© Springer-Verlag Berlin Heidelberg 2014Grace Lee, Judy Illes and Frauke Ohl (eds.)Ethical Issues in Behavioral NeuroscienceCurrent Topics in Behavioral Neurosciences1910.1007/7854_2014_329
The Use of Animal Models in Behavioural Neuroscience Research
Philosophy Group (CPT), Wageningen University, Wageningen, The Netherlands
Research Institute for Philosophy and Religious Studies, Utrecht University, Utrecht, The Netherlands
Animal models are used in experiments in the behavioural neurosciences that aim to contribute to the prevention and treatment of cognitive and affective disorders in human beings, such as anxiety and depression. Ironically, those animals that are likely to be the best models for psychopathology are also likely to be considered the ones that are most morally problematic to use, if it seems probable that (and if indeed they are initially selected as models because) they have experiences that are similar to human experiences that we have strong reasons to avoid causing, and indeed aim to alleviate (such as pain, anxiety or sadness). In this paper, against the background of contemporary discussions in animal ethics and the philosophy of animal minds, we discuss the views that it is morally permissible to use animals in these kinds of experiments, and that it is better to use less cognitively complex animals (such as zebrafish) than more complex animals (such as dogs). First, we criticise some justifications for the claim that human beings and more complex animals have higher moral status . We argue that contemporary approaches that attribute equal moral status to all beings that are capable of conscious strivings (e.g. avoiding pain and anxiety; aiming to eat and play) are based on more plausible assumptions. Second, we argue that it is problematic to assume that less cognitively complex animals have a lesser sensory and emotional experience than more complex beings across the board. In specific cases, there might be good reasons to assume that more complex beings would be harmed more by a specific physical or environmental intervention, but it might also be that they sometimes are harmed less because of a better ability to cope. Determining whether a specific experiment is justified is therefore a complex issue. Our aim in this chapter is to stimulate further reflection on these common assumptions behind the use of animal models for psychopathologies. In order to be able to draw more definite conclusions, more research will have to be done on the influence of cognitive complexity on the experience of (human and non-human) animals.
KeywordsAnimal modelsNeurobehavioural researchMoral philosophyPhilosophy of animal minds
Much research in behavioural neurosciences is aimed at the prevention and cure of cognitive and affective disorders in human beings. These disorders, such as anxiety, depression, and alcohol addiction, have a severe impact on individuals’ quality of life. While virtually anyone would applaud the aim of neurobehavioural science to relieve human suffering, the moral acceptability of the use of non-human animals in reaching this aim is a matter of controversy. It is significant that animal models are used precisely because we consider the use of human beings in such experiments morally impermissible. If the use of animal models is morally justified, there must be a relevant difference between human beings and the animals used in these experiments that justifies the differential treatment. Yet, if these animals are indeed good models for certain psychopathologies, it might be considered likely that they have experiences that are similar to human experiences that we consider to have strong reasons to prevent or cure (such as anxiety or sadness). What, then, justifies the use of animal models? Also, in the practice of animal experimentation we see that it is considered preferable to use animals that are less like or further removed from human beings, e.g. rats rather than apes, and zebrafish rather than rats.1 Is such a preference morally justified?
To determine whether the use of specific animal models is morally justified, we need, first, to determine the basis of moral status. This is an issue that is widely discussed in moral philosophy. The concept of moral status will be explained in more detail below, but roughly, it involves whether and how much a being should count in our moral considerations. We shall see that moral status is often linked to the possession of specific kinds of capacities, e.g. sentience (being able to have negative and positive physical and psychological experiences) or rationality.
