The Moral Status of Animals and Their Use in Research:A Philosophical Review
Times have changed. Twenty years ago discussing the moral status of animals probably would have qualified one as a kook. Today no moral philosopher can evade the subject. But despite increased attention to ethical issues involving animals, nothing approaching a societal consensus on their moral status has emerged. Opinions currently range from the view that the lives and welfare of animals are as important as those of humans, to the view that animals have no moral status. Thus, while ethical discussions concerning human subjects of research, for example, are quite refined—resting on substantial agreement about matters such as the importance of informed consent—academic debates about animals are at a more rudimentary stage.
In this article I offer a philosophical review of (1) leading theories of the moral status of animals, (2) pivotal theoretical issues on which more progress needs to be made, and (3) applications to the setting of animal research. Such an examination demonstrates, I believe, that the practical implications of leading theories converge far more than might be expected. In addition, I hope this review helps to clarify particularly troubling issues that remain so they can be treated adequately.
General Characterization of the Debate
The philosophical debate concerning animals is anomalous for a variety of reasons. First, the ethical theories underpinning the dominant views are polarized to an unusual degree: two of the contributors most commonly cited—Peter Singer and R. G. Frey—are among the purest utilitarians in philosophy; the theory of the other—Tom Regan—features rights that are nearly absolute. Their positions therefore run counter to the current trend of trying to bridge the gap between utilitarianism and rights theories (see, [End Page 48] e.g., Griffin (1986), Sumner (1987), and Beauchamp and Childress (1989)) or at least to modify a version of one to bring it normatively closer to the other.1
Also striking is the fact that there is no well-developed theory explicitly addressing the moral status of animals that supports such current practices as factory farming, animal research, and hunting. No philosopher who has developed his or her views to the point of publishing a book on the subject has vindicated the status quo. Michael A. Fox did write a book calling for only modest reforms in current animal research practices (Fox 1986), but his argumentation was severely criticized. Within a year, he recanted his views and joined those opposing the status quo (Fox 1987). Widely perceived to be a staunch opponent of the animal welfare movement, R. G. Frey is often invited to conferences as the sole opponent of Singer, Regan, and others considered radically proanimal. Yet while Frey vigorously opposes Regan's argumentation for animal rights, his own argumentation suggests he is almost an antivivisectionist (see, e.g., Frey (1987a)). This surprising clustering of the leading theorists on the side of animal welfare changes the meaning of "radical," "moderate," and "conservative" as one moves from society at large—which generally accepts meat eating, for example—to the academic arena of animal ethics.
Some will no doubt argue that I note this convergence too quickly, that I have overlooked lesser known philosophical efforts that attempt to justify more conservative positions on these issues. They will most likely point to articles by Carl Cohen and H. J. McCloskey, whose positions I will briefly summarize later, indicating why I do not think they represent significant contributions.
Another distinctive feature of this debate is a relative dearth of rigorous, sustained philosophical exploration. Not enough is done in the way of conceptual analysis, moral epistemology, the philosophy of mind, the philosophy of science, and so on. (One understandable reason for this is a desire on the part of some writers to reach a much wider audience than academic philosophy.)2 In my opinion only five authors have made a significant philosophical contribution to the endeavor of placing animals in ethical theory: Singer, Frey, Regan, Mary Midgley, and S. F. Sapontzis.3
The First Generation: Singer, Frey, and Regan
What I call "the first generation" of theories consists of the views of Singer, Frey, and Regan. Despite their differences, they are all, in an important sense, philosophically mainstream. First, working in the tradition [End Page 49] of liberal individualism, their moral focus is on the individual, whether as rights-bearer or as bearer of interests to be counted in utility-maximization. Thus they are not very receptive to community-based approaches, for example, which ground obligations in social relations more than in characteristics (such as sentience) possessed by an individual. Second, this group has great confidence in reason as the major, if not sole, arbiter of ethical disputes. Thus they abstain from—and are even hostile to—appeals to emotion in ethics. (Indeed, Singer and Frey go further and part ways with most of our tradition by castigating even the most cautious employment of moral intuitions.)4 In a related way, the first generation never seems to question the systematic approach to ethics, according to which once the right ethical theory is discovered, specific normative answers simply follow from it. Of course, some doubt the possibility of discovering such a comprehensive system because of significant objections to every major system in contention.
As we will see, these shared philosophical values have affected the shape of the debate. I would even argue that their orthodoxy causes them to underappreciate the contributions of what I call "the second generation." In any event, let us turn to the views of this first generation.
More than any other work, Peter Singer's Animal Liberation (1975) incited the philosophical debate about the moral status of animals. Singer presents his arguments with exceptional clarity. But the book, which was not written for a philosophical audience, lacks some depth; it races over or skips many philosophical issues treated in greater detail by other philosophers.
In Animal Liberation —and more explicitly in other writings—Singer argues from the perspective of act utilitarianism with a value theory consisting of preferences. Accordingly, everyone's good counts equally, and individual actions are justified by direct appeal to the standard of maximizing the good.5 (By contrast, rule-utilitarians justify actions by appeal to rules such as "do not lie." These rules, in turn, are justified by their tendency to maximize the good.) As a preference-utilitarian (Singer 1979, p. 12), Singer argues that the satisfaction of actual preferences is what is to be maximized. Singer uses the more common term "interests" in Animal Liberation, but he views all interests as preferences.
