A case for integrity: gains from including more than animal welfare in animal ethics committee deliberations

Abstract

From January 2013, a new EU Directive 63/2010/EU requires that research using animals must undergo a harm–benefit analysis, which takes ethical considerations into account (Art. 38 (2) d) – a so-called ‘project authorization’ (Art. 36). A competent authority in each member state has to ensure that no project is carried out without such a project validation process, but often delegates the actual assessment to an animal ethics committee (AEC) or its equivalent. The core task of the AEC is to formulate a justifiable balance between the animals' suffering caused by research and the potential human benefit. AECs traditionally focus on animal welfare issues, but according to the new directive other public concerns must also be taken into account. Taking the new EU Directive as a point of departure, the central aim of this paper is to discuss the evaluation process in relation to animal welfare and animal ethics through the concept of animal integrity. A further aim is to elaborate on possible improvements to project evaluation by considering animal integrity. We argue that concepts like animal integrity are often left out of project authorization processes within AECs, because animal ethics is often interpreted narrowly to include only certain aspects of animal welfare. Firstly, we describe the task of an AEC and discuss what has typically been regarded as ethically relevant in the assessment process. Secondly, we categorize four notions of integrity found in the literature to show the complexity of the concept and furthermore to indicate its strengths. Thirdly, we discuss how certain interpretations of integrity can be included in AEC assessments to encapsulate wider ethical concerns and, perhaps even increase the democratic legitimacy of AECs.

According to statistics from the European Commission, about 12.1 million animals were used in European research in 2008. ¹ Animals are employed in medical science, because their similarities to humans enable relevant results. However, these animals are only suitable as research subjects, because they are believed to differ sufficiently from humans in an ethical sense to justify that they are not given the same protection as human research subjects. This is a well-known contradiction in medical animal experimentation to which there is no simple response.

Motivated partly by EU Directive 63/2010/EU (hereafter also called ‘the Directive’), ² there is a growing interest within the European regulatory system for tightening approval procedures for animal use by emphasizing animal needs. This also acknowledges growing concern in many countries for animals and for the ways in which they are used for research purposes, in the agricultural sector, and as pets. From a legal perspective, these efforts have mainly stressed animal welfare. The same holds for the Directive of January 2013: according to this new Directive, animals must be treated well and protected from unnecessary pain and suffering (preamble 12, 13, Art. 33). However, research on animals can be performed, if ethically assessed and approved in a ‘comprehensive project evaluation’ or ‘project authorization’ (Preamble 38, see also 39 and Art. 36, 38). In the following, we refer to this process as an ‘ethical evaluation’ or ‘assessment’ and to the board by each member state performing the evaluation as an ‘animal ethics committee’ (AEC) (equivalent to the term ‘animal care and use committee’ (ACUC), as used in North America).

It is well known that public concern for animals (including livestock) and their moral status is not limited to the issue of welfare. Other aspects such as integrity, food quality, dignity, and naturalness also play important roles.^3–6 So, to what extent should such issues be taken up by AECs? One response is that only issues which fall under the sensu stricto definition of animal welfare as used in the Directive, should be considered.

In an ethical assessment core concepts need to be clear, and hence it is worth noting that ‘animal welfare’ is a matter of much discussion. ⁷ Traditionally three main perspectives are considered: the biological approach focusing on the animal's ability to cope with its environment; ⁸ the affective state approach which focuses on the animal's subjective experiences;^9–11 and finally the natural living approach which stresses species-specific or ‘natural’ behavior.^12–14 Further the relationships of these different parameters, their values, and ways to combine them in a full assessment are important.^15–17 Also, the concept of animal welfare is embedded in the broader ethical discussion of what constitutes a good life, such as hedonism, perfectionism, and preference theory. ¹⁸

Thus, by focusing solely on animal welfare, many other ethical concerns regarding the use of animals within research are not included in the assessment. If so, there could be a risk of a growing discrepancy between the scope of an assessment on the one hand and public expectations for the committee on the other. The aim of this paper is to discuss this discrepancy by taking the intention of the new Directive to meet public concern as the point of departure. By extension, the question must be considered whether or not it would benefit assessment procedures to be based on a broader understanding of ethics than is currently the case. The concept of integrity will be used as an example to illustrate the argument.

