A Conversation With … Peter Singer PhD, Ethicist, Philosopher, and Author of Animal Liberation Now

Peter Singer PhD. ©Andrew DaviesMy first experience performing medical research involved implanting telemetric sensors into rats in a series of experiments involving the circadian rhythms of parameters like body temperature, heart rate, and sleep-wake cycles. At the time, I was a college student, and I found the work satisfying and instructive. I learned about study design, animal husbandry, and humane care procedures, and I tried my hand at “surgery” for the first time. Although the animals were euthanized over the course of these experiments, both the principal investigators and I felt these studies had a reasonable chance of leading to translational research that could improve human health. The experience sparked my lifelong enthusiasm for the technical elements of our craft as well as for the process of discovery. The rats surely saw it differently. I now do, too. It turns out that this project had a lot in common with many animal studies I see in the course of my job as the editor here at CORR® and as a close reader of many other journals: As far as I know, the work we did never got published, and it did not influence anything that could remotely be called translational research. Although I learned from what I saw and did in that lab, I don’t think that justified the taking of life, even animal life. My experience was hardly unique; animal research is a “way into the profession” for many, many students and residents. To this, I would now say: That's too bad. I have come to believe that useless or nearly useless research that results in the sacrifice of animals actually models to learners a perversion of the very values we’re supposed to disseminate. This is especially problematic given how rarely animal experiments influence any subsequent research in humans, let alone lead to improvements in surgery or trauma care [1, 13]. By way of example, one paper in CORR [7] found that almost half of the studies using animal models that were presented over a 9-year span at the national orthopaedic congress of one large country (population 83 million) were never published, and of those that were published, almost 40% were never cited or were cited only once. Another paper [13] that analyzed studies published in the two clinical journals with the highest Impact Factor (Clarivate) in each of the 10 surgical specialties discovered that the median number of citations arising from papers using animal research in subsequent human/clinical research over a 10-year span was, wait for it, one (and the high end of the range was only five), suggesting little translation of animal studies to research in humans. Another analysis [11] determined that only about one animal study in 15 would pass ethical muster when the authors mapped animal suffering, likelihood of human benefit, and research importance, as measured by subsequent citations onto a three-axis rubric called a Bateson Cube. There’s obviously a tension here. I’ve written many times on these pages that citation counting is an impoverished approach to assessing impact. I also believe that many of the most important, lifesaving discoveries of the 20th century would not have been made without animal research, including blood transfusions [2, 12], organ transplantation, and many others [6]. In our own specialty, key discoveries about bisphosphonates for the treatment of osteoporosis—a condition whose complications kill more women than does breast cancer [20]—likely would not have been possible without animal research, either. Indeed, nearly all Nobel Prizes in Medicine awarded in the last several decades have derived key data from animal models, and more than 80% of Nobels awarded for outstanding contributions to medicine since the inception of that award in 1901 have involved experiments in animals [17]. In light of these observations, denials like those offered by People for the Ethical Treatment of Animals (PETA), such as, “We do know that animal experiments frequently give misleading results and many believe we’d probably be better off if we hadn’t relied on them” [10], strike me as overstated. Still, my own experience as the editor of a large, international, general-interest journal (and a reader of many others) has led me to believe that the vast majority of animal studies in our specialty take us nowhere and result in the deaths of tens of millions of animals each year, or more, in the United States alone [8, 9]. Because of this, about 5 years ago, we tightened our standards for this kind of research [5], and we publish less of it now than we used to. But is less still too much? To help get at this answer, I’m thrilled to be able to speak this month with Peter Singer PhD, a bioethicist and philosopher whose classic book, Animal Liberation (published over 50 years ago [14] and updated 3 years ago [15]), was primarily responsible for the contemporary animal rights movement. In it, he makes a compelling case against “speciesism,” defined in part as humans’ pervasive prejudice against taking the interest of nonhuman animals seriously. He also explains why there’s no philosophically sound basis for the belief that one species should have dominion over others, and why exercising that dominion in exploitive ways is especially wrong. He’s no absolutist, though. He has said that there are certain avenues of lifesaving research