Singer argues on the basis of a combination of behavioral, physiological, and evolutionary evidence that animals have interests—at the very least an interest in not suffering (Singer 1975, pp. 11-16). Indeed he identifies the capacity to suffer as the basic admission ticket to the moral arena: [End Page 50]
The capacity for suffering and enjoyment is a prerequisite for having interests at all, a condition that must be satisfied before we can speak of interests in a meaningful way.
Thus beings that lack sentience (the capacity to suffer and experience enjoyment) need not morally be taken into consideration, except as their treatment affects sentient beings.
From all this it follows that sentient animals, which he thinks include at least all vertebrates (Singer 1975, pp. 185-88), must be given equal consideration in ethical decision making. This does not entail precisely equal treatment. Dogs have no interest in voting, so equal consideration does not necessitate granting them the right to vote if we grant such a right to humans. However, it does mean that if a human and a mouse suffer equally in intensity and duration, their suffering has the same moral weight. Consequently, trampling on animal interests—in activities such as fur trapping and factory farming—for relatively marginal increases in our interests cannot be justified.
Singer's views have several difficulties, including the following: (1) the vulnerability of act utilitarianism to various objections (especially concerning justice), which has led or kept most moral philosophers away from this theory; (2) its difficulty with "marginal cases" (the problem that some humans are similar to animals in what are taken to be the morally relevant features, e.g., intelligence and sentience), as discussed below; and (3) the poverty of its value theory, which deems nothing valuable unless it is the object of someone's actual preference. One implication of (3) is that animals who lack the concept of life and therefore do not prefer to continue living are in no way harmed if killed painlessly in their sleep, even if in the midst of a flourishing sentient life.
One of the great merits of R. G. Frey's provocative Interests and Rights(Frey 1980) is the very thorough argumentation that Frey gives in conceptual analysis, ethical theory, metaethics, and the philosophy of mind. I suggested above that most scholars in animal ethics shy away from the "nitty-gritty" of philosophical problems, especially problems outside ethics. Frey is the most notable exception.
Like Singer, Frey is a preference act utilitarian.6 Unlike Singer, Frey contends, at least in this book, that animals have no interests. Although his views have now changed, the argument he presented proceeds as follows. First, he defends the view that the only morally relevant interests are desires (preferences). Then he contends that one cannot have a desire (to own a book, for example) without a corresponding belief (that I lack [End Page 51] the book or that the statement "I lack the book" is true). And he argues that a belief—which always amounts to a belief that a certain sentence is true—requires language. He concludes that since animals lack language, they lack the beliefs requisite for desires and therefore lack desires (Frey 1980, ch. 7). Thus animals have no interests.
It is more difficult to refute Frey's arguments than to disagree with his counter-intuitive conclusions. It is very difficult to believe that kicking a cat does not cause the cat to suffer and does not frustrate her interests. It is not surprising, then, to find considerable philosophical strain at the end of his book when he avoids the word "suffering," which he apparently links to "interests." Frey states instead that animals can experience "unpleasant sensations" and that wantonly causing unpleasant sensations is wrong (Frey 1980, pp. 170-71). It would seem difficult to argue that it is not in one's interests to avoid unpleasant sensations. It is also difficult to believe that beings who can have unpleasant sensations cannot suffer. Apparently, Frey agrees, because he now seems to allow that animals suffer (see, e.g., Frey 1987b), and that they have interests. (I base the latter attribution on numerous conversations with him.) But even if he has rejected some of his earlier conclusions, others might find the arguments compelling so it is worth confronting them on more than the intuitive level.
Most of the premises of his argument summarized above could be challenged, but the most vulnerable may be the claim that beliefs are always beliefs that a certain sentence is true. If this can be refuted,7 then animals can be acknowledged to have beliefs, leaving no obstacle to attributing desires (preferences), and therefore interests, to them. Frey also faces objections to act utilitarianism and problems with marginal cases.
Seizing on these problems, Tom Regan argues for an alternative. His painstakingly thorough The Case for Animal Rights(Regan 1983), which all things considered may be the best book in the field to date, begins by rejecting utilitarianism. Regan argues that because it is committed to maximizing the good with no prior commitment to how the good is to be distributed, utilitarianism fails to respect what has come to be called the separateness of individuals.8 If slavery is wrong, it is only because the practice fails to maximize the good, not because of the inviolability or dignity of persons. It is not even clear that the carefully concealed, painless killing of one unconsenting person to retrieve organs to save two other people is wrong based on utilitarian reasoning. Believing that our carefully considered moral intuitions have a significant place in ethical reasoning, [End Page 52] Regan cites the counterintuitive implications and methodology of utilitarianism in rejecting that theory.
