Introduction
In a recent paper in this journal on the use of animals in biomedical research, David DeGrazia and Jeff Sebo defend three necessary conditions of morally responsible animal research: (1) an assertion (or expectation) of sufficient net benefit, (2) a worthwhile-life condition, and (3) a no-unnecessary-harm condition.Footnote 1 We share DeGrazia and Sebo’s “middle-ground moral perspective” that accepts that animal use in biomedical research can be justified under certain conditions, and seeks to define the conditions under which that is the case. To use their terminology, we consider ourselves “morally serious” proponents of animal research.
The purpose of this commentary is to critique their first condition of morally responsible animal research: the assertion (or expectation) of sufficient net benefit (ASNB). Although we agree, in theory, that the production of benefits is a necessary condition of morally responsible animal research, we point out problems with how ASNB could be—and often is—operationalized in the prospective assessment of animal research studies. DeGrazia and Sebo work with a concept of benefits that is similar to the definition of benefits in the European Directive 2010/63/EU on the protection of animals used for scientific purposes.Footnote 2 Two authors (ME and HG) on this paper have previously criticized the EU Directive’s definition, arguing that knowledge is the only reasonably expected outcome of a research project, not benefits, and that knowledge can only be a necessary but not sufficient condition for the production of benefits.Footnote 3 In the following, we extend this criticism to DeGrazia and Sebo’s paper.
Our positive argument defends the replacement of ASNB, in practice, with a different condition: the expectation of knowledge production (EKP). EKP understands the benefits of research in terms of knowledge production rather than what we call ‘societal benefits.’ EKP has epistemic, moral, and practical advantages over ASNB, and EKP ought to replace ASNB in the assessment, justification, and ultimately approval, of morally responsible animal research.
The Assertion (or Expectation) of Sufficient Net Benefit
DeGrazia and Sebo’s first necessary condition of morally responsible research is the Assertion (or Expectation) of Sufficient Net Benefits (ASNB). ASNB captures the reasonable idea that using animals in biomedical research is morally responsible only if the research produces benefits. We generally do not condone torturing mice or rats for the sake of it; there must be some further benefit brought about by the action.
A similar, albeit weaker notion, of ASNB is reflected in European legislation governing the use of animals in biomedical research. In Europe, the Directive 2010/63/EU says that research evaluation and approval at the institutional level take into account, “a harm-benefit analysis of the project, to assess whether the harm to the animals in terms of suffering, pain and distress is justified by the expected outcome taking into account ethical considerations, and may ultimately benefit human beings, animals or the environment.”Footnote 4 In the United States, the Animal Welfare Act (AWA) and the Guide for the Care and Use of Laboratory Animals, provide guidance for research involving animals.Footnote 5,Footnote 6 Institutional Animal Care and Use Committees (IACUCs) review protocols and consider, among other things, the rationale for and purpose of the experiment. Footnote 7 This work is facilitated by asking researchers to describe the benefits of their research. For federally funded research studies, this is typically framed in terms of benefits to society at large. In sum, approval of animal research studies in both the EU and US is premised, in part, on the expectation that the study will produce societal benefits or serve some societal purpose, respectively.
While the European and American guidance incorporates something like ASNB in prospective evaluation of research involving animals, DeGrazia and Sebo provide a deeper analysis of it, and usefully show that their more demanding conception of ASNB involves (at least) three distinct claims. First, animal research offers unique benefits that could not be obtained, ethically, without animal research.Footnote 8 Second, these unique benefits outweigh the costs and any harms caused to human beings because of animal research, and, thus, animal research offers net benefits to humans. Third, this net benefit to humanity is sufficiently important that, when differences in moral status between humans and animals are considered, it justifies the harms caused to animal subjects.