that might be ethically justified in certain situations, although such situations are “quite rare,” and “the system that we have of assessing experiments is not rigorous enough to allow only those sorts of experiments to pass” [18]. I wholeheartedly agree that the current system is not rigorous enough to identify a priori which experiments would clear that bar, so I’m especially curious to learn from him on how then we ought to make choices, whether one happens to be a laboratory scientist, a clinician-scientist, or a journal editor. It's worth noting that the movement Dr. Singer was instrumental in initiating has already substantively changed minds at the highest levels of biomedical science—where previously there was resistance to change, there now are calls for change. Last year, for example, the FDA issued a “Roadmap to Reducing Animal Testing in Preclinical Safety Studies,” which opens, in part, with: “There is growing scientific recognition that animals do not provide adequate models of human health and disease. Over 90% of drugs that appear safe and effective in animals do not go on to receive FDA approval in humans predominantly due to safety and/or efficacy issues. Animal-based data have been particularly poor predictors of drug success for multiple common diseases including cancer, Alzheimer’s and inflammatory diseases” [19]. That document urges scientists to explore validated, alternative methods, such as organ-on-a-chip systems, computational modeling, and advanced in vitro assays, and it outlines specific steps that the FDA is taking to reduce toxicity testing in animals right now. But we will only get this right when we determine when and how those new research approaches will be most helpful, and when we achieve a culture shift in terms of how laboratory and clinician-scientists explore the important questions we’re charged with answering. Please join me in the conversation that follows with Dr. Singer to find out how we might do this. Seth S. Leopold MD:I’m so grateful that you’re willing to speak with me, as I could really use some guidance on this topic. I’ve progressively grown sour on animal research over my 25 or so years as a clinician-scientist and my nearly 15 years as the editor here. Still, there’s no question in my mind that some animal research is essential, as it has shown the potential to help us save lives and limbs of humans at scale. What do you see as “acceptable” animal research, and what falls outside of those bounds? Somewhere between 93% and 99% of animal research involves mice or rats [3]; to what degree does this kind of work differ to you than, say, experiments on primates? Peter Singer PhD: When it comes to inflicting pain or other forms of suffering on animals, I do not draw a sharp line between rodents and primates. Rats and mice are sentient beings, capable of feeling pain and suffering. Hence, we should only use animals in research when: 1. Every possible effort is made to avoid inflicting pain or other forms of suffering on the animals, and if some suffering is unavoidable, when every possible effort is made to minimize it. 2. The information gained can reasonably be expected to save lives or significantly reduce pain or suffering in a large number of humans or animals—significantly more than the number of animals used in the research. 3. There is no feasible alternative means of gaining the information, whether from artificial intelligence, tissue cultures, nonsentient organoids, or other nonanimal using methods, or with the fully informed consent of human volunteers. Dr. Leopold:What about research on animals that could result in improvements in human quality of life? For instance, in orthopaedic surgery, I can imagine new approaches to treat fractures that might be better than available ones, but that would to be tested on animals before trying them out in someone’s daughter, husband, or grandmother. How should we consider those sorts of lines of inquiry? Dr. Singer: That depends on how beneficial those new approaches might be for humans, and how much suffering the animal testing would inflict on the animals. Also, we should ask how risky it would be to try out these new approaches on patients with fractures who are in a position to give their informed consent to participating in a trial of the new methods? Dr. Leopold:On the flip side, considering how we humans have an ethical capacity and ability to conduct higher reasoning, doesn’t that then make us qualitatively different and not “just another animal”? An owl exercises its dominion over the mouse to the extent necessary to feed, but one could imagine that if it could exercise that dominion to extend its life, it would. Why shouldn’t we claim a broader swath of dominion under that same authority, particularly if humans are qualitatively different or special in the way I’ve offered in this hypothetical? Dr. Singer: Owls can catch and kill mice, but that doesn’t give them any moral right to do that. It’s just something that has evolved, and as Darwin himself pointed out, it is a mistake to think that we can draw moral lessons from evolution. Dr. Leopold:Fair point. So if the problem of animal research is principally the wrongness of causing sentient beings to suffer, and if we accept the premise that speciesism is wrong such that no kind of sentient being is more or less deserving of suffering than any other, then why would it ever be permissible to cause