, Cambridge University Press). For the views of Regan see REGAN, T. (1983) The Case for Animal Rights (Los Angeles, The University of California Press), and REGAN, T. (1982) All That Dwell Therein: Animal rights and environmental ethics (Los Angeles, The University of California Press). [10] Carruthers discusses his use of Reflective Equilibrium in CARRUTHERS, P. (1992), chapter 1. He uses an account that is largely borrowed from Rawls. See RAWLS, J. (1951) Outline for a decision procedure in ethics, Philosophical Review 60, pp An interesting discussion of Reflective Equilibrium and its uses in the animals issue can be found in DeGrazia, D. (1996), chapter 2. [11] See RAWLS, J. (1971) A Theory of Justice (Cambridge, MA, Harvard University Press). [12] Carruthers actually discusses the theories of both Rawls and Scanlon. He wants to accept what he thinks is the core of the contractualist theory without committing himself to idiosyncratic elements that arise from a particular formulation of that theory. I have chosen to use the simplifying procedure of using only Rawls s theory. Nothing of importance hinges on this decision. For Scanlon s version of contractualism, see SCANLON, T. (1982) Contractualism and utilitarianism in Sen and Williams (eds.) Utilitarianism and Beyond (New York, Cambridge University Press), pp [13] Carruthers, P. (1992), pp [14] DeGrazia, D. (1994), takes up this line of response. [15] Carruthers s responses can be found in Carruthers, P. (1992), chapter 5. [16] HARE, R., Moral Thinking: Its levels, method, and point (New York, Oxford University Press, 1981), part I. [17] Carruthers, P. (1992), p [18] Carruthers, P. (1992), pp [19] Carruthers is actually faced with a problem with his response at this point. For he spends one chapter of his book arguing that animals cannot have any conscious experiences at all. If that is correct, then it is hard to see why causing damage to an animal should lead one to have a bad character. If he truly believes that animals cannot have conscious experiences, then he seems forced to admit that there is nothing wrong with intentionally causing them damage for fun. I will not pursue this line of objection here.

13 Carruthers and the Argument from Marginal Cases 147 For Carruthers s views on the experiences of animals see Carruthers, P. (1992), chapter 8; and Carruthers, P. (1989) Brute experience, The Journal of Philosophy 96, pp [20] Carruthers, P. (1992), p [21] Ibid. [22] Kant, I. (1964) Groundwork of the Metaphysics of Morals, H. J. Paton (trans.), (San Francisco, Harper Torchbooks), pp [23] This interpretation of Kant s reasoning about this case is not entirely non-controversial. I have been greatly influenced by the writings of Christine Korsgaard on this matter, and believe that the interpretation I give of the example owes much to her. See KORSGAARD, C. (1996) Kant s Formula of Universal Law, in Korsgaard, Creating the Kingdom of Ends (New York, Cambridge University Press), pp See also, KORSGAARD, C. Kant s analysis of obligation: the argument of Groundwork I, in Korsgaard, C. (1996), pp [24] MIDGLEY, M. (1983) Animals and Why They Matter (Athens, GA, The University of Georgia Press), pp [25] Midgley, M. (1983). [26] Carruthers seems to argue for this point in Carruthers, P. (1992), pp [27] For more on this issue, see my Indirect Duties to Animals (forthcoming). An objection to this line of thinking can be found in RICHARD HOLTON and RAE LANGTON, Empathy and animal ethics, in Jamieson, D. (ed.) (1999) Singer and His Critics (Malden, Ma:, Blackwell Publishers), pp [28] KANT, I., Duties in regard to animals, in Regan and Singer (eds.) (1989) Animal Rights and Human Obligations (Englewood Cliffs, NJ, Prentice Hall), p. 24. [29] This is essentially the charge that Carruthers makes when he says that a concern for animals is a sign of moral decadence.