Regan goes on to propose that we regard individuals as possessing equal inherent value. Who are "individuals?" They are beings to whom the ascription of inherent value is meaningful because they have a welfare. The notion of a welfare involves that of faring well (or badly) over time. Thus animals who have beliefs, desires, and a psychophysical identity over time—so-called "subjects-of-a-life"—have inherent value. This includes at least normal adult mammals. Inherent value implies a basic Respect Principle, which in turn implies the Harm Principle, that "subjects" are not to be harmed. (It is significant that Regan includes both inflictions and deprivations as harms. Therefore, death, which deprives one of the opportunities that life holds, is a harm whether or not one has a concept of life.) A full examination of the Respect Principle leads to the thesis that "subjects" have rights that are not to be overridden in the name of the common good, except in rare, carefully circumscribed circumstances.
Regan's theory has its weaknesses; his strong rights view has failed to convince many philosophers. It would not allow killing a small number of rats in research in a promising attempt to fight a raging epidemic, for example. He takes this view even though failing to kill the rats might show disrespect for those who would eventually die from the epidemic. At the same time, the fact that Regan specifies situations in which rights can be overridden (e.g., when every rightholder on a lifeboat will drown if none is sacrificed) might be incompatible with his strong interpretation of the notion of equal inherent value. This point is well made in Jamieson (1990).
The Second Generation: Midgley and Sapontzis
The first generation of animal ethics philosophers proffered some carefully worked out theories of the moral status of animals. Subsequent work has typically used these theories as a point of departure. I think the contributions of the philosophers in the second generation, Mary Midgley and S. F. Sapontzis, have been underappreciated for two reasons: (1) the almost habitual scholarly focus on the first generation; and (2) the fact that the debate tends to be carried out in terms encouraged by the philosophical orthodoxy I described previously.
The less orthodox flavor of the second generation is most evident in the fact that theirs are mixed theories with no visible overarching framework. Neither is a utilitarian exactly, nor a committed rights theorist, nor for that matter a virtue theorist. Neither can formulate his or her theory in [End Page 53] the form of a single principle or a neat set of principles with clear relations between them. They also have considerably less confidence in reason as the arbiter of ethical disputes, displaying a somewhat more pragmatic bent. And evincing great impatience with Frey-like skepticism about animal interests, the second generation invests somewhat less in issues concerning the philosophy of mind.
Much of Mary Midgley's Animals and Why They Matter(Midgley 1984) is devoted to discrediting the view that animals are morally unimportant. In this endeavor she profitably distinguishes "absolute dismissal" from "relative dismissal." The latter allows animal interests to count, but only after all human interests are satisfied. In arguing against absolute dismissal, she stops short of entirely rejecting the idea that the needs of those closer to us have moral priority over the needs of those less close. Indeed, in a qualified endorsement of such a perspective, she departs from the individualist mainstream and invokes social-bondedness as morally paramount (while revealing her antirationalist sentiments):
The special interest which parents feel in their own children is not a prejudice, nor is the tendency which most of us would show to rescue, in a fire or other emergency, those closest to us. We are bond-forming creatures, not abstract intellects.
By way of analogy she goes on to argue that a preference for our own species is acceptable within limits, in no way justifying either complete dismissal or relative dismissal. (For example, she deplores the current practice of factory farming.) So instead of painting a picture of the moral concerns of family, kin, species, etc., as forming concentric circles with "me" in the middle, she portrays them as overlapping concerns.
Midgley's discussion of "rationalism"—a term she uses extremely broadly to refer to all traditions that emphasize reason and downgrade emotion—is of interest both in motivating her normative view and in its own right. In criticizing "rationalism," she apparently is attacking both (1) the view that reason is the basis for moral status, and (2) views that regard intellect, as opposed to feelings, as the proper guide in ethics. She argues, for example, that the notion of rights is so unclear as to be almost completely unhelpful. This point is well made in Russow (1985, p. 173).
Midgley's arguments bring theoretical fresh air into the debate and provide a novel account for very stubborn convictions concerning the strengths of our obligations to family, neighbors, members of the same species, etc. Still, I think her proposals have the following difficulties: (1) [End Page 54] consideration of long-term utility may be equally capable of explaining such convictions, vitiating one of the strongest supports for her approach; (2) it is never successfully explained why racism could not be justified along the lines of her defense of giving priority to those "closer" to us; (3) the relation between reason and emotion is never clearly explicated (Russow 1985, p. 174); and (4) the normative suggestions are very vague and leave no clear guidelines for the treatment of animals, so that giving animals very little consideration does not seem to have been precluded (even if the book's tone suggests otherwise).
A somewhat clearer normative position is taken by S. F. Sapontzis in Morals, Reason, and Animals(Sapontzis 1987). Eschewing attempts to ground ethics in ahistorical, transcendental norms, Sapontzis regards ethics as a pragmatic endeavor rooted in cultural traditions but capable of development or progress within a tradition. He contends that while the Western tradition does not question our casual consumption of animals, "there are fundamental elements of that morality that point in the direction of animal liberation" (Sapontzis 1987, p. 110). In view of the lack of a clearly dominating ethical theory, and suspicious of relatively simple frameworks (like utilitarianism and Regan's rights view), Sapontzis treats three fundamental goals of our moral tradition as on a par in addressing ethical issues. He concludes:
Liberating animals from our routine sacrifice of their interests for our benefit would seem to be the right thing to do in order better to pursue our moral goals of developing moral virtues, reducing suffering, and being fair.