Besides determining the sensory or cognitive capacities required for moral status, we need to investigate whether specific species of animals (rats, zebrafish etc.) have these capacities. This will also help us determine whether their interests differ from ours and vary between different kinds of non-human animals. If non-human animals suffer much less from the experiments performed on them than human beings suffer from the pathologies we aim to cure, this might be considered a reason to regard the use of these animal models justified. To find an answer to the question what capacities specific species of animals have, and what they can be thought to experience, we require empirical evidence on different species of animals, e.g., data on their behaviour and neurophysiological responses in certain situations. However, there is an interpretational gap between data and meaning: between test results and what they actually tell us about what certain animals can do and experience. This is why this is also an issue in what is called “philosophy of mind”. Philosophy of mind studies the nature of the mind and consciousness, and its relation with the brain.
We do not intend to give an exhaustive discussion of all positions in animal ethics (or moral philosophy, more generally) and the philosophy of animal minds. These are very rich and complex fields, and we cannot fully do them justice in this chapter. We have more modest aims. First, to bring to the fore some of the more important questions that need to be considered to determine whether using animals in neurobehavioural research is morally acceptable and whether it is more justified to use certain animals than others. Second, we want to show that common assumptions about the moral status or capacities of animals that may lie in the background of the use of animal models in the behavioural neurosciences are not uncontroversial, and indeed, that there is good reason to question them.
We will argue that common defences of the view that human beings have a higher moral status than animals (or even that non-human animals lack moral status altogether) involve implausible assumptions or implications. We will present two very divergent positions in contemporary moral philosophy that nevertheless both defend attributing equal moral status to all beings that consciously strive to attain goals, and point out the comparative merits of these views. Furthermore, we shall also question views that less cognitively complex animals have a somehow lesser sensory and emotional experience than more complex beings across the board (or even lack consciousness altogether). We shall argue that while there are good reasons to assume that there are differences in the way that different kinds of animals are affected by negative sensory or emotional states like pain, anxiety and depression, this does not necessarily mean that less complex animals are not seriously harmed by these states.
From the outset, it is important to note that it is extremely difficult to generalise about the cognitive and sensory capacities of animals; thus, different taxa may have widely different capacities for suffering , or for coping with any suffering which may be experienced: mammals may have totally different experiences in a given situation than fish or insects . While animal ethicists tend to talk rather loosely of animals in general, especially for the purposes of this chapter, it makes a lot of difference what type of animal we are discussing. Where appropriate, we will try to specify what group of animals we are discussing, although there remains the problem in many cases that at present, we do not have perfect knowledge about the emotional and cognitive abilities of those different animal taxa, nor do we have enough knowledge on the influence of cognitive complexity on different kinds of emotional suffering.
In our considerations below, we presuppose that all neurobehavioural experiments involve some kinds of physical and environmental interferences with animals, which are aimed at making them models of specific human psychopathologies. The question is whether specific examples of such interferences are morally problematic.
2 Moral Status
To determine whether it is morally acceptable to use specific kinds of non-human animals in experiments in the neurobehavioural sciences, the first question that we need to answer is whether these animals have moral status. If animals have moral status, this means that we should take them into account in our moral decision-making. There are, however, different ways in which things can figure in our moral decision-making: directly or indirectly. Some people have thought that we only have indirect duties regarding animals. One example of such a view is that we should not treat animals cruelly only because this is likely to harden us to suffering and therefore to make it more likely that we will violate our duties to other human beings (e.g. Kant 2000, p. 6, 442). Also, it might be thought problematic to harm an animal, because in doing so, we harm the owner of that animal. However, the concept of moral status is generally used to signify that a being counts in its own right. If animals have moral status, we that should treat them in a certain way (e.g. not treat them cruelly) for their own sake, rather than for the sake of others, say, human beings. We then do not merely have duties regarding animals, but also to them.
To determine whether animals have moral status, we need to know what is a necessary and sufficient basis for moral status to be accorded to them. We shall first critically discuss some justifications of attributing unequal status to human beings and the other animals, and to animals with different degrees of cognitive complexity. These are based on some general assumptions about the nature and basis of morality that we will argue involve implausible assumptions or implications. Then we shall discuss two different approaches in moral philosophy, that both advocate attributing equal status to all conscious animals. As these two authors also conclude, we will argue that it makes sense to consider moral questions from the perspective of all beings that have an evaluative perspective.