a sentient animal to suffer in order to improve the health of humans? Dr. Singer: For the same reason that it is sometimes justifiable to cause one human to suffer for the benefit of many others, which we all accept in certain situations. For example, during the COVID-19 lockdowns, we prevented children from going to school, although they were at very low risk from the virus, to reduce the threat of them passing COVID-19 to individuals who faced higher risks from the virus. Beyond that, because animals have more limited cognitive capacities than humans, they understand less of what is happening to them, and so, with sufficient care, it is possible to use them in ways that do not cause great suffering. This type of thoughtful and empathetic decision-making process is what is missing from medical research. Dr. Leopold:Let’s shift gears from the philosophical to the practical, because we’re really lousy at anticipatinga prioriwhich studies will clear the ethical bar that you [15] and I [5] have tried to set. A possible approach that we use in other areas of moral risk would be to say that all of it is out of bounds, but seems too severe here. How should we manage that real-world problem? Dr. Singer: It needs to be a case-by-case approach, using the criteria I outlined earlier. As we have already agreed, today’s system of regulation is not sufficiently rigorous. One problem is that most of those who sit on ethics committees overseeing research on animals are scientists who tend to approve each other’s research. It would be better either to have impartial assessments or to have the number of scientists from animal-using disciplines balanced by an equal number of animal-welfare advocates. And there needs to be more transparency. It is a disgrace that in the United States, no official statistics are even kept on the number of rodents used in research, although they constitute the overwhelming majority of all vertebrate animals used. Hence, no one really knows how many animals are used in research in the US. This is in sharp contrast to the European Union, the UK, and many other countries. Dr. Leopold:Reducing our use of animal models in medical research will take a massive culture change, one that will need to include scientists, institutional review boards, and government agencies. But there’s some cause for hope—the FDA document [19] I mentioned that promotes alternative approaches makes it seem like the FDA itself is making specific changes. But you know perhaps better than most people how difficult it is to make changes happen in the real world. What can we do to hasten the culture change needed to implement these alternative approaches to animal models, particularly when most physicians do not perform research of any sort? Dr. Singer: I’m hopeful that it is starting to happen now, and that it has gone beyond the “roadmap” that you are referring to. The pharmaceutical and biotech industry’s initial response to the FDA announcement was very promising—many CEOs reported using the new approach already because “it works,” and that they were still testing animals to satisfy FDA requirements for animal data [16]. If the FDA drops this requirement, we can expect that much of this unnecessary use of animals will cease. Recent developments in artificial intelligence may also be useful in reducing the need for animal testing. I hope that physicians will contact the FDA and NIH to voice their support for these moves. Dr. Leopold:My (admittedly limited) understanding of utilitarianism causes me to want to estimate the gravity of the sin being committed here. Some 2400 years ago, a proto-utilitarian named Mozi wrote: “The task of the benevolent is surely to diligently seek to promote the benefit of the world and eliminate harm to the world and to take this as a model throughout the world. Does it benefit people? Then do it. Does it not benefit people? Then stop” [4]. By that standard, we’d get a pass altogether, but it’s clear to both of us that this is far too permissive. From the utilitarian standpoint, how large is the magnitude of harm being committed? Dr. Singer: If we amend the quote from Mozi [4] so that “people” is replaced with “sentient beings,” then it’s not too permissive at all; it’s an accurate account of how to live as a utilitarian. But to answer your question, worldwide, animal experimentation uses about 200 million vertebrate animals every year—which is a rough estimate, since the United States keeps no official record of the number of rodents used in research—and it inflicts severe suffering on many of those animals, so it’s a big issue. As I describe in Animal Liberation Now [15], humans raise and kill more than 200 billion animals every year for food, the vast majority of whom are raised in intensive farms that treat them as machines for converting low-priced grains, soy, and fishmeal into higher-priced flesh, eggs, and milk, with no concern for their well-being. Nothing about this process is necessary or life-saving, and all of it wastes most of the nutritional value of the crops and the fish we feed to the animals. It also contributes to climate change, local environmental pollution, and public health risks (including the risk of future pandemics, as well as antibiotic misuse). Between the industrialized food and medical research industries alone, the magnitude of harm and suffering humans cause animals is vast, unconscionable, and not sustainable.