Thus his theory is, in the end, an amalgam of considerations of virtue, utility, and rights.
Although Sapontzis argues in favor of extending rights to animals who have interests, he does not settle for simple solutions. Perhaps surprisingly, he rejects, in Midgley-like fashion, Singer's idea that we owe all animals with interests equal consideration:
In common morality we are not under an obligation to give equal consideration to everyone. On the contrary, we are . . . even obligated to give priority to the interests of families, friends, colleagues, and compatriots.
In spite of this inegalitarian move, Sapontzis ends up condemning current animal-consuming practices, although (as I note below) he allows for very [End Page 55] limited use of animals in research. Along the way, and in taking up further topics, there are many illuminating discussions that display both subtlety and originality.9
An excellent, underappreciated work, Morals, Reason, and Animals nevertheless has its shortcomings. The argumentation is unevenly rigorous. Sapontzis's discussions of, for example, the replacement argument, the environment, and the disvalue of death are excellent. His discussion of the limits of reason, however, often seems irrational, while his treatment of the problem of marginal cases seems based on a logical misunderstanding. Also, while his disenchantment with monolithic ethical theories is understandable and his pluralistic approach attractive, the latter is no stronger than the basis he provides for it: our moral tradition. Why assume that our moral tradition is ethically in the right? If one turns to it for fear that there is no other plausible foundation, there remains this problem: every exhortation to "progress" invokes normative standards that either (1) are independent of that tradition, ensnaring the metaethic in self-contradiction, or (2) leave us no reason to respect them.
Some Outstanding Theoretical Issues
Progress on ethical issues involving animals could be facilitated by clarifying important theoretical issues that have thus far resisted adequate treatment. Let us turn, then, to three of them: equal consideration; the value of life; and marginal cases.
Assuming that the interests of animals are morally important, do they have as much moral weight as human interests (as both the utilitarians and Regan assume), or do they have less weight, that is, deserve less consideration (as Midgley, for example, has maintained)?10 This issue, though rarely discussed explicitly, is of the greatest importance in determining our moral relationship to animals.
It has been argued by Singer and R. M. Hare that equal consideration is a formal requirement of morality, that the very concept of morality, or the logic of moral language, includes this requirement.11 This is surely false. Would a putatively moral system that stipulated that everyone's interests counted equally, except for Buddha's, which counted twice as much as the others, not qualify as a moral system?
On the other hand, the principle of universalizability seems to establish a presumption in favor of equal consideration. This principle states that if one makes a certain judgment in a particular set of circumstances, one must make the same judgment in circumstances that are similar in relevant [End Page 56] ways. The upshot is that anyone who claims that the interests of animals have less weight than human interests must produce a relevant difference between them. What makes this issue so difficult is having to determine what differences are morally relevant.
One could try to meet this burden of proof in a number of ways. I will discuss only two. As we have seen, one possibility is to cite degrees of social-bondedness, or special relationships, as underlying differences in how much consideration to give the interests of others. Midgley and Sapontzis would justify some greater degree of consideration for other humans based on the argument that family members, friends, and so on deserve different degrees of moral consideration. To make a plausible case, a proponent of this view must show at least how these arguments do not support what seem to be unjust forms of discrimination, such as racial discrimination. One must also say more to the utilitarian who vindicates giving varying degrees of moral attention to members of different groups, but only on the basis of long-term efficiency (since, for example, children will generally be better taken care of if we care for our own).
Perhaps the most troubling challenge to equal consideration is what I call the "sui generis view"—a view that is almost never taken seriously by philosophers yet seems to capture the convictions of many people. Justifications for what counts as a morally relevant characteristic, the argument goes, must ultimately depend on an undefended assumption. Utilitarians typically conclude that having interests is the relevant trait—on the assumption that the essential job of morality is to promote interests. But while they sometimes argue for this assumption, such arguments never amount to a rigorous proof. Many discussants invoke traits like sentience, self-consciousness, etc., but no logical demonstration is ever offered that species membership per se is not morally important. Intuitively, it seems relevant. Further, its relevance would explain the tendency to think that our duties to humans are stronger than our duties to animals and would dissolve the problem of marginal cases, which I discuss later.
In spite of some attractions, the sui generis view seems vitiated by the fact that universalizability establishes a presumption in favor of equal consideration—so that a convincing argument, not just a stalemate, is needed to defeat equal consideration.12 (To my mind no such argument has been found.)
The value of life
Another issue on which a great deal turns concerns the value of life. What sort of being must one be for one's life per se, as opposed to the quality of the experiences it contains, to have moral weight? [End Page 57] Does the act of killing an animal painlessly in its sleep have any disvalue by itself? The U.S. Public Health Service, in its Guide for the Care and Use of Laboratory Animals, suggests not (Department of Health and Human Services, 1985). It regards the infliction of suffering—but not: killing—as demanding justification.13 How one answers affects how one is likely to judge hunting, which often entails little or no suffering, as well as raising meat in nonintensive family farms. Let me sketch three rough positions on this issue of killing.