2.2 Unequal Moral Status
One view of the basis of morality is the idea that it is in our mutual self-interest to accept moral constraints in our dealings with one another. It might be thought that animals do not have moral status, as we cannot make a mutually advantageous agreement with them, and expect them to uphold their side of the bargain by reciprocating (e.g. Morris 2011). However, we think that the incapacity of animals to reciprocate does not give us a sufficient basis for denying them moral status. Undoubtedly, a lot of rules in social life and much of the practice of politics centre around the idea of reciprocity, but this does not seem to cover the whole content of even human morality. After all, we take it to be wrong to exploit people who are too weak (or too far removed from us) to reciprocate or take their revenge on us. If we think morality goes beyond the confines of mutual interest through reciprocation, we need to find another basis for such duties.
Another proposal for the basis of (human) morality is social sentiment. Most humans are not only motivated to pursue their self-interest, but are at least to some degree sympathetic to others. The famous 18th century philosopher Hume based morality on sympathy . However, he noted that we have limited sympathies, and that our sympathy is greatest for those closest to us and similar to us (Hume 1978; Cohon 2010). While our sympathies are not limited to human beings,2 it has been noted that we are generally more emotionally attached to members of our own species (Midgley 1998). Wenz (1988), suggesting a “concentric circles ” model of justice: we have the strongest duties to those we are in a closest relationship with, and our duties to others become less strict with distance. We do not want to deny here that human social sentiments and capacity for sympathy may play a very large role in morality. We do want to question the view that our basic moral duties vary with how close we feel to the other, or what relationships we have with others, especially duties not to harm others. Hume himself noted that our moral judgments on the characters of those who harm or help others do not vary along with our sympathies for those affected. He proposed that we estimate the effects of people’s character from a “common point of view”, which abstracts from our own self-interest but rather involves the viewpoints of everyone affected by the action (Hume 1978, T 3.3.1).3 It might be argued that we have stronger positive duties (duties to assist) those whom we have relationships with, but it seems implausible to hold that negative duties (duties not to interfere) depend on the strength of (affective) bonds. Such a view could justify harmful treatment of those with whom one is or feels less connected, like those with a different ethnic background or those on the other side of the world.
A final way to argue for unequal moral status would be to resort to ‘everyday moral judgment ’ which says that rational beings, such as humans, matter more than merely sentient beings, such as many animals. Balzer et al. (2000), for example, say that it fits better with our considered intuitions to assign a hierarchy of inherent moral standing to different kinds of beings. Similarly, DeGrazia (2008) argues that moral status varies with the capacities of beings, e.g. being conscious, self-aware, moral agency, language, and so on. This does appear to be the common view. However, is this view justified? We need to ask why exactly it matters whether a being is capable of language or is a moral agent for how we ought to treat them. Sure, it would be problematic to defend a moral theory that has no connection at all to our views about the content of morality. However, we think that a view being commensensical alone does not suffice to justify moral claims. After all, we now consider views that were once common, such as the view that slavery is morally right, as completely morally unjustified. We think we need to dig a little deeper to determine whether our everyday moral judgments are indeed justifiable.
It is important here to consider what a hierarchy of moral status actually means. It means that different creatures would all have moral standing, but would have so to a varying degree. In other words, if we need to decide how to treat two different creatures, the creature with higher moral status would automatically receive preferential treatment, regardless of the specific interest of the creatures involved in that specific dilemma. So, for example, if we must choose to hurt either a rat or a human being, even if their pain would be equally severe, we should choose to spare the human being, because her/his interests matter more in principle. However, this begs the question as to why this human being’s interests matter more. It cannot be because she/he experiences more pain, because in this example the pain was equally severe for the rat and the human. Could it then be because the human can use language or is a moral agent? This raises the question why these differences would be relevant in this context. Again, more than a simple reference to common sense is necessary to explain such a position.
