------"'_...•...

THE RESPONSIBLE USE OF ANIMALSANIMALS IN BIOMEDICAL RESEARCHRESEARCH

Edwin Converse HettingerHettinger College of CharlestonCharleston

arl Cohen's defense of the use of animals for biomedical research in The New C EnglandJournal ofMedicinel raises most of the major issues in the moral controversy con­ cerning human treatment of nonhuman animals. It exhibits the major lines of attack against both advocates (such as Tom Regan2) and utilitarian animal-liberationists (such as !!). It is also a showcase of the most common mistakes made by those who seek to defend the current human use of animals. Cohen argues that although we do have obliga­ tions to animals - for example, not to be cruel to them - we have no obligations to animals based on their rights to such treatment. According to Cohen, the biomedical use of animals does not violate their rights, since by their very nature animals cannot have rights. Cohen rejects the util-

PHILOSOPHY

Summer 1989 123 The Responsible Use ofAnimals in Biomedical Research

itarian argument that much of the biomedical use Do All Humans But No Animals Have Rights? of animals is an unjustified subordination of the most vital interests of animals to relatively minor Cohen argues that only humans beings can have human concerns. He thinks a proper utilitarian rights. assessment of animal experimentation counsels the increased use of animals in biomedical Rights arise, and can be intelligibly defended, research, rather than its reduction or elimination. only among beings who actually do, or can, Because there is a stronger justification for animal make moral claims against one another. use in biomedical research than for any other use Whatever else rights may be, therefore, they of animals (e.g., for food or clothing), Cohen are necessarily human; their possessors are argues that opponents of animal experimentation persons, human beings. (p. 865) must adopt what he feels is the absurd position Cohen is correct in maintaining that rights which opposes all use of animals. cannot arise unless there exist moral agents for In response, I argue that Cohen cannot secure whom these rights claims make sense. To say that the rights of severely retarded humans while some being has a right is to say (at least in part) denying that psychologically sophisticated animals that some other being has obligations to treat the have rights. Cohen can reach his conclusion that right holder in certain ways specified by that right. we should increase our biomedical use of animals So if there were no beings more cognitively and only because (1) he counts animal pain and suf­ morally capable than pigs or dogs, there would be fering as less important than equivalent human no rights. pain and suffering, (2) he ignores the frequent However, the fact that rights claims require the misuse of animals in biomedical research, and (3) existence of duty bearers does not imply that only he overlooks the abundant alternatives to current those duty bearers can have rights. Even Cohen animal experimentation. I propose the limited would grant that human infants have rights, yet use of animals based on their degree of psycho­ they are not duty bearers. Thus, some creatures logical sophistication as a consistent and attractive possess rights despite being unable to invoke them alternative to the extreme views of both Cohen against others or to recognize and respect others' and his absolute prohibitionist opponents. I close rights. by suggesting that only if researchers would be Cohen attempts to avoid this objection by willing to experiment on severely retarded shifting his criterion of rights possession to the humans at comparable levels of psychological capacity for being a moral agent, rather than sophistication are their experiments on animals actually being a moral agent. morally permissible. Animals ... are not beings ofa kind capable of exercising or responding to moral claims. Animals therefore have no rights, and they can have none.... The holders ofrights must have the capacity to comprehend rules of duty.... (p. 866) However, most people would grant that severely retarded humans have rights (Cohen does), and yet they do not have "the capacity to comprehend rules of duty." Thus if having the capacity to be a duty bearer is necessary for the possession of rights, then severely retarded humans cannot have rights. Cohen responds to this point with his talk of ''kinds.''