One answer, implied by the brand of utilitarianism advocated by Singer and Frey, is simple: all and only beings who have a preference to continue living (which requires having a conception of oneself as living) have lives with moral weight.14 Only the satisfaction of preferences is valuable; if one does not prefer X, removing X does not cause harm. Probably all animals show a tendency to struggle for survival, but if it is instinctual and unaccompanied by particular mental states, then it fails to reveal a preference to live. Perhaps then, the members of few species are harmed by death per se, although identifying them, using this criterion, poses a formidable challenge to the philosophy of mind and animal psychology.
Of course, if not all interests are thought to be actual preferences, this view is less attractive. Consider a second position: the lives of all and only morally considerable beings (usually beings with interests) have value, but these lives may differ greatly in value.15 Moreover, the mentally more complex being generally has the more valuable life. Different theorists explain this judgment in different ways, but I do not think anyone has explained it in sufficient depth. Regan's account captures the spirit of the central point. In defending the sacrifice of a dog over a normal human if one of them must be thrown off a lifeboat to prevent all aboard from drowning, Regan argues that death is less of a harm for the dog. Death is a harm because it forecloses all opportunities one has for obtaining the satisfactions available to members of one's species, and more opportunities are closed off in the case of a human than in the case of a dog (Regan 1983, p. 324). Rachels (1983, p. 254) similarly argues that "when a mentally sophisticated being dies, there are more reasons why the death is a bad thing."16 I have contended that human life generally "has more interests and more very important interests [than animal life has], all of which are thwarted by death" (DeGrazia 1989, p. 99).
A third view reflects the hesitation of some philosophers to value lives differently, and in particular, to assign superior worth to normal human lives. In what way are more opportunities—or more interests—thwarted [End Page 58] by a human death than by an animal death? Why are there more reasons that human death is an evil? Pressed to explain, proponents of this view might say that while humans have the kinds of experience that animals have, they also have other kinds of experiences out of the reach of animals, for example, the pursuit of life projects and the cultivation of deep personal relationships. In response, Sapontzis notes that animals have a multitude of experiences that humans cannot appreciate:
We cannot enjoy the life of a dog, a bird, a bat, or a dolphin. Consequently, we cannot appreciate the subtleties of smell, sight, sound, and touch that these animals can apparently appreciate. Here we are the boors.
A common rejoinder switches emphasis from quantity to quality: some of the riches of typical human lives that are not shared by animals have special weight or value, and therefore, human lives are generally more valuable than the lives of animals.17 Sapontzis retorts that we cannot know the value of an experience unless we have had it, so we have no business arguing that our experiences, and therefore our lives, are morally weightier than those of other animals (Sapontzis 1987, p. 219). Suffice it to say that the issue is unresolved.
Perhaps the most agonizing of the unresolved issues is the problem of marginal cases. This is a problem only for those who, unlike Regan, believe that we may sometimes harm animals for human benefit. Medical research probably offers the strongest case. If we may harm animals in this way, then either (1) we may likewise harm humans who are similar to animals in relevant ways, or (2) we must explain why no humans are similar in relevant ways to the animals we are justified in harming. The latter is difficult. If we refuse to conscript humans for involuntary harmful research, what do all humans have that animals lack? Obviously not all humans are rational, self-conscious, autonomous, or even potentially so. On the other hand, the former is repellent to all of us, even those who take that tack.
Frey is famous for biting the normative bullet and arguing for the extremely limited (involuntary) use of humans in research. Whatever standard is employed to justify using animals (e.g., quality of life below a certain threshold) must be used consistently, so that some humans will be equally eligible. He believes the demand for consistency makes the case for antivivesection very strong, but as a utilitarian, he cannot justify foregoing the benefits of all animal research.18 His view cannot be accused of [End Page 59] logical inconsistency or evasive fudging. Yet only a perfectly uncompromising anti-intuitionist can deny that justifying the conscription and harming of innocent humans is an extremely unattractive feature of a theory.
A possible way to avoid this result without rejecting all animal research is to appeal to side effects. Although it is frequently mentioned, I do not think this possibility has been carefully enough explored. Anyone who is aware of the uproar caused by the debate over whether to use the organs of anencephalics—who are not even sentient—will tremble at the prospect of a public policy of conscripting sentient humans for research. People read newspapers; dogs do not. Just imagine the riots, threats against researchers, widespread paranoia, and so on. The suggestion that education about the justification of using some humans could lessen such side effects (see, e.g., Frey 1987a, p. 97) seems unrealistic, as does the claim that the conscription and use could be kept secret. But arguments on both sides need further development.
Another possibility is to allow, in principle, the use of humans who fail the tests that research animals fail, while emphasizing that such use is not obligatory. We may, as it were, confer rights on such humans simply because we do not want to use them. My giving a gift to one person does not oblige me to give gifts to others, for no gift-giving (special circumstances aside) is owed in the first place. Of course, some will object to justifying the use of humans in principle. Another problem is that exempting humans may be unfair to animals remaining in the eligibility pool, because they thereby become more likely to be conscripted.
