Proceedings for Pain Management and Humane Endpoints
Philosophical and Ethical Issues Concerning Pain
George Washington University
Any thorough discussion of pain in the context of animal research must face (1) certain philosophical issues concerning the concept of pain, and (2) certain ethical issues concerning the causing of pain to animal subjects. Philosophically, or conceptually, the most basic question is, "What is pain?" Other conceptual issues concern the relationships between pain and other unpleasant mental states, such as distress and suffering, as well as the nature of these related states. But because pain is (at least typically) unpleasant, causing pain to animals also raises ethical issues. A basic question here is whether it is ever right to cause pain to an individual -- human or animal -- for the sake of benefitting others. In the animal research setting, this raises the issue of whether our obligation to animal subjects is to minimize their pain (distress, suffering) or to eliminate it entirely. And if it is sometimes right to cause pain to animals in the pursuit of biomedical progress, other ethical issues that arise include the following. First, is there a degree of pain beyond which it is unethical to go? Second, how promising must research be to justify causing a certain amount of pain to animals? Third, how aggressively should the biomedical community pursue alternatives to the painful use of animals in research? In this talk, I will briefly address each of these questions.
Philosophical (Conceptual) Issues
What is pain?1 This is a surprisingly tricky question. Let's start by getting some examples of pain in view: toothaches, sunburns, cramps, back pains, the pains caused by pinching or pricking a finger. First, we can say that pains -- painful experiences -- are, or involve, sensations. (I take "mental pain" and "emotional pain" to be metaphorical extensions of the literal use of the term "pain.") In our consciousness, we can identify sensory properties of a specific pain: its location, duration, intensity, and features that permit classifying it as a particular kind of pain, such as an ache, twinge, or sting.
In addition to these sensory qualities, a pain also has an affective aspect -- how much we mind it. It seems that pains that are comparable in their sensory aspects are sometimes experienced very differently. Puzzling cases include wounded soldiers removed from battle who endure, with little apparent distress, what would normally be excrutiating pain, and lobotomized patients, who often report that they feel pain but don't mind it. Less pathological cases include the welcomed soreness that follows one's first 10K race, the stoicism that some women display in the face of the pain of childbirth, and the enjoyment one can get from painfully -- yet deliciously -- probing a loose tooth with one's tongue.2
The fact that the sensory and affective aspects of pain seem separable makes it hard to define "pain." According to the sensation model, pain is simply a particular sensation, with no intrinsic affective dimension. But that has what may be a paradoxical implication: that the most intense pains could be liked just for the way they feel. On this view, there might be a planet populated with beings who like all pains for the way they feel -- and the more intense, the better! One wonders what "pain" means, then: What kind of sensation is it, exactly?
Another approach, the attitude model, emphasizes the affective dimension of pain -- its unpleasantness -- and states that pain is any sensation that we dislike for its own felt qualities. This model is challenged by at least some of the cases mentioned earlier, in which pain does not seem to be minded. Also, we dislike certain sensations other than pain -- such as nausea, irritating itches and tickling sensations, and the discomfort that comes from sitting too long -- so the model owes us a way of distinguishing pain from these other unpleasant sensations.
Fortunately, we can reconcile these two accounts for all practical purposes involving animals. For clearly the vast majority of pains are unpleasant. Moreover, our ethical concern about animal pain concerns pain that is unpleasant; we don't need to worry about pain that is enjoyed or not minded by animals (if such a notion makes sense). Practical purposes favor a working definition of "pain" that makes reference to unpleasantness, since it is this feature that makes causing pain ethically problematic. So I suggest that pain is an unpleasant or aversive sensory experience typically associated with actual or potential tissue damage.3 (Note that defining "pain" in terms of its connection with tissue damage may serve to distinguish pain from other unpleasant sensations; it also connects pain with nociception, the detection of potentially tissue damaging stimuli by neural end organs known as nociceptors.)
Because pain is only one kind of unpleasant experience animals can have, it would be remiss not to say something about pain's connections to other kinds of unpleasant experience -- which are no less morally important than pain is. Let's begin with suffering, which is not the same as pain. Pain without suffering can be caused by an ordinary hand pinch. Nor can suffering be equated with distress, though the two are related; the mild distress of a professor on the first day of class normally does not involve suffering. Mild pain and distress don't add up to suffering. These points are captured in the following definition: Suffering is a highly unpleasant emotional state associated with more-than-minimal pain or distress.4
If suffering is defined partly in terms of distress, what is distess? Certainly animal subjects often experience this state. Distress is a typically unpleasant emotional response to the perception of environmental challenges or to equilibrium-disrupting internal stimuli. Distress can be caused by the belief that one will fail, the sight of charging lions, or diarrhea. And it can be caused by, or take the form of, various more specific mental states, including fear, anxiety, discomfort, and perhaps others (though I won't analyze these concepts here).5 Now, just as extreme pain is unlikely to occur without suffering, the same is true for extreme distress: A person having an anxiety attack, for example, clearly suffers.