Between the Species 124 Summer 1989 ~~~~~~~~~~~~~~~~~~~~~------

The Responsible Use ofAnimals in Biomedical Research

The capacity for moraljudgment that lacking the ability to realize that capacity. But why distinguishes humans from animals is not a should we accept such an attenuated notion of test to be administered to human beings one capacity? Certainly capacities can be left unrealized, by one. Persons who are unable, because of but if there is no possibility that they could ever be some disability, to perform the full moral developed, what sense is there in claiming that the functions natural to human beings are capacity is present? I see no reason to accept the certainly not for that reason ejected from the notion that there can be unrealizable capacities. moral community. The issue is one of kind. Humans are of such a kind that they may be Is Defensible? the subject of experiments only with their voluntary consent. The choices they make Perhaps Cohen would agree that severely freely must be respected. Animals are of such a retarded humans lack the capacity for moral agency kind that it is impossible for them, in but thinks this is unimportant. He may be arguing principle, to give or withhold voluntary that we should treat the severely retarded as human consent or to make a moral choice. What beings and that since human beings have rights humans retain when disabled, animals have (presumably because many of them are moral never had. (p. 866) agents), severely retarded humans have rights as well. On this reading, Cohen is suggesting that we Cohen seems to be claiming that the capacity for treat individuals according to their biological kind moral agency is essential to human beings and is and ignore their individual characteristics. Moral necessarily lacking in other animals. Thus, severely status is to be determined by species membership, retarded humans, because they are human, retain not individual qualities. This is "speciesism": the the capacity for moral agency even in their view that species membership is by itself a morally retarded state. Animals by their very nature lack legitimate reason for treating individuals differ­ this capacity. Since the capacity for moral agency ently. confers rights, severely retarded humans have Peter Singer and others have argued that rights, whereas animals do not. speciesism is "a form of prejudice no less objec­ But many severely retarded humans could never tionable than racism or sexism."4 Cohen's speciesist carry out even the quasi-moral functions that some perspective concerning the moral status of animals animals can perform. Dogs, for example, can be vis-a-vis humans does coincide uncomfortably with obedient, protective, and solicitous, while there are the outlook of racists and sexists towards blacks and severely retarded humans who could not achieve women. Both judge according to class membership these minimal moral abilities despite our best while ignoring individual qualities. efforts. Given this fact, it just is not plausible to Cohen responds to this charge of speciesism by claim that severely retarded humans have the embracing it: capacity for moral agency, while claiming that psy­ chologically sophisticated animals do not. Cohen I am a speciesist. Speciesism is not merely certainly has not given us any reason to accept this plausible; it is essential for right conduct, claim. He simply assumes that being a member of a because those who will not make the morally biological species guarantees that one has certain relevant distinctions among species are almost capacities, despite overwhelming evidence that certain, in consequence, to misapprehend marginal members of species often lack capacities their true obligations. The analogy between normal for that kind of creature. We need a strong speciesism and racism is insidious. Every argument before we should reject the obvious point sensitive moraljudgment requires that the that some animals have a greater capacity for moral differing natures of the beings to whom behavior (however minimal) than do some severely obligations are owed be considered. (p. 867) retarded human beings. Cohen might argue that severely retarded This passage defends the truism that there often humans have the capacity for moral agency despite are differences between members ofdistinct species which are morally relevant in determining how we

Summer 1989 125 Between the Species The Responsible Use ofAnimals in Biomedical Research

should treat them. But this is not what is at issue in The analogy between speciesism and racism or the debate over speciesism. Singer, Regan, and sexism is deficient in one respect. Species classifi­ other opponents of speciesism are not suggesting cation marks broader differences between beings that we ignore morally relevaI).t differences than does racial or sexual classification. Thus between members of different species and treat attempting to justify differential treatment on the them all identically. (They are not suggesting, for basis of species membership alone (as Cohen example, that dogs be allowed at the dinner table does) is not just as morally objectionable as doing or be allowed to vote.) What rejecting speciesism so on the basis of race or sex, since members of commits one to is being unwilling to use difference different species are more likely to require differ­ in species by itself as a reason for treating indi­ ential treatment than are members of different viduals differently. Similarly, rejecting racism and races or sexes (within a species). For example, in sexism commits one to not using race or sex by determining what sort of food or shelter to itselfas a reason for differential treatment. Cohen's provide, it would be much more important to truism does not support speciesism in this prob­ know a creature's species than it would be to lematic sense. know a person's race or sex. But this does not imply that difference in species by itself is a morally legitimate reason for treating individuals differently, while difference in race or sex considered by itself is not. Arguing that a woman should be prohibited from combat nless Cohen can show us because of her sex fails to provide a morally rel­ evant reason for the recommendation. Arguing U that there is some morally for this on the grounds that this woman lacks the required physical capacities is to provide a morally relevant difference between relevant reason. Similarly, arguing that a chim­ panzee should be experimentally sacrificed rather severely retarded humans and than a human, simply because it is a chimpanzee, gives no morally relevant reason for the recom­ psychologically sophisticated mendation. However, arguing that the chim­ panzee does not value or plan for its future ~ife to animals, his position is open to the the extent that the human does is to provide such a reason. following objection: if experimenting Thus even though considerations of species are frequently more closely correlated with morally on severely retarded humans is a relevant features than are considerations of race or sex, species membership by itself (like racial or violation of their rights, then sexual class membership) is not a morally legit­ imate reason for differential treatment. experimenting on psychologically Speciesism is thus a moral mistake of the same sort as racism and sexism: it advocates differential sophisticated animals violates their treatment on morally illegitimate grounds. The illegitimacy ofjudgments based on species rights, as well. membership alone becomes especially clear when comparing the moral status of a severely retarded human with that of psychologically sophisticated animals, since here the individual does not have what most members of the species have. The morally relevant differences which usually exist between individuals of two different biological