James Nelson has provided a different response to the problem of marginal cases (Nelson 1988). He argues that humans with greatly subnormal capacities have suffered a tragic harm in that they will never enjoy the sort of lives normal people experience. A normal rat, in contrast, does not merit special sympathy because of its limited capabilities. To those who would argue that we should feel remorse for the rat because it lacks traits of normal humans, Nelson has a response:
The rat could not be the possessor of [such] traits . . . and still be the same rat, or perhaps even a rat at all; such a radical alteration of its genetic structure would constitute an essential . . . change in its identity.
For three reasons, this difference justifies sparing such humans from research or at least greatly strengthening the presumption against using them: (1) such unfortunate humans call forth our compassion; (2) justice in distributing burdens requires us to prefer using those not already [End Page 60] harmed; and (3) not treating them with special compassion would have pernicious consequences in the form of lessened moral sensitivity and discernment. I think this intriguing argument deserves further consideration, but I am inclined to argue instead that humans of subnormal capacities who never had greater capacities have neither been harmed nor suffered a loss. One is harmed only when one's situation is changed; one can lose only what one has.19
Applications to the Research Setting
Having examined several theories of the moral status of animals and a number of unresolved theoretical issues, let us turn to the research setting and to specific working principles that are derived from, or at least suggested by, the theories. It will become apparent that there is significant convergence among the principles that flow from the theories; I believe that recognition of this convergence can facilitate progress in animal ethics.
To bolster my argument for this convergence, I will briefly explain why the contributions of philosophers H. J. McCloskey and Carl Cohen, who defend conservative views, fail to vitiate the others' positions. In "The Moral Case for Experimentation on Animals," H. J. McCloskey bases the case for animal research on prima facie duties—obligations that can conflict, none of which always trumps the others—to maximize good over evil, and to respect persons as persons, meaning respect their rights. What is the argument for these prima facie duties? McCloskey answers that "appeal must ultimately be made to the self-evidence of the principles and of the possession of these rights by persons [only]" (McCloskey 1987, p. 65).
The weakness of his case should be evident. Many utilitarians (e.g., Singer and Frey) do not find it self-evident that persons have such rights. For many others it is not self-evident that only humans have rights.20 In essence, McCloskey has made an undefended assumption on precisely the crucial issue: the relative moral status of humans and other animals. Later, he claims that to forbid animal research but permit research on volunteering human subjects is "morally outrageous" (McCloskey 1987, p. 70). This is debatable to say the least—since the humans are presumed to consent—and reveals the futility of his earlier appeal to the self-evidence of the principles on which he bases his claims.
Carl Cohen's "The Case for the Use of Animals in Biomedical Research" is equally question-begging. Against the claim that animals have rights, he offers an analysis of rights that nearly excludes animals by definition: [End Page 61]
they are . . . claims, or potential claims, within a community of moral agents. Rights arise, and can be intelligibly defended, only among beings who actually do, or can, make moral claims against one another.
Cohen offers no defense of his analysis whatsoever.
Against utilitarians who oppose animal research, he argues that, even granting equal consideration of interests, maximizing utility requires increasing animal research because of the benefits it procures (Cohen 1986, p. 868). But because benefits are always only hoped for (and often seem obtainable by alternatives), while the harms to animals are certain, Cohen's calculations seem highly implausible. Singer, at least, has marshalled considerable empirical evidence that points in the opposite direction (Singer, 1975).
Other shortcomings are noteworthy. First, Cohen fails to note the distinction between equal consideration and equal treatment—arguing that equal consideration entails either that animals and humans both lack rights or that they have precisely the same rights (Cohen 1986, p. 867). As I indicated above, Singer obviated this non sequitur in his early work; many philosophers find the mistake particularly egregious, especially in relation to the value of life. In addition, Cohen responds to the problem of marginal cases by claiming that subnormal humans are of the same kind as normal humans and therefore have the same moral status (Cohen 1986, p. 866). But he never explains why being of a particular kind, and not one's own characteristics, determines one's moral status. Moreover, he does not defend the metaphysical claim that we are of the human kind, as opposed to primate, mammal, sentient creature, animal, etc. Finally, Cohen concludes his article with ad hominem arguments against antivivisectionists who inconsistently engage in other animal-consuming practices, as if their moral character determined the moral status of animals (Cohen 1986, p. 869).
Cohen and McCloskey seem to represent the most prominent philosophical efforts to justify a strongly proresearch position, so the failure of their arguments supports the thesis of convergence towards very progressive views.
Let me then turn to Regan's position on animal research, the simplest of those I will go on to discuss. In stating what his theory implies for animal research, Regan notes the comfort it offers for subnormal humans:
Those who accept the rights view . . . will not be satisfied with anything less than the total abolition of the harmful use of animals in science—in education, [End Page 62] in toxicity testing, in basic research. But the rights view plays no favorites. No scientific practice that violates human rights, whether the humans be moral agents or moral patients [who are not moral agents], is acceptable.