Pain, distress, and suffering are all mental states that are unpleasant, aversive, or disagreeable. That is why we generally try to avoid these states and why causing them in others raises ethical issues. In the animal research context, the recognition of these points is suggested by policies requiring the use of anesthesia and analgesia to reduce the pain and distress of animal subjects. But, experientially, suffering is, in a sense, worse than pain or distress: Suffering is highly unpleasant, by definition, whereas pain and distress can be mild. Yet in biomedical circles, there has been considerable resistance to attributing suffering to animals; goverment documents concerned with humane use of animals often mention only pain, distress, and sometimes also discomfort.6 The reasons for this may have been partly political: Precisely because suffering is highly unpleasant, saying that animal subjects can suffer is tantamount to saying that certain procedures performed on them raise very significant moral issues. In any case, one indication that resistance against speaking of animal suffering may be weakening is the use of such language in the recently adopted "NASA Principles for the Ethical Care and Use of Animals."7
Another possible reason for resisting the idea that animals can suffer concerns one of the most interesting facts about suffering -- namely, that whether and how much one suffers can vary according to attitudes or expectations about the associated pain or distress. Let me explain with examples. Even the mild pain of a common headache can lead to great distress and suffering if the pain endures with no end in sight, or if the subject believes the headache is a symptom of impending physical deterioration. On the other hand, soldiers have sometimes incurred major injuries yet apparently not suffered (or not suffered much), due perhaps to the relief of expecting to be removed from combat, or to positive attitudes about the context of the injury (as in "I'm a hero"). Long-distance runners who experience pain and discomfort in a race may or may not suffer, or may suffer more or less, depending on such psychological factors as how they evaluate their efforts, and their attidudes (maybe relaxed, maybe fearful) about the rest of the course. Thus any meaning one attaches to the context, and one's expectations for the future, are important factors in whether and how much one suffers.
Now because meaning and expectations are important factors in human suffering, and because animals are often thought to be incapable of assigning meaning and anticipating the future, it might be thought that only humans can suffer. This reasoning is unsound, however. First, to say that meaning and expectations for the future are important factors in human suffering doesn't entail that either is a necessary condition for suffering. The definition I gave earlier of "suffering" makes no reference to those factors. And it seems plausible that if scalding water spilled on a human infant, the infant would suffer even if he or she could assign no more meaning to the event than a dog could. Further, even if expectations for the future are necessary for suffering, the idea that only humans could suffer would depend on the premise that animals have no sense of the future at all. As I have argued elsewhere, that is highly implausible in the case of "higher" animals.8 For one thing, much work in cognitive ethology suggests that their behavior cannot be adequately explained without attributing at least some sense of past and future to them. In sum, the combined empirical and philosophical case for attributing suffering -- not just pain and distress -- to a wide range of animals appears very strong.9
The fact that a wide range of animals -- and therefore many research animals -- can experience pain, distress, and suffering raises ethical issues. In discussing these issues, for the sake of simplicity and in keeping with the topic of this conference, I will generally refer only to pain.
The first issue is so basic that it might easily escape notice: Is it ever right to cause pain to another individual -- human or animal -- for the sake of benefitting others? I put the question this way because most animal research is nontherapeutic. Therapeutic research, which aims to benefit the animal subjects themselves, is in itself morally unproblematic. But most animal research aims to benefit only individuals other than the animal subjects who endure pain -- and usually the intended beneficiaries are human beings, who would gain from biomedical advances.
Note that the issue here is not whether the goal of improving human health is a worthy goal. Of course it is. No morally serious person could deny this. No, the issue is what means to this worthy goal are ethically permissible. It is a familiar moral point that not all means to worthy goals are permissible. For example, it would be wrong to force human subjects to take part in highly painful, lethal research in order to advance biomedicine. So, since it doesn't follow from the fact that improving human health is a worthy goal that any means to this goal are permissible, the question arises whether we may cause pain to animal subjects in pursuit of this goal.
In the case of human beings, we generally agree that no competent human subject should be exposed to harm, or risk of harm, without the subject's informed consent. As for minors, who cannot give informed consent, it is generally accepted that they should not be involved in nontherapeutic research that involves more-than-minimal risk. Now, of course, animal subjects can never give informed consent to endure pain or undergo risks in the name of biomedical progress. However, if it is acceptable to involve human children in research that involves only minimal risk, it's hard to see why it would not be acceptable to involve animals in minimal-risk research. Such research may involve a small amount of transient pain, such as that associated with a blood draw. Thus, it seems implausible to say that it is always wrong to cause pain to animal subjects; very minimal pain seems no more problematic than minimal-risk research involving human children. (That doesn't mean, however, that we shouldn't try to eliminate pain wherever possible.)
