Invasive research on nonhuman animals has played an important role in the recent tremendous growth in our understanding of the nervous system. From Hodgkin and Huxley’s early investigation of action potentials in squid giant axons to our modern understanding of neurotransmitters, neuroanatomy, neurology, and just about anything else with the prefix “neuro,” most of our knowledge about how brains work can be traced, at least in part, to investigations involving other species.

Understanding of the brain is both one of the great frontiers of science and immensely valuable in practical terms. Phenomenal consciousness is currently one of the greatest puzzles, arguably the greatest puzzle, in the biological sciences, and it is unlikely we will be reach any firm conclusions about the nature of consciousness without a thorough understanding of how the brain works (see Braithewaite and Droege, this volume). On the more practical side, brain diseases like Alzheimer’s and Parkinson’s affect millions of people worldwide and cures would have the potential to greatly enhance the lives of many people. Moreover, the greatest feelings of elation as well as the most intense forms of suffering , as well as any experience in between, are all mediated by the operations of the brain. Thus it is no surprise that modern science has invested a huge amount of resources on investigation of the mind.

On the other hand, few who have looked into the issue would disagree that neuroscience research has been responsible for at least some suffering in the unwilling animals that have served as the subjects in tests. Animals are poked, prodded, shaken, spun, “sacrificed,” dissected, and held captive as a means to human ends. Thus, the research has led to great benefits and, in some cases, great costs, with the costs born by nonhuman animals and the benefits accrued almost exclusively by humans. (But see Ohl et al., this volume).

The desire to advance modern medicine and our understanding of the mind comes into frequent conflict with the desire to not cause suffering , and we are left both with a constantly evolving set of regulations and with, at times, advocates at two sides of a debate that are deeply distrustful of one another. The current status quo is not satisfactory to many animal advocates, yet any changes that further restricted research would likely be vigorously opposed by the scientific community.

What if, however, we could eliminate many of the harms caused by the current system while preserving the momentum of the recent cognitive revolution? Wouldn’t it be clear that we should take this opportunity?

I will argue that we already have, or are extremely close to having, the capacity to dramatically reduce the amount of suffering caused in biomedical research via genetic modification of the animals used in research. The significance of this possibility will vary according to different views; for those who think that the advancement of knowledge and human well-being trumps all concerns about animal welfare, there will be an opportunity to eliminate large amounts of animal suffering without any impairment in the forward march of science. For those who think that progress needs to be carefully balanced against any harms inflicted on animals, the implications for current practices will be even more dramatic. But, most importantly from my perspective, on almost any view that takes the elimination of unnecessary suffering to be a valuable end of moral conduct, there are options available that can reduce suffering with minimal or even, I would argue, nonexistent cost.

My plan in setting forth this case begins with providing an overview of different ways that animals can suffer in neuroscience research, in particular classifying the results based on the relationship between the capacity to suffer and the potential knowledge gained from the research. I will then provide a brief review of experiments that have already arguably reduced the capacity to suffer in rats and mice, the species that make up the overwhelming majority of research animals. Next I argue that, consistent with any view that considers nonhuman suffering morally significant, we should change from the current status quo to a system where welfare-enhanced (more on this terminology below) animals are used whenever doing so does not impede the likelihood of gaining knowledge. I will also argue that we can further reduce suffering by using welfare-enhanced animals in initial tests in order to reduce the number of nonenhanced animals who suffer. I finally examine how my view fares against various objections, starting first with criticisms that have been raised against using welfare enhanced animals in agriculture, and then moving on to consider other objections.

Before proceeding, I need to flag a couple of assumptions that are central to my arguments. First, I assume that all vertebrates are sentient. Though there are still some skeptics who argue that only humans have phenomenal consciousness , I think the best evidence to date suggests that vertebrates, and particularly mammals, are likely to be conscious (Report of the Committee on Recognition and Alleviation of Pain in Laboratory Animals 2009, see also Kaldewaij and Bovenkerk, this volume). I further assume that though it is conceptually possible for an organism to be sentient but not capable of suffering (as will be important later), vertebrates and especially mammals are also capable of suffering.

Suffering is an evocative term. Many researchers prefer to use the terms “distress” or even “pain” when referring to nonhuman animals to avoid the connotations of suffering and to leave open the possibility that there might be differences between human and nonhuman experience. However, I use the term to refer to unpleasant overall experiences, which should be unobjectionable to anyone who agrees that some nonhuman animals are sentient. We can remain neutral on the question of whether human suffering is more intense or more profound than that of other species while still agreeing that it is bad from the perspective of the sentient organism to be in a state of suffering as it is defined here.

I link suffering to “overall experience” because there may be individual aspects of experience that are unpleasant but which are outweighed by other features. For example, runners might feel some aches and pains while jogging which are trivial compared to the overall rush of endorphins. Or, as Rollin (2012) has pointed out, dogs may choose to undergo a mildly painful procedure because they are looking forward to the reward at the end. So when I use the term suffering, I don’t intend to refer to mild or short-lived experiences that are part of overall positive experiences; rather, I mean experiences sufficiently long and intense to cause the overall experience to be unpleasant.