In the present context I will treat Singer and Frey as having a single view because there seem to be no significant differences between them (at least since Frey's apparent acknowledgement that animals have interests). Their utilitarian position allows for harming animals only when the beneficial consequences of doing so are likely to offset the harms. The principle of equal consideration, which precludes weighting consequences differentially according to species, also justifies harming humans when doing so would maximize the good. (But see the discussion of marginal cases above.)
It is notoriously difficult to translate utilitarianism into specific rules for the research setting. Clearly, however, they would be extremely restrictive because the harms done to animals must always be measured against benefits that are only hoped for. The hoped-for benefits must be multiplied by the probability of achieving them and must outweigh the virtually certain harms. The probability is always less than one—usually much less. And the lower the probability, the greater the amount of benefit must be to meet the utilitarian standard. Rarely could a research protocol meet this standard. Still, unlike Regan, a utilitarian would probably justify the use of a small number of rodents if doing so were likely to lead to an important medical breakthrough, and if there were no less harmful known way of achieving this benefit.22
Although Sapontzis's general theory resists the simple formulation permitted by the Regan and Singer-Frey views, he extracts some clearly stated principles from his considerations of utility, fairness, and virtue. Notice that the first and second principles generalize from federal guidelines covering adults, who can give informed consent, and children, who (like animals) cannot. Experiments are to be allowed only
i. on those who freely and with understanding consent . . .
ii. when, in situations beyond the subjects' ability to understand . . . , a guardian determines that participating in the experiment would (likely) be either innocuous to or beneficial for the research subjects and freely and with understanding consents for them, or
iii. when conducting this research on these subjects is the only available way to attain a clear and present, massive, desperately needed good that greatly [End Page 63] outweighs the sacrifice involved . . . and where such sacrifice is minimized and fairly distributed among those likely to benefit from the research. (Sapontzis 1987, p. 226)
Most animal research is neither innocuous nor performed for the benefit of animals, so (iii) is the key principle. Because it requires a "good that greatly outweighs the sacrifice . . ." (emphasis added), Sapontzis's view appears to be normatively between Regan's view and utilitarianism. He says the practical implication is that "virtually all of that research would have to be radically restructured or terminated" (Sapontzis 1987, p. 228).
Midgley's theory does not provide anything like clear guidelines for the use of research animals. However, it seems safe to conclude from the tenor of the entire book, and from her criticisms of the typical justifications researchers provide for their excesses (see, e.g., Midgley 1984, pp. 37-39, and Midgley 1981), that Midgley's view supports considerably tighter restrictions than those currently in place. At the same time, it is clear from her arguments supporting a limited degree of species favoritism that her restrictions would be looser than those of the other views we have examined.
This review of the policy implications for animal research of what I consider the five leading theories in the animal ethics literature, leads to a striking conclusion: all the theories are very progressive. They range from total abolition to an implicit call for considerably tighter restrictions.
Let us conclude with a look into the future. What should be on the agenda for animal ethics? One issue requiring more careful treatment is the question of which species of animals are worthy of moral consideration. In tackling this question, the first task is to arrive at a more uniform understanding of what is required, in principle, for moral status. If the many who identify interests as the moral sine qua non are correct, residual issues in analyzing this concept need to be cleared up. Perhaps most importantly, can one have an interest in something of which one has no conception? This issue is enormously important for determining the values of different animal lives—one of the major outstanding issues I discussed earlier. Whatever our answers to these questions, an even more difficult task is determining which animals satisfy the conceptual conditions for moral status. Although there is a near consensus that at least vertebrate animals can suffer,23 beyond that range much is in dispute, indicating the need for extensive empirical research. [End Page 64]
Further exploration of the other outstanding issues that I raised earlier is also needed. Are all beings with moral status owed equal consideration? Or is some degree of preferential treatment for our own species justified? While nonhuman animals with interests clearly deserve some consideration, and while the leading theoretical efforts all conclude that they deserve significant consideration, it is not enough to leave it at that. The difference between allowing some prohuman discrimination and requiring equal consideration is significant, and little that is penetrating has been written on this issue. Finally, if we permit some harmful use of animals for human benefit, how are we to regard those humans who are in no greater possession of morally relevant characteristics than the animals? Is there any way of exempting all nonvolunteering humans from harmful use—short of abolishing animal research?
Another item on the agenda of animal ethics—which overlaps with the first—is to achieve a deeper and more comprehensive understanding of animal interests or, less technically, animal welfare. To effectively promote something requires understanding it. To make a responsible estimate of how large a monkey's cage should be, we need to understand in what sense and to what degree confinement is harmful to the monkey. To decide reasonably how many monkeys should be placed in a cage, we need to understand the companionship needs of monkeys and how monkeys interact in groups. And to do all of this requires grasping a great deal about the nature of monkeys—biological, psychological, and social. As another example, when we give repeated electric shocks to rats, how bad, qualitatively, are their experiences? Do they experience only pain, or also highly emotional forms of suffering? Answering such questions is necessary for determining the extent of harm caused by particular cases of inflicted pain, suffering, and other unpleasant experiences. Of course, the absence of such precise information in no way justifies not promoting animal welfare as well as possible under the circumstances.