These claims constitute a far more demanding standard for morally responsible research than the conception of ‘harm-benefit-analysis’ in the European Directive, because the European Directive only considers harm to animals in the harm-benefit analysis of animal research. DeGrazia and Sebo expand the harm analysis to include harms to humans as a result to animal experiments. DeGrazia and Sebo explain what they mean by human costs and harms:
Specifically, the costs of animal research include not only the financial and opportunity costs entailed by conducting animal trials but also the following: (1) false toxicity negatives , in which interventions appear safe for animal test subjects yet prove harmful to humans; (2) false toxicity positives, in which interventions appear unsafe for animal test subjects though they would be safe for humans; (3) false efficacy negatives, in which interventions that fail to work in animal test subjects would work in humans; and (4) false efficacy positives, in which interventions that work in animal test subjects prove useless in humans.Footnote 9
ASNB is more demanding than current regulations in that it requires subtracting costs to both humans and animals in the harm-benefit analysis of proposed animal research. According to DeGrazia and Sebo the upshot of their view is that much of animal research would not produce sufficient net benefits, such that it would meet the condition. Current regulations would tend to undervalue the harms of animal research when calculating the sufficient net benefit of animal studies.
While this may be right, we are more concerned in this paper about the tendency to overstate the benefits of animal research, rather than undervalue the harms. We are not alone in this worry. In a 2014 article in the British Medical Journal, Pandora Pound and Michael Bracken argue that the benefits of animal research remain unproved.Footnote 10 In what follows we argue that the focus on benefit, both as a condition of morally responsible animal use, and in animal research oversight, while intuitive, is flawed in that it commits us to epistemic mistakes, leads to researchers overstating the benefits of research, and therefore exposes animals to unnecessary harms and suffering and erodes trust in the research enterprise.Footnote 11 We find DeGrazia and Sebo’s proposal theoretically defensible, but take issue with how benefits tend to be understood and assessed in practice, that is, as ‘societal benefits.’
After explaining in more detail these problems, we propose that prospective assessment of animal research should rely on an analysis of the expectation of knowledge production (EKP) instead of the expectation of net benefits. EKP breaks down into two further claims: (1) that morally responsible research generates knowledge worth having, and (2) that morally responsible research is designed and executed to produce generalizable knowledge. Generalizable knowledge refers to experimental results that generalize to a larger population beyond the animals studied. Generalizable scientific knowledge is reliable, replicable, and accurately descriptive. High scientific standards tend to secure generalizable knowledge. Thus, the second condition requires that the research team conducts its studies in compliance with high scientific standards.
Our analysis distinguishes production of knowledge from production of what we’ll call ‘societal benefits.’ This term is not explicitly used by DeGrazia and Sebo. We introduce it to highlight the differences between our views. Societal benefits refer to the creation of direct, translational benefits, typically to humans. DeGrazia and Sebo cite the following example of a benefit of animal research: learning how to perform a new surgical procedure by trying it on animals before trying it in humans. In this example, the benefit of research is not the mere acquisition of knowledge (‘know-how,’ in this case) but its application. Another example of a paradigmatic ‘societal benefit’ is the development of novel therapeutics following basic research with animals that revealed pathological mechanisms.Footnote 12,Footnote 13 A type of societal benefit of animal research, not typically discussed, but also not categorically excluded from an analysis of the societal benefits of animal research, is the employment of animal caretakers, handlers, and lab technicians, or the benefits to institutions from receiving grants to undertake animal research, such as funding for ‘fringe benefits’ and indirect costs.
The problem is that scientific experimentation does not aim to produce benefits; more modestly, it produces knowledge. There is a mismatch between the goals of science and the moral justification for research. Very few experiments involving animals produce direct societal benefits. To take DeGrazia and Sebo’s example, merely knowing how to do brain surgery is not by itself a societal benefit—only when this knowledge is applied in a surgery will the experiment have produced societal benefit. As has been previously argued, the knowledge produced in an experiment is a necessary, but not sufficient, condition for the generation of societal benefit.Footnote 14
Our argument for replacing ASNB with EKP in the prospective assessment of animal research is based on the idea that prospective assessment of knowledge production is more accurate and feasible than prospective assessment of the production of societal benefits, and it tracks more closely the goals and practice of science. At the outset of an experiment, assessments of societal benefits amount to professional ‘guesstimates.’ Our use of animals in research ought to rest on more solid evidence that some good will come of animal use, not mere speculation or hope.Footnote 15 In addition, EKP makes clear that ancillary societal benefits of animal research, such as employing more humans, do not count in the moral justification for animal use.