Between the Species 126 Summer 1989 The Responsible Use ofAnimals in Biomedical Research

kinds (and hence which would frequently justity Cohen rejects the utilitarian critic's position that treating them differently) are lacking when com­ the like interests of humans and animals should be paring severely retarded humans with psychologi­ given equal moral weight. He denies that similar cally sophisticated animals. Any plausible morally amounts of human and animal pain are equally relevant characteristic - whether it be rationality, morally significant. self-sufficiency, ability to communicate, free The first error is the assumption, often choice, moral agency, psychological sophisti­ cation, fullness of life, and so on - is possessed by explicitly defended, that all sentient animals have equal moral standing. Between a dog and some animals to a greater extent than by some have equal moral standing. Between a dog and severely retarded humans. In this case, to classity a human being, according to this view, there is by biological kind and to argue for differential no moral difference; hence the pains suffered treatment on that basis alone obscures and by dogs must be weighed no differently from ignores morally relevant features rather than the pains suffered by humans.... Ifall forms of relying on them. We should not treat individuals animate life ... must be treated equally, and if on the basis of group or kind membership when therefore in evaluating a research program the their individual characteristics are readily pains ofa rodent count equally with the pains ofa human, we are forced to conclude (1) apparent and relevan t. ofa human, we are forced to conclude (1) Thus, Cohen's argument fails on this second that neither humans nor rodents possess interpretation, as well. His appeal to biological rights, or (2) that rodents possess all the rights kind to justity differential moral status of severely that humans possess.... One or the other must be swallowed if the moral equality of all retarded humans and psychologically sophisti­ be swallowed if the moral equality ofall cated animals is an unjustified form of speciesism. species is to be defended. (p. 867) Unless Cohen can show us that there is some This argument misses the mark. To claim that morally relevan t difference between severely animals "'have equal moral standing" and should retarded humans and psychologically sophisti­ have their like interests treated equally implies cated animals, his position is open to the following neither that there are no moral differences objection: if experimenting on severely retarded between humans and animals nor that we should humans is a violation of their rights, then experi­ treat animals in the same manner that we do menting on psychologically sophisticated animals humans. violates their rights, as well. From the utilitarian position that the right act is the one which maximizes the net satisfaction of Does Utilitarianism Iustify Animal interests it follows that it is morally preferable to give a human a slightly less amount of pain than to Experimentation? give an animal a slightly greater amount of pain (or vice versa). If the pains are of equal intensity Utilitarians hold that the right policy is the one and consequence, then one should be morally whose consequences maximize the satisfaction of indifferen t. The fact that one is the pain of a interests. In this calculation the interests of all human and the other is the pain of an animal is affected parties are fairly taken in to accoun t. not by itself morally relevant. Utilitarians who oppose animal experimentation This is not to say that the same type of exper­ do so not on the grounds that animal rights are iment on a human and an animal would cause violated but because they think that the overall each the same amount of pain and suffering and good resulting from these experiments is not suffi­ that we should be indifferent to which being we cient to justify their negative consequences. The use. Giving a typical chimpanzee a deadly virus in

benefits which result from animal experimentation order to test a v~ccine is likely to cause less pain (such as an increase in scientific and medical and suffering than giving a typical human the knowledge) either do not outweigh the costs (e.g., deadly virus for the same purpose. The greater psy­ animal pain and death) or could be achieved in a chological sophistication of the human, its greater less costly fashion. intelligence and self-consciousness, makes possible