2.3 Equal Moral Status
So far, we have argued that three of the most common arguments for attributing unequal moral status to humans and animals are problematic. What bases could there be for attributing equal moral status? In this section, we will discuss the views of the prominent practical philosophers from two very different moral-philosophical backgrounds. Peter Singer is a proponent of the theory of utilitarianism , and a prominent animal ethicist. Christine Korsgaard is a Kantian philosopher, and has in recent years discussed the place of animals in her wider philosophical work. While there are important differences between them, the two authors both think that we have moral duties to others that are not dependent on reciprocity or sympathy for others and both are critical of everyday moral judgments. We will now explain how they justify moral claims.
Singer (1999) takes a basic starting point for the moral point of view to be that one should consider what ought to be done not just from the standpoint of self-interest, but from the interests of all involved. The basis of morality, in Singer’s view, is the principle of equal consideration of interests: all comparable interests should be weighed equally. If interests differ, however, then this should be taken into account. For example, all people have an equal interest in mobility, but for disabled people this means getting access to facilities like a wheelchair, while for able-bodied people it doesn’t. Equal consideration of interests, then, may lead to dissimilar treatment. Singer suggests that not only human beings, but also certain species of animals may have interests. Singer understands interests in terms of the satisfaction or frustration of preferences. The question then is what animals can have preferences. In Singer’s view, a minimal requirement to be able to say that a being can form preferences is that the animal can have positive or negative experiences. Singer appears to regard all negative affective states as forms of suffering which they have a preference to avoid and all positive affective states as forms of joy which they have a preference to strive for. 4 If an animal can suffer negative experiences such as pain, or fear, it will have a positive motivation, a preference, to not suffer. Such animals may also have preferences for positive states, unconnected simply to the avoidance of suffering, e.g. play or food or being with conspecifics.
Singer is a utilitarian, and that means that he thinks that in determining the right thing to do, we ought to compare, aggregate and maximise the interests of everyone involved. Thus, for example, in choosing whether to help someone with her homework, or bring someone with a serious injury to the hospital, we ought to do the latter, because that is here the more important interest. Singer noticed that in practice, even when human and animal interests are considered comparable, for example when humans and animals are thought to experience the same amount of pain after a specific procedure, the human interest is generally considered more important than the animal interest. He posed critical questions about this, and popularized the term “species-ism ”, meant to signify discrimination on the basis of biological species, which he considers as unjustified as sexism and racism. Only when different species in fact have different interests, it is justified to treat them differently. For example, dogs cannot benefit from human education, so it would not be speciesist to deny them access to schools. He also attacked the idea that it is specific capacities of human beings that make them especially morally significant, such as rationality or their being moral agents. After all, we also think that human babies’ pain matters equally to adult beings’ pain, even if they are less rational than adult humans, and we accept that just as we may not harm rational humans, nor should we harm intellectually disabled humans.
As a utilitarian, Singer thinks that we should always maximise the satisfaction of the interests of everyone involved. Traditionally, this approach to morality is most contrasted with the moral views inspired by the 18th century philosopher Immanuel Kant . Kant (1785, and more recently reprinted 1998) thinks that we should not act morally for the sake of an external goal, such as self-interest or even the interests of others, but simply from respect for moral law. He sees the moral law not as legislated by an external authority, such as God, but as a law of our own reason. In acting on the moral law, human beings are autonomous (literally: self-legislating). Kant claims that the capacity of autonomy makes human beings “ends in themselves”: we ought to respect them for their own sake, not only use them as means to another end (e.g. our self-interest). Kant thinks we do not have any direct moral duties to animals, as they lack the capacity of autonomy. He does think we ought not be cruel to animals, but that is because it undermines a duty to ourselves: to cultivate those capacities (e.g. sympathy) that enable us to do our moral duty (Kant 2000, p. 6, 442).