Note, too, that our analysis is limited to prospective assessment of research studies involving animals. Our paradigm case is the use of animals in basic and translational science of some relevance to human health. Retrospective analyses of research studies usefully assess societal benefits of animal research. Retrospective analyses of societal benefits are, in turn, material to designing prospective research studies, for analyzing costs of research, and for communicating publicly the importance and the actual benefits of animal research.Footnote 16 Because retrospective benefit assessment does not suffer from the same epistemic limitations as prospective benefit assessment, there remains a place for it in the assessment of morally responsible research. However, we’ll argue that prospective assessment of animal research should be based on the expectation of knowledge production (EKP) instead of ASNB. We also confine our analysis to the use of animals in biomedical research, and do not address the necessary conditions for the morally responsible use of animals in agriculture, education, or any other sector.
ASNB’s Flaws
Epistemic Mistakes
The nature and epistemic task of a scientific experiment is to test a hypothesis whose truth is unknown. At the outset of a research study, it is unclear what the outcome is going to be, i.e., whether the hypothesis will be verified or falsified. Indeed, if the outcome of an experiment was already known beforehand, there would be (a) no new knowledge gained from the experiment, and (b) in the case of animal research, the experiment would be morally unjustified.
At the same time, an experiment’s assertion of societal benefits is typically premised on the verification of the starting hypothesis. The thinking here is that if the experiment successfully verifies a hypothesis, this research will lead to the production of societal benefits down the line. For example, if basic research involving animals contributes to characterizing the role of an essential metabolic protein in food digestion, that research might support the development of therapeutics for people who are missing that protein. Prospective benefit assessment of research involving animals typically considers these societal benefits in the rationale and justification of animal use. This involves a fallacy: assuming the truth of an uncertain conclusion of the research, to support unwarranted expectations of societal benefit, when assessing the justification of animal use. One cannot guarantee a future that is, by definition, uncertain.
Just how uncertain are future societal benefits? As has been argued elsewhere, knowledge is only a necessary but not sufficient condition for the generation of societal benefit.Footnote 17 There are many intermediary steps from basic science to production of societal benefit. The development of a new drug shall serve the purpose of a proof of principle. The knowledge produced in a long series of basic research studies involving animals that build upon one another (it is hardly ever the finding of one project alone that leads to a new treatment) might be taken to support the theory that a particular chemical compound will modify disease trajectory in humans (e.g., the particular compound will modify insulin production or protein expression). This compound will need to be tested in human trials. Once approved for application, a drug first has to be produced, distributed, paid for, and administered. This might sound trivial; however, market potential, lobbying, and patient compliance are all factors beyond research that will decide whether promised societal benefits will manifest.Footnote 18
Quantifying the likelihood of production of societal benefit is incredibly difficult, and arguably impossible, owing to the number of unknown variables: (1) will the experiment verify or falsify its hypothesis? (2) will the lab be funded to produce follow up work? (3) will subsequent research verify and build upon the knowledge generated in this experiment? (4) will drug developers see market value in investing in this particular disease-modifying compound? (5) can human trials be conducted ethically? The list could go on. These and other known unknowns complicate—and arguably disallow—good-faith attempts to quantify the value of animal research in terms of its societal benefits.
It is possible to imagine cases with fewer variables. DeGrazia and Sebo’s case may be one: the surgeon who practices procedures on nonhumans before attempting them in humans. In this scenario societal benefit seems more likely. However, the argument that there are hurdles from knowledge production to societal benefit still holds. Societal benefit is only realized once the surgery has been performed on humans. Moreover, these cases are not the norm. By and large, most animal use in biomedical research falls under the umbrella of basic science and applied research that is not yet tangibly close to medical application. ASNB’s flaws are most obvious in cases where the distance from experiment to realizing societal benefit is great.
Together, the uncertainty regarding the outcome of an experiment and the uncertainty regarding the production of societal benefits make the prospective assertion of sufficient net benefits an epistemically implausible criterion for morally responsible animal research. We cannot know or even predict with much accuracy the likelihood that an experiment would produce societal benefits.