Summer 1989 127 Between the Species The Responsible Use ofAnimals in Biomedical Research

a greater degree of pain and suffering. Cohen can reach this conclusion only by aban­ (Sometimes the reverse is true, however.5) doning utilitarianism (and its principle of equal Even though pain and suffering would often be consideration of like interests), by adopting the minimized by experimenting on an animal instead speciesist position which treats animal pain and dis­ ofa typical human, that does not show that we may tress as insignificant when it is a means to human morally discount the pain and suffering ofanimals. benefit, and by being overly pessimistic about the We must still count the pain and suffering of possibility ofalternatives to animal use. animals equally with the like pain and suffering of humans. But in cases where a human will suffer The Possibility of Substitution more, we should prefer the use of animals (and vice versa) . Whether research using living creatures is Cohen is thus mistaken in thinking that giving justified on utilitarian grounds depends in large equal consideration to the like interests of animals part on the availability of substitute procedures. A and humans makes moral discriminations between utilitarian benefit/cost analysis (which must the two impossible. For a utilitarian, equal consid­ consider alternative, less costly ways to achieve eration (or equal moral standing) does not imply these benefits) would find that some, perhaps identical treatment. Cohen has given us no cogent many but certainly not all experimen ts, using reason for rejecting the view that the like pains of animals are morally justifiable. Some use of living humans and animals must be given equal moral beings continues to be necessary and justifiable. weight. Since the pain of the animals on whom we Even developing alternatives to the biomedical use experiment cannot be discounted, Cohen's utili­ of animals often requires the use of animals. At tarian justification for the biomedical use of present substitute techniques are not sufficiently animals becomes far more difficult to achieve. developed to eliminate this use entirely (and they Cohen argues that even if "the pains of all may never be).6 animate beings must be counted equally" (p. 868), Nevertheless, Cohen is overly pessimistic about a utilitarian calculus would still come out in the possibility of alternatives to the current support of the biomedical use of animals: biomedical use of animals. His speciesism prevents him from appreciating or even acknowledging the The sum of the benefits of their use is is numerous substitute procedures that are being utterly beyond quantification. The The developed. A recent report by the U.S. Congress' elimination ofhorrible disease, the the Office of Technology Assessment (OTA) on alter­ increase oflongevity, the avoidance of of natives to animal use in research, testing, and edu­ great pain, the saving oflives, and the the cation is much more encouraging about the improvement of the quality oflives (for (for potential for alternatives.7 This study presents humans and for animals) achieved achieved numerous suggestions involving the replacement, through research using animals is so so reduction, and refinement of the use of animals. In incalculably great that the argument of of addition to the promising techniques of in vitro these critics, systematically pursued, pursued, experimentation and computer simulation (which establishes not their conclusion but its its Cohen mentions), the OTA report suggests: reverse: to refrain from using animals in in biomedical research is, on utilitarian utilitarian (1) Coordinating investigations and sharing grounds, morally wrong. (p. 868) 868) information (to reduce duplicative experiments when unnecessary for validating the original Substantial benefits have resulted (and continue research) ; to result) from biomedical experimentation, much ofwhich involves the use of animals. And although (2) Replacing the use of higher animals with with a utilitarian benefit/cost analysis would reach the lower animals (invertebrates for vertebrates and and conclusion that it would be wrong to stop the use cold-blooded for warm-blooded animals); animals); of animals entirely, it would notjustity Cohen's call (3) Using plants instead of animals; for an increase in the biomedical use of animals.

Between the Species 128 Summer 1989 The Responsible Use ofAnimals in BioJfUldical Research

(4) Sharing animals (e.g., getting several tissues and articles have persuasively documented that from one animal); many experiments using animals have been unpro­ fessional, of dubious scientific merit, repetitive, or (5) Designing experiments which use statistical cruel.12 Two video tapes are especially persuasive; inferences and whose design provides reliable "Unnecessary Fuss," about head injury research information despite the use offewer animals; involving baboons at the University of (6) Decreasing the pain and distress in animal Pennsylvania,l~and "Tools For Research," a general experimentation by altering the experimental review of research using animals over the last design and by using anesthetics and tranquilizers; twenty years.I4 The flurry of recent legislation con­ cerning cited above shows a growing (7) Using non-living chemical and physical public recognition of the misuse of laboratory systems that mimic biological functions; animals. Government regulations for the care of (8) Using human and animal cadavers; and laboratory animals have been developed to prevent these sorts of experiments, as well.I5 Cohen's sug­ (9) Teaching by demonstration instead ofby gestion that we encourage the wide and imaginative individual student use ofanimals. use of live animal subjects, instead of limiting this Recent amendments to the Animal Welfare Acts use and working to 'find substitute techniques, and the Public Health Service Act,9 as well as legis­ shows blatant disregard for this widely acknowl­ lation concerning the education of health profes­ edged problem. sionals,lO all encourage alternatives to the current methods of animal use.ll Cohen's pessimistic assessment of these alternatives flies in the face of a growing trend of using already existing alternatives and of developing new substitute procedures. Experiments which cause animals pain, distress, or death are clearly not justifiable when such sub­ stitute procedures are available.