Christine Korsgaard, a prominent contemporary Kantian author, has offered an internal criticism of Kant’s position.5 Korsgaard (2011) argues that Kant was wrong in thinking we only have duties to autonomous beings. Like Kant, she takes morality to be based on a law that human beings legislate to themselves. As humans, we cannot simply go along with our impulses, but we need to have reasons for what we do. Insofar as we consider our choices rational, we must think that the objects of our choices are objectively good. Korsgaard emphasises, however, that the content of our reasons cannot be given by respect for autonomy itself. Rather, we find reasons in what is naturally good for us (Korsgaard 2011, p. 108). While things can be said to be good or bad for plants, only conscious animals care about their own natural good (Korsgaard 2009a, pp. 34–35).6 Animals can act purposively, to avoid things that they dislike, and to attain things they want (Korsgaard 2009b, pp. 10–15). When we avoid pain and suffering, we act for a purpose we share with other conscious animals. But even if we value ends that other animals do not share, we still value what is good or bad for the kind of beings that we are. When we, rational beings, act for the sake of an aspect of our own good, we take something’s being naturally good for us as objectively good: as a law for ourselves and others (Korsgaard 2011, pp. 107–108).
Korsgaard says that we thereby accord ourselves a certain standing: of an end in itself. Kant thought that we only have to respect ourselves as ends in ourselves insofar as we are rational, or autonomous. Korsgaard explains that Kant conflates two different conceptions of the end in itself : (1) the source of legitimate moral claims that should be recognised by all rational agents , and (2) someone who can give the force of law to his claims, or participate in moral legislation. She notes that a law can protect someone who did not participate in the making of it (2005, p. 21). In legislating a law that what is naturally good or bad for us is objectively good or bad, we confer value on our animal selves. We therefore have to accept duties to all those who have a good that they care about, even if they cannot claim respect for it. Korsgaard argues that on the basis of this reasoning, conscious animals too should be regarded as “ends in themselves” (2011, pp. 108–109). We should respect their good for the sake of the individual animals involved, and not just treat them as means for our own ends.7
Utilitarianism and Kantianism are usually understood as very different approaches, and some important differences will come to the fore when we apply these theories to the practise of using animal models in neurobehavioural research (in Sect. 4). Here, we want to point out what these specific variants of these approaches have in common. They offer basically the same reason for extending equal moral status to all animals that strive to attain goals on the basis of preferences. They attribute moral status to sentient animals, but not, say, to plants, because we can only consider what should be done from the perspective of beings who have preferences or who care about what happens to them. Cars or plants don’t care what happens to them, while sentient animals do. We can put ourselves in the place of animals, because it matters to an animal what happens to it.
We cannot completely defend these views of Singer and Korsgaard here, as this would require much more sophisticated reasoning in moral philosophy. We just want to point out that, if we think that those to whom we attribute moral status is something that is not based on reciprocity , then it seems to make sense to take perspective not just from ourselves but also from the other as an experiential being. What animals actually belong to the class of experiential beings is a matter of discussion, even amongst biologists. For example, biologists disagree about the question whether and if so, which, fish can experience pain, and whether they have capacities such as memory and flexible learning. Regarding insects and crustaceans there are even more unknowns. As we will see, an answer to this question depends on how we interpret consciousness , and this requires reflection in the field of philosophy of mind. While this subject is treated exhaustively in the chapter by Droege and Braithwaite (this volume), we are not able, nor intend, to resolve these complex discussions here, but restrict ourselves to pointing out where more research is needed and how this is relevant for animal ethical considerations.