Finally, when assessing net benefits, DeGrazia and Sebo claim that one also must take into consideration counterfactual claims about what would have happened without research involving animals. Even if one acknowledges the counterfactual claim on a theoretical level, one still has to question its practical relevance with regards to prospective project evaluation, since this argument suffers from same uncertainty as prospective benefit assessment.
Moral and Practical Problems of Overstating Societal Benefits
A further problem with ASNB is that it might incentivize scientists to overstate the societal benefits of their research proposals, making accurate and objective study evaluation more difficult, for example, for IACUC members and their analogs in EU member states. In project proposals in EU member states, it is mandatory to elaborate on potential benefits in publicly posted “non-technical project summaries.”Footnote 19 This further encourages scientists to overstate the significance of their research. While one must acknowledge the role that science plays in public health, it is highly unusual for any single experiment to lead directly to breakthroughs or societal benefits. The process of new discovery is slow and windy. It is possible to retrospectively assess that a particular experiment was crucial or necessary to advancing a new discovery or leading to a breakthrough, but impossible to know beforehand which experiment will do so. In being asked to articulate the societal benefits of research to society at large, investigators are incentivized to overstate the significance and societal benefits of their research study.Footnote 20 This may also lead to false positives in the determination of which experiments meet the condition of ASNB in an assessment of morally responsible research. One may come to falsely believe that the benefits outweigh the harms, and thus approve morally irresponsible research. This is likely to occur when the societal benefits are overstated or overemphasized in the moral justification.
The overstatement of societal benefits—the false positives problem—in the prospective justification of research has at least two morally problematic implications. First, it may lead to unnecessary harms to animals. If research is approved on the assumption that an investigator’s assertion of net benefit is correct, when in fact it is not, then animals would have been exposed to harms and suffering in morally irresponsible research. Second, making promises by overstating societal benefits and then repeatedly falling short will inevitably erode the trustworthiness of science.Footnote 21 Some recent evidence shows that the false positives problem is not mere speculation. Vogt and colleagues show that authorization of animal experiments is based on confidence rather than evidence of scientific rigor.Footnote 22
False Negatives
A final flaw of using ASNB in the task of prospectively assessing animal research is that it might produce false negatives. Since, as we’ve said, scientific discovery and the production of societal benefits is slow and windy, it will often take many years of basic research before a ‘breakthrough’ is discovered. The slow, incremental accumulation of knowledge may not portend societal benefits, but nonetheless produce knowledge that is highly valuable within a certain scientific field or subfield, e.g., because it advances theory construction or methodological innovation. When societal benefits are not in sight, a reviewer of research might conclude that the benefits of the experiment do not justify the harms to animals and humans. In some cases, this may be a mistake, because although it may not be possible to show that the experiment will produce societal benefit, it can be shown to advance scientific knowledge and an important body of work.
Expectation of Knowledge Production
As we have shown, ASNB, while on its face seemingly quite reasonable, commits research reviewers, and scientists proposing animal research, to epistemic mistakes and may contribute to both false positives and false negatives in the prospective assessment of animal research. This is because it premises the moral justification of research involving animals on the production of societal benefits. Accurately predicting societal benefits—and so determining whether an experiment would produce sufficient net benefits—is never possible with certainty. And rather than promote scientific integrity, it may incentivize overstating the benefits of research, to morally problematic consequence. We argue instead, that the benefits of animal research should be understood in terms of knowledge production, and so, ASNB ought to be replaced with the expectation of knowledge production (EKP) in the prospective analysis of animal research. EKP breaks down into two further claims, discussed below: (1) morally responsible animal research will produce ‘knowledge worth having,’ and (2) morally responsible animal research will be designed and executed to produce generalizable knowledge. The first we call ‘knowledge worth having.’ The second we call ‘responsible science.’