The ethical limits of animal biotechnology are frequently debated.^19,20 Often these ethical discussions are divided into issues of: (1) human and environmental risk; (2) animal welfare; and (3) other moral concerns. ⁵ The focus is more often than not on the first two. ²¹ This approach, however, often overlooks the value-based decisions that underlie the first two concerns and regards them as scientific discussions only.^5,22,23 Furthermore, it neither reflects legal documents nor public concerns about biotechnology, where an issue such as integrity plays a larger role than is often expressed in scientifically-based debates on risks.^5,24–29

The concept of integrity also plays a small yet increasingly persistent role in academic ethical discussions of animal biotechnology (see examples^22,30–38). This paper examines the notion of integrity (definition below), and discusses how and why the concept could be included in AEC assessments, drawing on different understandings of integrity found in public surveys and academic ethical sources.

In the first and second section of the paper we describe the task of AEC project authorization, which according to the new EU Directive, is also called ‘project evaluation’. This is initially done by drawing parallels to the task and structure of the Swedish AECs, which are of special interest since Swedish national legislation from 1988 in relevant respects is close to the new Directive, and where there is an obligation to relate approval to ‘public interest’. ³⁹ We here distinguish between legislation claims, i.e. what would be an ideal situation and what seems to be current reality. In the third section, we describe the concept of integrity, looking at how it is used in both scientific and philosophical debates and at different levels. This gives rise to a taxonomy involving four different classifications of integrity. In the fourth section, we discuss the possible application of this concept of integrity in AEC assessments. Finally, we explore how integrity might be better suited to capturing ethical concerns about the human–animal relationship that animal welfare, even in a broad sense, does not. In turn this leads to the idea that the inclusion of integrity might contribute to a democratization of AEC decisions.

To bring these aspects into the context of practical use, we use EU Directive 63/2010/EU on research animals, but also the Swedish regulation for AECs showing that aspects other than medical benefit and animal welfare are also to be considered in the ethical assessment, insofar as such aspects are present in society. We thereby aim to show that reflecting on a broad spectrum of ethically relevant issues can meet public concerns, democratize the assessment procedures, and help them fulfil the statutes under which they operate. A different strategy would be to leave the discussion of narrow animal welfare issues to the AECs and to create a corresponding body that should deal with the broader ethical issues pertaining to the use of animals in research. This body could then recruit members with special knowledge about these subjects and relevant stakeholders. This is a far from ideal way of handling the task: from our point of view it is crucial to bring the welfare and integrity discussions together as they are but two aspects of the same ethical discussion. Keeping these aspects separate will only result in a deeper dichotomy between animal welfare thinking and other animal ethics considerations – a divide we seek to overcome by focusing on whether and how the concept of integrity can be brought into the discussions of the AECs.

1. AECs and the role of the public

In this section we discuss the role of animal welfare and inclusion of public concerns in the project authorization process. As the construction of a system fulfilling the requirements in the new Directive is a complex task, the experiences of those member states that already have such institutions is relevant. Here the Swedish AECs are of particular interest, because they have a policy of including lay members.

1.1. Project authorization in the light of public considerations

Historically, some arguments in favor of public control were based on an idea of a self-evident, Christian duty to consider animal suffering to be morally relevant and on the belief that scientists generally share this view, i.e. that control would be no obstacle. Arguments against control were based on trust in methodological necessity and sufficient moral judgment among scientists. Hence, many issues other than a concern for animal welfare have traditionally influenced the debate on research animals,^40,41 not seen in the new Directive. Rather, arguments for public control refer to animals as sentient and morally relevant beings (preamble 12), and public control – in terms of a comprehensive project evaluation of experiments by an AEC – is presently a precondition for performing experiments on animals in many countries. This is often underpinned by the argument that it ensures public acceptance of research methods and objectives.^2,41,42. In the same preamble the EU Directive not only indicates the importance of the public being informed about projects, but also that ‘[t]here are also the ethical concerns of the general public as regards the use of animals in procedures’. ²