A further item is to maximize points of contact in theories of the moral status of animals. Less emphasis on the differences between, for example, Regan's and Frey's views, and more on the common ground of rights-based and utilitarian approaches should be made. Even better, the development of theories that form clear and coherent compromises between the two basic approaches would be welcome, especially since the uncompromised versions presented by the first generation seem open to serious objections and are likely to satisfy only small sects of devoted dogmatists. Sapontzis has taken a big step in this direction, although I think he leaves [End Page 65] much to be desired at the foundations. Wayne Sumner has also taken such a step (see Sumner 1987 and Sumner 1988), though his main theoretical work has not focused on animals in particular.
A final item on the animal ethics agenda takes us onto a somewhat different stage. If my thesis about convergence is correct, policy changes will be in order. If in addition, further efforts to improve theories and resolve outstanding issues lead to greater specificity about what actual policies should look like, the question of how to effectuate such changes will become prominent. Hence the vexing issue of what strategy to adopt in the face of institutional and political realities; resistance to significant change will be enormous. This is an issue of both prudence and ethics: what works and what is permissible? Which is more effective, reform from within or more radical measures? Is civil disobedience or even violence ever justified in the name of combatting injustice? If so, how egregious does the injustice have to be? A detailed exploration of these issues will have to await another occasion.
David DeGrazia, Ph.D., is Assistant Professor of Philosophy, George Washington University.
1. See, for example, Brandt (1979), in which utilitarianism takes the form of a system of rules, and Dworkin (1978), in which rights (which usually trump considerations of utility) may be overridden in particular cases if the consequences of preserving them would be extremely pernicious.
2. This desire clearly affected the style of Bernard Rollin's Animal Rights and Human Morality (1981). The book is rich with good ideas and the beginnings of good discussions, but sails through issues in a superficial, if highly readable, manner.
3. It is worth mentioning Bernard Rollin's The Unheeded Cry (1989), and James Rachels's Created from Animals: The Moral Implications of Darwinism (1990). The former is a significant contribution to the philosophy of mind, but presents no ethical theory; the latter is of interest to numerous areas of philosophy and includes illuminating discussions of various points implicit in utilitarianism, but presents no novel theory of the moral status of animals.
4. See DeGrazia (1989, ch. 3), for an attempted refutation of such extreme anti-intuitionism.
5. For an explanation of how he arrives at this particular moral theory, see Singer (1989, p. 11).
6. It is difficult not to notice the influence here of the famous British philosopher R. M. Hare. Singer was also his pupil. [End Page 66]
7. For attempted refutations see Regan (1983, pp. 37-49) and DeGrazia (1989, pp. 239-47). Perhaps the best attempt is a section in Rollin (1989), "The Claim that Animals Lack Concepts."
8. This term was popularized with Rawls's eloquent arguments in Rawls (1971).
9. See, for example, his careful discussion of what has been called "the replacement argument" (ch. 10) and his argument that, in an important sense, many animals can act virtuously (ch. 3).
10. This issue is addressed in DeGrazia (1989, ch. 2).
11. This argument is given in Singer (1979, pp. 10-11).
12. I also find compelling James Rachels's argument that, if a Martian came to Earth to teach in a boys school and behaved just as humans do, we would not think it justified to treat him with less consideration just because he is not homo sapiens (Rachels 1989, pp. 103-4).
13. The attitude is well reflected in this statement: "[Various] methods can be used for euthanasia [any killing] of anesthetized animals because the major criterion of humane treatment has been fulfilled" (Health and Human Services, Department of, 1985, p. 38).
14. I do not know that they unequivocally hold this position, but it is implied by their version of preference-utilitarianism.
15. Singer and Frey also allow for differences in the value of lives, but for simplicity, I ignored this feature of their position.
16. A similar argument is made, though more tentatively, in Jamieson (1983, pp. 145-46).
17. Mill in his classic Utilitarianism writes that it is "better to be a human being dissatisfied than a pig satisfied," arguing that certain kinds of pleasures that humans experience are qualitatively superior to other, baser pleasures. In keeping with Sapontzis's arguments, Edward Johnson (1983) attempts to discredit Mill's thesis.
18. He writes: "Clearly we have here a powerful argument for antivivisection: it is one sure way to bar human experiments. The reason I cannot endorse it is that it [ignores] the benefits to be derived from (some, by no means all) experimentation" (Frey 1987a, p. 98).
20. McCloskey might take comfort in my discussion of the sui generis view above, in which I said it was not logically impossible to argue that being human per se confers special moral status. But again, a compelling argument, and not just a stalemate, is needed to meet the burden of proof set by the principle of universalizability. McCloskey would also need to explain why so many reasonable people fail to perceive what he takes to be self-evident. [End Page 67]
21. This unequivocal abolitionist view, as well as Regan's general theory, were endorsed by speakers at the well-attended animal rights march in Washington, D.C., on June 10, 1990.
22. At the same time, Sumner (1988) notes considerable convergence between Regan's view and utilitarianism in the research setting, and forges a compromise between them.
23. For a particularly comprehensive article, see Rose and Adams (1989).