What about research that imposes more-than-minimal risk (or harm)? Is it ever justified to cause moderate to severe pain to animal subjects (who, again, cannot consent to such procedures as competent human subjects can)? Here matters are more controversial. Many animal advocates would answer negatively. Most people professionally connected with biomedicine would probably answer affirmatively, following the official positions of their organizations. But, if it's right to impose moderate to severe pain on animal subjects (who cannot consent), but wrong to impose such pain on human subjects (without their consent), there must be some morally relevant difference between all human subjects and all animal subjects that justifies different standards with respect to each group. That is, there's a burden of justification on proponents of such a double standard -- and, in my opinion, this burden has never been successfully carried.
My own view is that causing some amount of pain (distress, suffering) to animals is as serious a moral matter as causing the same amount of pain (distress, suffering) to humans. I realize that it can be very difficult to assess how much pain an animal (or even a human) is likely to undergo from a certain procedure, but the present claim is one of ethical principle: Pain is not less important just because it is an animal who stands to experience it. This standard would severely restrict the types of animal research that would be deemed morally permissible. But, unlike many people who share roughly this view, I believe that the opposing arguments, suitably developed, are nearly as strong as those that support my position. So, as I have explained elsewhere, I have respect for a view that's somewhat more permissive about animal research than my own view -- and I can understand how a reasonable person might take that more permissive view, which would probably justify a fair amount of research that causes more-than-minimal pain on its animal subjects.10
Whether one takes a view like mine or a more permissive view on animal research, a few points should be agreed upon. First, again, it is sometimes permissible to cause small amounts of pain to animal subjects consistent with the minimal-risk standard that applies to human children (assuming alternatives methods cannot replace all worthy animal research, a topic to be taken up later). Second, since causing pain is an ethically serious matter, the Humane Society's goal of eliminating pain caused to animal subjects11 can be endorsed -- at least as an ideal, if not as an absolute moral requirement.
Assuming it's sometimes permissible to cause pain to animal subjects, is there a degree of pain beyond which it is unethical to go? One might argue, for example, that we should never force animal subjects to undergo severe pain. In this spirit, Patrick Bateson, a prominent animal researcher at Cambridge University, has argued that any experiment that is expected to result in a high degree of animal suffering (his terms) should not be performed.12 Such a perspective might be defended either by an appeal to principle or more pragmatically. An appeal to principle would claim that it is simply wrong to cause some high degree of pain -- pain beyond a stated threshold -- to any animal. A pragmatic argument would contend that, even if it is sometimes right in principle to cause such a high degree of pain, we had better adopt a policy that sets a limit; otherwise, overeager researchers and those who review their work will not go all out to minimize pain caused to research subjects, resulting in a great deal of unnecessary harm in the long run. Several countries -- including Britain, Canada, the Netherlands, Sweden, and New Zealand -- have adopted pain scales of various kinds, either prohibiting experiments that would cause some high degree of pain, or at least requiring that ethics committees consider expected degrees of pain in evaluating protocols.13
So one basis for evaluating research protocols is by considering how much pain they are expected to cause. But that isn't the only important basis for evaluation: We must also consider how important the research is. How promising is it, in terms of (1) the benefits it aims to achieve and (2) the likelihood of achieving those benefits? In general, how promising must a protocol be in order to be justified?
Most parties to the animal research debate can probably agree that only animal research that is in some sense necessary should be conducted. But people will differ as to what they consider necessary. Whenever someone states that such-and-such animal research is necessary, it is appropriate to ask "Necessary for what?" Research is never intrinsically necessary; if it is necessary at all, it is necessary relative to certain goals. But what are the appropriate goals of animal research?
An animal advocate who is open to some animal research might hold that the only necessary animal research is that which has a strong chance of leading to a vaccine or cure for an illness that represents a major public health calamity, such as AIDS or a common form of cancer. (At an animal-rights rally in Washington, DC in 1990 Christopher Reeves proposed roughly this standard, but he was nearly booed off the stage by absolutist opponents to animal research.) Someone who is much more sympathetic to the animal-research endeavor might hold that all animal research that has any decent chance of leading to any biomedical breakthrough, or even confirmation of an apparent breakthrough, is necessary. And there are all sorts of possible positions in between.
Suppose animal experiments have some reasonable chance of informing us about the possible effectiveness and toxicity of a potential treatment of severe acne. Are such experiments really necessary? Well, they might be necessary to learn about this potential treatment. But is this treatment of severe acne necessary? One's answer is likely to depend on at least three factors.