In Sweden, AECs' decisions on whether to approve or reject an application have been legally binding since 1998, and the political aim has been to ensure that the ethical assessment process is accepted from a democratic point of view. ^41,43 The Swedish Animal Welfare Act states: ‘The committee may only approve an animal experiment application if such use can be considered important in the public interest and the conditions set out in section 19, points 1–3 are fulfilled’. ³⁹ Hence, in the Swedish AECs, ‘public interests’ need to be taken into consideration in each project evaluation. Clearly research is one such interest, but other interests may be regarded as being more important than specific research interests. Such an inclusive view of what is to be considered in a project authorization therefore follows two ‘channels’, both relevant for ensuring a democratic evaluation process: (1) considering what research is of common interest; and (2) reflecting public views on animal experimentation given the structure of lay members appointed to the AECs.

In addition, all applications sent to the Swedish authorities by default become public documents. This ensures a high degree of transparency, which is a cornerstone of a democratic system. Although preamble 41 of the Directive states, ‘To ensure that the public is informed, it is important that objective information concerning projects using live animals is made publicly available’, the requirement for openness in the actual article (Art. 43) is limited to an anonymous, non-technical summary of the project. ²

The general task of project authorization, according to both the Swedish Animal Welfare Act and the new EU Directive, is to consider whether there is a need to use animals and, if no alternatives are found (Directive preambles 10, 12), to evaluate the importance of the research (Directive Art. 38). Each project authorization must ensure that: (a) animals are not used if their use can be replaced by other research methods; (b) that the principle of the 3Rs (replacement, reduction, and refinement) has been fully applied; (c) that animals are not caused undue suffering; and finally (d) that there is no unnecessary repetition of research.^2,39 As expressed in a detailed, Swedish, administrative provision: ‘The AEC must make an analysis of suffering and benefit from an ethical point of view where animal suffering is balanced against expected benefit of the experiment for humans, animals or the environment’. ⁴³ This, (and the parallel claim in Art. 38 2 d) expresses a goal-oriented (teleological) framework, maximizing human benefits and minimizing animal suffering. Such balancing is not an easy task and requires skills in recent (medical) research, animal science, animal welfare, and in applied ethics, since defining ‘unnecessary suffering’ is dependent upon the interpretation of complex factors, e.g. the importance of the suggested research goals, the extent of animal suffering, and an evaluation of what is ‘undue’ or ‘unnecessary’.

The statement in Directive preamble 12, that ‘animals have an intrinsic value which must be respected’, becomes very relevant in the light of respecting public views on animals. Given the structure of Swedish AECs, it is possible to expect a nuanced discussion on animal welfare and inclusion of wider ethical concerns such as animal integrity in the committee's justification of its decisions.

2. Ideals and realities in assessing research on animals

The following section discusses the tension between the ideal described in the Directive regarding project authorization, and the practices, as described in the literature, of some AECs. Although the literature builds on studies performed before the Directive was implemented it is relevant as it outlines a kind of ‘state of the art’ of the process of project evaluation. Moreover, changing of procedures is a slow process, i.e. after the implementation of the new Directive each country needs time to amend its project authorization process and system. Further, in order to ensure that the evaluation process fulfils the requirements in the Directive, current difficulties must be considered.

Our point of departure is that a committee structure or authorization practice that does not request sufficient information on animal suffering and research relevance runs the risk of institutionalizing deficient application scrutiny. However, as public ethical concerns go beyond animal welfare concerns and may relate to, or be expressed in terms of, concepts such as animal integrity^5,24–27 this implies that an assessment based only on a biological and/or affective view of animal welfare could fall short of fulfilling its task of including relevant public ethical concerns. If public concerns are not taken into consideration, AECs run the risk of losing legitimacy in the assessment process. For this reason, it is a necessary reponsibility for democratic societies to handle decision-making in multidisciplinary, hierarchical contexts where expert knowledge is combined with ‘laymen’s' views. ^44–46 For instance, what issues are to be discussed and how can human benefits be compared with animal suffering? There seems to be an inherent conflict when seeking to ensure a scientific basis for the decisions, all the while legislators claim to include non-scientific views in a broader project authorization.^40,41 Covering all public concerns is of course impossible, but as mentioned, by including an equal number of lay members and experts in the Swedish AECs the intention of the legislators was to ensure an ethical and societal legitimization for using animals in research. ⁴¹ Hence we find it justifiable to assume that the inclusion of lay members enhances the chances of broader ethical concerns being addressed in the committee discussions, rather than would be the case if only scientists decided the agenda.^45,47,48 An AEC constituted of an equal number of lay members and experts increases the possibility that ethical concepts such as integrity and intrinsic value are included in the discussion and thus ensures that public concerns are mirrored in a better way than in an AEC consisting only of scientific experts.