Knowledge Worth Having
The outcome of scientific experimentation is knowledge. Our claim is that morally responsible animal research will produce ‘knowledge worth having,’ and that the prospective moral justification of animal research ought to rest on the value of the knowledge the experiment will produce. Knowledge worth having refers to knowledge that would advance an important interest. We leave open for now what it means for knowledge to be worth having, because a full defense would require clarifying what constitutes an important interest to society. It is certainly possible that some people have an interest in generating knowledge for its own sake, and so an argument could be made that some knowledge is intrinsically valuable. However, the most obvious cases of ‘knowledge worth having’ will be cases where the value of knowledge is that it is a means to an end, e.g., we need to learn how certain proteins work to develop new targeted therapeutics, to reduce disease-related morbidity and mortality. Humans’ interest is in saving lives or improving the quality of life; the basic research that makes that possible creates ‘knowledge worth having’ relative to this further end.
Our central contention is that it is possible to discern prospectively whether the knowledge produced by an experiment is worth having, but it is not possible to prospectively assess the societal benefits of the experiment. Thus, meeting EKP does not require proving that experiment will produce societal benefits such as new therapeutics or reduction in morbidity and mortality. Rather, it requires that experiments involving animals test a hypotheses that, whether verified or falsified, would advance important interests. It would be the investigators responsibility to show how the hypothesis being tested is related to other experiments and ongoing research projects, and that, whether proved or disproved, ascertaining the truth of the hypothesis advances an important interest.
One way to do this would be to describe how the knowledge gleaned in an experiment will advance a particular scientific field or line of inquiry. This is tantamount to showing that the knowledge produced would advance an interest. It would be important show that both verification and falsification of a starting hypothesis would advance the field. Both results ought to be of some use and value to fellow scientists. In her analysis of the social value of knowledge in international research, Danielle Wenner proposes that we ask the following three questions of a proposed study to determine its value:Footnote 23
1) What might we learn?
2) How probable is it that we will learn it?
3) How will that information alter the decisionmaking of relevant stakeholders?
This can be made more specific for our purposes:
1) What is the starting hypothesis?
2) How probable is it that the hypothesis will be proven or disproven?
3) How will verifying alter the decisionmaking of fellow scientists, and how will falsifying the hypothesis alter the decisionmaking of fellow scientists?
The relevant stakeholders to research involving animals are fellow scientists whose theories or ideas might be either refuted or bolstered by the knowledge gleaned from an experiment. We want to explicitly exclude from the list of relevant stakeholders those who await societal benefits, such as patients, investors, or funders. This is because the production of societal benefits depends on far more variables: funding, market forces, and clinical trials in humans, among others. It would be important to show, thus, not how the knowledge produced would be applied to produce societal benefits, but how it would affect the decisions and theories of one’s own peers and how it would advance scientific inquiry. This is equivalent to showing it would advance scientific interests. It ought to be shown not only how confirmation of a hypothesis would affect other scientists’ decisions, but also how falsification of the hypothesis would alter their decisionmaking, too.
The variables that go into assessing the value of the knowledge produced in an experiment are, thus, fewer and knowable. Responsible researchers that use animals will know and understand the relevance of their proposed experiment to their field of study, including how ascertaining the truth of a hypothesis would alter the decisions of fellow scientists, as well as their own decisions. Uncertain variables include whether a particular experimental finding will be published and whether, if published, anyone would read it. One problem is that journals are much less likely to publish ‘negative’ findings, and so other scientists might never find out that their theory has been debunked, or that their proposed experiments duplicate other studies. This is a problem. Nonetheless, it also a problem for ASNB. The advantage of EKP over ASNB in the assessment of prospective research studies is that the variables that go in to determining whether an experiment will produce knowledge worth having are fewer than the number of variables that go in to determining whether an experiment will produce societal benefits.
Nearly all knowledge advances an interest—very few things are worth knowing for their own sake. Whether a particular kernel of knowledge advances an important interest remains up for debate. Reasonable people can disagree about what counts as an important interest. A recent experiment created limbless mice to ascertain whether a particular gene may have played a role in the evolution of snakes.Footnote 24 Does knowing whether this particular gene played a role in snake evolution advance an important interest? We do not endeavor to defend a view here about what makes certain interests sufficiently important as to justify animal use. But we suggest below, how we might decide.