First, it will depend on the importance one places on finding a successful treatment for severe acne and the projected likelihood of achieving this benefit. Second, it will depend on how one views the moral status of animals and what principles one holds regarding the harming of animals for human purposes; in applying these principles, one will factor in the expected harm (including pain caused) to the animal subjects. On this matter of animals' moral status, major disagreement is likely. If one sees animals as essentially existing for human use, one is likely to perceive many ways we can use them for our benefit as necessary. If one sees animals as having significant moral status, on the other hand, one will probably be much more conservative in deploying the term "necessary."
The third factor that plays a role in determining whether some possible biomedical advance seems necessary is one's beliefs about the viability of alternatives to animal research, such as mathematical models, computer simulations, and in vitro biological systems. The issue here is epistemological -- whether the data we can obtain from nonanimal methods are as valuable as the data we can obtain from animals -- but the discussion of this issue often seems more political than scientific. Some representatives of biomedicine apparently hold that no alternative methods can inform us as animals can because the alternative methods just aren't animals; there's no substitute for the real thing. Why they don't also hold that no animal subjects can inform us about human beings -- there being no substitute for the real thing -- is less than crystal clear. Meanwhile, some animal advocates claim that anything we could learn from animal studies could be learned from alternative methods. But I wonder how they know this a priori, since the question is an empirical one. Now, of course, many people who represent biomedicine and many within the animal protection community espouse more thoughtful and nuanced positions between these sweeping overgeneralizations.
My own view is a humble one: We really won't know how much alternative methods can achieve until we make a full-throttle effort to study them. (Similarly, no one knew whether we could travel to the moon until we went all out in an effort to get there.) The mission of studying alternatives, it seems, has only begun. Those engaged in this effort often feel that the tradition of beginning biomedical studies with animal models is so deeply entrenched that calls for much more aggressive study of alternatives are commonly met with great resistance and little real support. But such resistance may be misplaced. Even those who are strongly in favor of animal research should acknowledge some good reasons to take an active interest in alternatives, beyond the ethical advantage of reducing harm to animal subjects. According to a major British report, sometimes an alternative method is the most valid way to approach a particular scientific question; and alternatives are frequently cheaper.14 Further, biomedicine may enjoy stronger public support if it responds to growing societal concern about animal welfare with a very serious investment in non-animal methods. This means not just using alternatives wherever currently feasible, but also aggressively researching the prospects for expanding the use of such methods. Thank you very much.
- In answering this question, I draw extensively from my Taking Animals Seriously: Mental Life and Moral Status (Cambridge: Cambridge University Press, 1996). pp. 106-7.
- The last three examples appear in L. W. Sumner, "Welfare, Happiness, and Pleasure," Utilitas 4(1992): 209.
- A similar definition appears in Hyram Kitchen, Arthur L. Aronson, et al, "Panel Report on the Colloquium on Recognition and Alleviation of Animal Pain and Distress," Journal of American Veterinary Medical Association 191(1987): 1187.
- Cf. Kitchen, Aronson, et al, "Panel Report," p. 1188; and David DeGrazia and Andrew Rowan, "Pain, Suffering, and Anxiety in Animals and Humans," Theoretical Medicine 12(1991): 201.
- The above discussions of suffering and distress draw from Taking Animals Seriously, pp. 116-17. For analyses of the other mental states just mentioned, see ibid, pp. 117-20.
- See, e.g., National Research Council, Guide for the Care and Use of Laboratory Animals (Washington, DC: National Academy Press, 1996); and Council for International Organizations of Medical Sciences (CIOMS), International Guiding Principles for Biomedical Research Involving Animals (Geneva: CIOMS, 1985), pp. 18-19.
- NASA, "NASA Principles for the Ethical Care and Use of Animals" (NASA Policy Directive 8910.1; effective March 23, 1998)
- Taking Animals Seriously, especially pp. 167-71.
- See Margaret Rose and David Adams, "Evidence for Pain and Suffering in Other Animals," in Gill Langley (ed.), Animal Experimentation (New York: Chapman and Hall, 1989): 42-71; and DeGrazia and Rowan, "Pain, Suffering, and Anxiety in Animals and Humans." There is also much evidence that many animals can experience anxiety (ibid; and Jane A. Smith and Kenneth M. Boyd, Lives in the Balance [Oxford: Oxford University Press, 1991], ch. 4).
- See my "The Ethics of Animal Research: What are the Prospects for Agreement?," Cambridge Quarterly of Healthcare Ethics (forthcoming December 1998).
- This is the stated goal of a new initiative of the Humane Society of the United States (HSUS), which expects the initiative to expand to Humane Society International (Andrew Rowan, Vice President of HSUS, personal communication, June 1, 1998).
- "When to Experiment on Animals," New Scientist (February 20, 1986): 30-31.
- See F. Barbara Orlans, In the Name of Science: Issues in Responsible Animal Experimentation (New York: Oxford University Press, 1993), pp. 118-19.
- Smith and Boyd, Lives in the Balance, p. 334.