3 Consciousness in Animals
In the previous section we have seen that certain capacities are taken to be the criterion for moral status. Singer takes preference satisfaction as morally important, and Korsgaard argues that all beings who consciously pursue purposes have moral status. These capacities involve that the being in question is sentient, and that it has positive attitudes towards certain goods—such as food or playing—and negative attitudes towards others—such as threats. These attitudes correlate with affective states. For example, fear does result in aversive behaviour because it constitutes an unpleasant feeling that motivates a being to avoid what it is afraid of. According to Singer and Korsgaard, consciousness makes a crucial difference with regard to moral status: if a being is not conscious what we do to it will not matter to it (although, of course, it may matter to us). On the other hand, if a being is conscious, it matters to the being in question whether we frustrate or aid its pursuit of goods.
In this section we will focus on the questions whether and how we can know which animals are conscious, and whether there is a difference in consciousness between humans and other animals. Of course, the list of animals that are deemed to have consciousness depends on how one defines consciousness in the first place and is constantly changing, as more research is done on species that were previously assumed to be unconscious. For example, cephalopod molluscs such as the octopus and squid were previously not considered to be sentient and cognitive beings, but are now being recognised as such. They have even been given the status of ‘honorary vertebrates’ in legislation on animal experiments in many countries (Kolar 2006).8 Like much of the literature about animal consciousness, we will focus on the question whether animals have phenomenal consciousness , which refers to the experience of sensing what is around you and the feelings and emotions that this creates; also termed ‘raw experience’ (Block 1995). We assume that when you are conscious there is ‘something it is like’ to be you (Nagel 1974). The question then is ‘can we say it is like something to be an animal’? Another way of describing this type of consciousness is as ‘the subjective state of feeling or thinking about objects and events’ (Griffin and Speck 2004, p. 6).9
As we will explain later, we think there are good reasons to believe that consciousness is not an ‘on or off’ notion, but rather that it is a matter of degree. If so, it may very well be possible that negative experiences as a result of experimentation also come in degrees. The question whether animal consciousness differs in important ways from human consciousness is important in the context of this chapter because it might be thought that, while animals do have moral status, it is less problematic to experiment on animals, if they experience less negative consequences from these experiments. This view seems to be based on the idea that animals are somehow less conscious of what happens to them. But what reasons do we have to conclude that animals are less conscious of pain and suffering than humans and therefore do not have the same interest in avoiding the negative experiences associated with experimentation as humans? The main difference between humans and animals in this context appears to be humans’ greater cognitive complexity . Therefore, we need to address the question what the influence of cognitive complexity is on suffering. In order to do this, we first need to ask if and how we can know whether animals are conscious.
3.2 Can We Know Whether Animals Are Conscious?
Both in the philosophy of mind and in biology, we encounter scepticism about the question whether animals are conscious. One reason for this scepticism is that we simply do not—and in a strict sense cannot—know exactly what animals experience. What animals actually belong to the class of experiential beings is a matter of discussion, even amongst biologists. For example, biologists disagree about the question whether and if so, which, fish can experience pain, and whether they have capacities such as memory and flexible learning. Regarding insects and crustaceans there are even more unknowns. This has led scientists in the past to ignore the study of animal consciousness. As Griffin and Speck (2004, p. 5) put it, ‘many behavioural scientists have been extremely reluctant to consider non-human consciousness on the grounds that it is impossible to obtain objective evidence about subjective experiences’. Therefore, some remain agnostic about animals’ consciousness and others simply assume that an animal doesn’t have experiences and cannot suffer pain. The obvious problem with this last line of reasoning is that it commits the fallacy of ignorance: lack of knowledge of a certain fact doesn’t make the opposite true.
Moreover, as Panksepp (2011) convincingly argues, neuroscience does now give us objective evidence about animal feelings, at least about mammals. As he explains, historically, it was believed that ‘emotional feelings are a subset of cognitive processes’ and many still believe this to be the case (Panksepp 2011, p. 4). This has meant that without higher cognitive functions, animals were not regarded as being able to experience emotions. However, animals, including humans, that had their brain’s cortex removed, still showed emotional responses (Panksepp 2011, p. 6).