Should We Increase Biomedical Animal Use?

Cohen argues that in order to achieve maximum safety for humans "the wide and imaginative use of live animal subjects should be encouraged rather than discouraged" (p. 869). Cohen is right that some experiments which subject humans to risk could be conducted using animals without loss in the significance of the results. Furthermore, risky experiments which are necessary should be per­ formed on psychologically less sophisticated crea­ tures. An increase in psychological sophistication brings with it a wider range of interests, a greater ability to experience satisfaction (and dissatis­ faction), and the possibility of leading a fuller life. Inflicting suffering or death on these creatures causes greater harm. In advocating an increase in the biomedical use of animals Cohen not only ignores the available alternatives but disregards the widespread experi­ mental misuse of animals, as well. Numerous books

Summer 1989 129 Between the Species The Responsible Use ofAnimals in Biomedical Research

Can A Consistent Position Concerning Animal Use consistently advocate the limited use of animals in Be Developed? other areas, as well. Both extremes - the absolute prohibition of all animal use, as well as Cohen's Cohen charges his anti-speciesist opponents with speciesist encouragement of such use - should be inconsistency or absurdity: "Scrupulous vegetari­ avoided. anism, in matters of food, clothing, shelter, com­ merce, and recreation, and in all other spheres, is A Test Biomedical Researchers Should Use the only fully coherent position the critic may adopt" (p. 869). The person who eats veal and then I have suggested that it would be morally strenuously objects to the killing of cats in relatively preferable, ceteris paribus, to give a deadly virus to painless medical experiments is inconsistent. We an animal rather than to a typical human being. do not need to eat animals for food (certainly not The pain, suffering, and distress caused by the two mammals); carefully chosen vegetarian diets are experiments, as well as the significance of the loss perfectly healthy. We do need the ongoing results of life, would be minimized by experimenting on of biomedical research, and for some of this the animal. However, this argument in support of research the use of living creatures continues to be the experimental use ofanimals rather than typical required. humans does not give us a reason for preferring Cohen is right that the use of animals in animal experimentation to "marginal case" human biomedical research is less difficult to defend than experimentation. Since many animals (e.g., chim­ are other uses of animals. (Only one out of every panzees) and severely retarded humans would suffer hundred animals used is for this purpose.I6 ) But equally from such an experiment, the pain and dis­ the anti-speciesist critic of current biomedical uses tress of the experimental subject gives us no reason of animals need not be committed to prohibiting to prefer the use of one to the other. Furthermore, all uses of animals. Since anti-speciesism allows for given their rough equivalence in psychological discriminating between animals, critics can consis­ sophistication, the value of the two creatures' lives tently object to the raising, slaughtering, and con­ is about the same. Whatever moral rights such crea­ sumption of veal calves while not objecting to tures have, ifany, are also comparable. commercial shrimp farming and shrimp con­ Thus, an important test to determine if an sumption. A critic might also object to repeated experiment is signifIcant enough to justify the surgery on healthy animals in the training ofveteri­ pain, suffering, and (perhaps) death of the narians and not object to the use of chick embryos creature involved is to ask the following question: for toxicity testing. The recommendation that Would the investigator still think the experiment experimenters substitute cold-blooded animals for justifiable if it were performed on a severely warm-blooded ones or invertebrates for vertebrates retarded human at a comparable psychological is also perfectly consistent. These suggestions are level as the animal? If not, then the experiment not speciesist, since species membership per se is should not be conducted. Only an arbitrary pref­ not the justification offered for differential erence for members of our own species could avoid treatment. Differences in the fullness oflife, in psy­ this conclusion. chological sophistication, and in the capacity for If this test were used, and I am arguing that it is suffering are what motivates these suggestions. the appropriate test, many - though certainly not Thus, one can argue for limiting animal use in all - experiments on animals would cease and be biomedical research without embracing the replaced by alternatives. Biomedical researchers extreme position prohibiting all uses of any would do well to keep this test in mind.I7 animals for whatever reason. Cohen can success­ fully saddle only his most extreme opponents with - " this consequence. A more circumspect skepticism ~~~)I".~~~ about the legitimacy of a significant portion of lab­ oratory animal use is possible. Advocates of lim­ iting the use of animals in biomedical research can "