Responsible Science
Meeting EKP requires not only that an experiment produce knowledge worth having, but also that generalizable knowledge is producible, given the experimental design, number of animals, and other features of the experiment. This means that investigators must also show that their research study involving animals will be designed and conducted in such a way as to produce generalizable knowledge. Adhering to the highest standards of scientific integrity is essential to knowledge production, e.g., it has been shown that bias undermines the production of generalizable knowledge.
One need not look far to find evidence of ‘sloppy science’ in basic research involving animals—cases where researchers may have endeavored to produce knowledge worth having but failed to produce generalizable knowledge due to flaws in design or execution. Recent reviews of the translational benefits of research involving animals suggest that research involving animals have showed that scientific standards are not has high as one would expect, given the high stakes for animals involved.Footnote 25
Scientific experimentation is under scrutiny, and as we look at what might contribute to ‘sloppy science,’ it is worth considering that emphasizing indeterminate societal benefits in the justification of research involving animals may distract from determinate steps that ought to be taken to improve the quality and integrity of research, whether or not societal benefits accrue. Investigators must show that their experiment is designed in such a way as to prove or disprove their hypothesis, and that their team can effectively carry out the experiment. It is the moral responsibility of scientists working on animals to comply with high scientific standards that achieve the epistemic task of scientific experimentation and maximize the production of generalizable knowledge.Footnote 26
The Flaws of EKP
One might foresee a few flaws with our proposal to replace ASNB with EKP in the prospective assessment of animal research. First, one might argue that EKP suffers from the same epistemic flaw as ASNB regarding the uncertainty of the outcome of an experiment. However, as mentioned above, the advantage of EKP over ASNB in the assessment of prospective research studies is that far fewer and more easily predictable variables, e.g., verification/falsification of hypothesis, and its implication for the research field, that go in to determining whether an experiment will produce knowledge worth having than go in to determining whether an experiment will produce societal benefits. We grant that there is still uncertainty in this prospective evaluation, but the uncertainty is far less compared to ASNB.
Second, and relatedly, the verification of a hypothesis might further advance an important interest than its falsification, while at the same time verification and falsification of hypothesis might be equally likely to occur, making it difficult to assess the extent to which knowledge to be gained from an experiment will advance an important human interest. While we again agree with the claim that verification of one’s hypothesis is often more valuable to a research agenda than its falsification, it is still important to note two things. One, this also holds true for ASNB, but when prospectively assessing societal benefits, most often only verification of a hypothesis can be shown to contribute to the societal benefits upon which the use of animals is justified. Using EKP however, both the verification as well as the falsification of a hypothesis would have to be shown to advance important interests within one’s scientific field. Two, the likelihood of verification and falsification, and its implications within one’s field of scientific inquiry, can be more accurately assessed prospectively than the implications of verification and falsification on the production of societal benefits.
Third, one might think it is a limitation of EKP that it does not consider prospectively the societal benefits that might result from an experiment, and that could lead to unintuitive results. Consider the following: scientist 1 proposes Experiment A to test the efficacy of a therapy for a highly infectious, lethal disease in an animal model (e.g., Ebola Virus Disease), in what is thought to be the last step before human trials. Scientist 2 proposes Experiment B to characterize a relatively unknown and relatively poorly understood signaling pathway thought to have some relevance to malignant tumor development, but that theory has not been experimentally verified. It might seem obvious to think that Experiment A is, in some sense, more important than Experiment B, even though we might stipulate, from a scientific point of view, that both are equally well designed to generate generalizable knowledge that would advance an interest, insofar as it would equally impact other scientists’ decisions.Footnote 27 Here, it is important to distinguish between interests and important interests.
Whether it advances an important interest is a matter of ethical analysis, and of societal importance. We all have an interest in what research field or programs receive funding from, or are prioritized by, national funders. The prioritization of biomedical research ought to be a socially negotiated endeavor. The ethical analysis of what knowledge advances an important interest should not be part of the project evaluation by competent authorities and committees, but should be decided publicly.Footnote 28 Below, we explain the practical implications of EKP for the review of research at the scientific level and at the societal level.