On this definition, many animals suffer in a wide variety of behavioral neuroscience experiments. Mammals have been used as animal models of acute pain , depression , anxiety , fear, nausea, disgust, startle, as well as a number of pathological conditions and models of neurological disease such as Alzheimer’s disease, Parkinson’s, chronic pain, and so on. In many experiments, the animals’ condition is managed with anesthetics that render the animal unconscious or with analgesics that diminish the pain (at least during invasive procedures; Carbone 2011 has pointed out that there is often very little information in studies about how pain is managed post-surgery). In other cases, it is deemed necessary for the experiment that the animals do not receive pain management. Of course, it is worth noting that many neuroscience experiments do not involve any procedures that would necessarily cause any suffering.

Much of the harm to animals via the institution of animal research might come not from the experimental procedures themselves, but rather from aversive experiences caused by conditions the animals live in, their handling, their transportation, the technology used to measure brain activity, and any imperfect execution of the welfare policies put in place by various laboratories. Rollin (see this volume) has emphasized that the confined environments of laboratory animals does not allow them to exhibit a wide range of species typical behavior, or telos as he puts it. A related point can be made purely by focusing on experience; first, pleasure often exerts inhibitory influence on unpleasant feelings (Leknes and Tracey 2010), so an environment that provides very little opportunity for pleasurable experiences might amplify any potential negative feelings. And second, the inability to act on biological drives might itself be a negative experience. Thus, the mere existence of animals being used for research, even if the experimental design does not itself cause any discomfort, may count as harming the animals, if the unpleasantness of their conditions outweighs any positive aspects of well-being included in their life.

Putting aside the harms that may occur outside of the experimental context, it will be useful for my purposes to sort various neuroscience experiments into three categories. The first category of experiment is one where any suffering experienced by the animal is completely irrelevant for the scientific outcome that is being tested. For example, one set of experiments involves measuring the neural activity in rhesus monkeys as they are restrained and presented with visual information. Surgery is performed to implant a skullcap on the monkeys, which facilitates the direct measurement of individual neurons or groups of neurons. Animal activists and researchers strongly disagree about whether this set up is inhumane and causes aversive experiences for the animals. I have no intention of weighing in on this debate here; however, I think it is clear in this situation that suffering is in no way necessary for the results of the experiment. Assume that the experiments were performed on rhesus monkeys that lacked the ability to feel pain, anxiety, and fear. There is no (or rather very, very little) reason to think that an inability to suffer would in any way interfere with early visual neurons’ response to their stimulation. Thus, the validity of these results would not be compromised by using animals incapable of experiencing those feelings but otherwise identical to the animals that were used. I will refer to these sorts of experiments as Aversion Independent.

A second category is experiments where a form of suffering or a behavior correlated with a form of suffering is a direct object of study. In these cases, the scientific validity of the experimental design depends on the capacity of the animals to suffer. Experiments investigating the neural underpinning of pain, depression, and learned helplessness are examples of this. In these cases, if the animal’s capacity to suffer is eliminated, then the experiment will no longer be capable of obtaining the information it is intended to measure. Clearly, for example, it would not make much sense to study the efficacy of a proposed new analgesic on an animal incapable of feeling pain . I will refer to these types of experiments is Aversion Dependent.

And between these two categories exists another category where states associated with suffering are not the object of study, but nevertheless might be plausibly thought to indirectly contribute to the mechanisms or capacities under investigation. Consider research that investigates the effects of a certain drug on a neurodegenerative disease. While the direct object of study is not directly related to suffering, one might think that interfering with certain negative emotions could potentially cause downstream effects that influenced the strength of the model. For example, say that mild stress influences the body’s reaction to the drug, or the immune system’s response, or the rate of degeneration; in these cases, the elimination of suffering could impact the validity of the results, although the extent is not entirely clear. I will refer to this type of study as Aversion Related.

Thus we are left with three types of experiments: those where the direct object of study is a form of suffering or a mechanism associated with suffering, those where the object of study might be indirectly influenced by the capacity to suffer, and those where the object of study is entirely unrelated to the capacity to suffer.

Neuroscientists now have the ability to “knock out” certain capacities via the insertion of altered DNA sequences with inactivated genes into embryos that develop into adult animals. This technique is routinely used to study the development of certain capacities and the cellular building blocks that compose various mechanisms governing behavior. Though few researchers make claims as dramatic as saying they have eliminated an aspect of consciousness , I think it will be clear from the range of capacities that have already been eliminated under various circumstances that we are already close to having a tremendous opportunity to diminish the capacity for suffering in rats and mice, the animals that make up the vast majority of nonhuman research subjects.

As I have detailed elsewhere (Shriver 2006