Between the Species 130 Summer 1989 The Responsible Use ofAnimals in Biomedical Research

Notes

1 Cohen, Carl (1986), 'The Case for the Use of Animals in Biomedical Research," NewNIJW EnglandEnglandJoumal Joumal ofMedicine, 315, pp. 865-870. All page references in the text are to this article.

2 Regan, Tom (1983), The Case fOTfor Animal Rights (Berkeley, California, University of California Press).

!l~ Singer, Peter (1975), (New York, Avon Books). TREATING ANIMALSANIMALS 4 Singer, Peter (1986), "Animal Liberation," in VanDeVeer, Donald and Pierce, Christine (eds.), People, Penguins, and PlasticPl=tic NATURALLY?NATURALLY? Trees (Belmont, California, Wadsworth), p. 31.

5 Singer, Peter (1979), Practical Ethics (Cambridge, Cam­ bridge University Press), p. 53. Holmes Rolston III Colorado State University 6 Office of Technology Assessment (OTA) (1986), Alternatives to Animal Use in Research, Testing, and Education (Washington, D.C., U.S. Government Printing Office, OTA-BA-273), p. 138. Editors' Note: This article is a 7 Ibid. response to an article by Professor 8The Food Security Act of 1985 (Public Law 99-198). Peter S. Wenz, "Treating Animals 9 The Health Research Extension Act of 1985 (Public Law 99­ Naturally," published in Between 158). the Species, vol 5, no. I, pp. 1-10. 10The Health Professions Educational Assistance Amend­ ments of 1985 (Public Law 99-129). f the nine chapters in Environmental llSee OTA, AllemativesAlternatives to Animal Use in Research, Testing, and Education, Ch. 13. Ethics I most expect criticism on the one on higher animals - not because my 12See Singer, Animal Liberation, Ch. 2; Ryder, Richard (1985), O treatment of animals is socially controversial but "Speciesism in the Laboratory," in Singer, Peter (ed.), In Defense ofAnimaL!'Animal< (New York.York, Basil Blackwell) pp. 77-88;Jamieson, Dale because it isn't. The chapters on organisms, and Regan, Tom (1982), "On the Ethics of the Use ofAnimals in species, and ecosystems all depart more radically Science," in Regan, Tom and VandeVeer, Donald (eds.), And from current thought. My value theory in the book Justice For All (Totowa, NewJersey, Rowman and Littlefield), pp. is objective, running upstream against a torrent of 169-196; Rollin, Bernard (1981), Animal Rights and Human subjectivity. But my account of animals will disap­ Morality (Buffalo, Prometheus Books), Ch. 3. point animal activists. I eat animals and leave them I!lAvaiiablel~Available from People for the Ethical Treatment of to perish in the wild. I kill goats to save a few Animals, P.O. Box 42516, Washington, D.C. 20015. endangered plants. I tolerate , under 14Available from Bullfrog Films, Inc., Olney, PA. ecosystemic conditions. I accept some wildlife com­ merce as a management tool. I seem to have no 15See National Institutes of Health (1985), Guidelines for the mercy. Care and Use ofLaboratory AnimaL!'Animal< (Bethesda, MD, NIH Pub. No. Frankly too, I am less than confident in applying 85-23). my theory to the examples I cite. I changed my 16See OTA.OTA, AllemativesAlternatives to Animal Use in Research, Testing, and mind about some of them while researching the Education, p. 43. book. My theory leadsle<,tds to unexpected conclusions. 171 would like to thank Beverly Diamond, John Dickerson, Martin Perlmutter, and Hugh Wilder for helpful suggestions on earlier drafts of this paper. DISCUSSIONDISCUSSION

Summer 1989 131 Between the Species