A Practical Proposal
We think that the prospective assessment of the benefits of animal research should occur at different stages of project evaluation and in different forums than prescribed by either the EU directive or US Animal Welfare Act. The question of what constitutes an important human interest is a matter of ethical theorizing and public deliberation, and should thus be decided in the political and public arena, rather than in project evaluation by committees of mostly scientists. This happens most obviously when states and other funders prioritize research agendas, such as prioritizing Ebola research over basic research on signaling pathways. The question, “Is this knowledge worth having?” ought to be debated not in a behind-closed-doors institutional committee, but in public view.Footnote 29
This would free up institutional committees tasked with approving animal research to focus on responsible science, e.g., scientific parameters such as experimental design, inferential validity, construct validity, strength of hypothesis, quality of personnel, etc., as well as the extent to which the knowledge produced in an experiment would advance the scientific field of inquiry. The focus in project evaluation should be on guaranteeing high scientific standards to enable the generation of knowledge in the pursuit of advancing science, ensuring that the experiment produces knowledge. This may require changing the composition of IACUCs and similar committees to ensure that the committee has the proper scientific expertise. It would not be the purview of IACUCs or their analogs in the EU to assess, or even consider, the societal benefits of animal research in their assessment of EKP. They would not be tasked with debating whether the knowledge produced is important enough. Their primary function would be to assess the scientific merit of the experiment and its likelihood to produce generalizable knowledge that would affect the decisions of other scientists. This is not a new proposal; others have called for a more thorough review of scientific aspects of experiments involving animals since IACUCs were first created in the 1980s.Footnote 30 Still, many institutional committees are hesitant to undertake scientific merit review, as it is outside their purview, and many research review committees lack the expertise and time to do so. Realizing our vision would likely require changing the composition and mandate of institutional committees that oversee animal research.
Conclusion: Is this Just a Semantic Dispute?
One might think that our dispute with DeGrazia and Sebo is merely semantic, insofar as producing knowledge worth having is a benefit of animal research. On the one hand, that’s right. DeGrazia and Sebo’s ASNB may well encompass ‘knowledge worth having’ in their conception ‘benefit.’ On the other hand, ‘benefit’ casts a much wider net. It captures much more than knowledge production. As noted above, one might assert that a societal benefit of animal research is that it employs lots of people (veterinarians, laboratory managers, laboratory technicians, animal handlers, etc.). Should we consider this benefit in this justification of animal research? No, but that is not categorically excluded from an assessment of the benefits of animal research. As we’ve argued, relying on ASNB in the prospective assessment of societal benefits produces both false positives and false negatives in its determination of what experiments meet ASNB as a condition of morally responsible animal research. Correctly defining and measuring the ‘benefits’ of research matters. At stake are the lives of thousands of animals, and the integrity of scientific research. Our proposal is to limit the prospective benefits of a study—which figures in its moral justification—to the knowledge it would produce.
We suggest replacing an assessment of the societal benefits of animal research in the prospective assessment of morally responsible animal research, as ASNB requires, with the expectation of knowledge production (EKP). Assessing ASNB raises epistemic, moral, and practical problems. We do not disagree that the consequences of animal research, namely the benefits it produces, should justify animal use. But it is important to make a distinction between societal benefits and knowledge, as well as prospectively evaluating knowledge production and retrospectively assessing societal benefits.
To be sure, EKP is not the only condition of morally responsible future animal research. It is a necessary, but not sufficient condition.Footnote 31 We agree with DeGrazia and Sebo that morally justified research will also meet further conditions, such as the worthwhile life condition and the no-unnecessary-harm condition. We also agree that ASNB ought to be used to retrospectively justify animal research. But looking forward, we think that relying on EKP instead of ASNB in the practical assessment of future research studies will produce both epistemic and morally responsible research. It allows agnosticism about the truth of one’s hypothesis, refocuses the purpose of scientific inquiry on the production of knowledge rather than the production of benefits, and does not incentivize the overstatement of research benefits. We predict that this would contribute to a reduction in sloppy science and, ultimately, more benefits to society, but only rigorous empirical research would prove our hypothesis.