Several recent studies indicate that many committees consider the benefit of potential research outcomes of greater importance than animal suffering. ^{6,45–47,49–51} Some studies on Swedish AECs have shown disparate perceptions of ethics and of the purpose of ethical assessment among committee members. ⁴⁵ In addition, there have been examples of very limited discussions of broader ethical issues in Swedish AECs. ^{41,45–47,49–51} This is not an exclusively Swedish phenomenon, as Schuppli et al.^6,52,53 have demonstrated in a Canadian study. Kolar & Ruhdel ⁵⁴ and Alzmann ⁵⁵ have discussed whether German AECs perform too narrow an ethical evaluation. Also, Lehto & Hirsjärvi ⁴⁷ have shown how Finnish AECs have a narrow understanding of the issues at stake and how the arguments formulated by scientists are given the greatest weight. These studies indicate problems pertaining to an emphasis on details (such as small changes in the number of animals used, or technicalities); to excessive time being spent on clarifying aims and medical discussions; and on the wording of the protocols (that according to law should be, but are often not, accessible to laymen and researchers from other fields) in many committees. Consequently issues such as alternative methods, the appropriate level of suffering for each individual animal, and scrutinization of animal welfare parameters are seldom discussed in any depth, nor are other ethical aspects such as intrinsic value or animal integrity taken into consideration.^6,41,45,47

In the studies mentioned above it is shown that harm–benefit analysis is often understood in narrow terms, and excludes broader citizen perspectives on animal research. Rather, the boundary for ethical reflection in an AEC is set through what is already common practice. ⁵¹ In yet another study initiated by a responsible governmental authority, Dahlborn found a variety of discussion themes and decision methods as well as differences in how the assessment procedure (e.g. handling of insufficient information and discussion of alternative methods) was completed. Deficiencies were found, e.g. regarding how comprehensively the harm–benefit analyses were performed and with regard to comparability between the committees in similar applications. In sum, the ethical considerations were found to be lacking in depth. ⁵⁶ Taken together this clearly indicates a significant risk that these AECs regard information about potential human benefits to be more important than knowledge about the conditions for the individual animal and its suffering, i.e. they have difficulties in fulfilling the ethical balancing of harm and benefit. A similar conclusion is drawn in other studies by committee members who are non-scientists ⁴⁹ and researchers. ⁴⁷

The picture of limited or lacking ethical dimensions included in the committee discussions makes it possible to sketch at least three problems. Firstly, animal welfare is regarded as the only ethically relevant variable concerning animals in the assessment process. Secondly, acceptance of deficient and one-sided information about the animals and their welfare makes ethical harm–benefit balancing virtually impossible. Thirdly, if project authorization fails to take public concerns into account there is a latent risk of a growing mistrust towards an AEC system ⁵⁷ (see also Rollin ⁴² ). Based on this discrepancy between ideals and the actual project authorization processes, we see a risk of public trust in the evaluation system deteriorating, if there was a wider perception of such an imbalance.

In conclusion, it seems fair to state that: (1) compared with potential human benefit animal welfare issues are often not given equal weight in the AEC assessment process; and (2) other ethically relevant aspects are included to an even lesser degree. This is partly due to a narrow idea of what ethical issues are relevant to the assessment and – we would claim – insufficient breadth of perspective and conceptual inclusion as is needed for any balanced ethical discussion and assessment. A clarification of how a concept such as integrity might generate a more inclusive ethical discussion on the use of animals that better mirrors the ethical concerns of citizens will therefore be useful in examining how the scope of AECs can be broadened and thus embody the ideals in the new EU Directive. Thus, the following section will elaborate the notion of integrity as a tool for facilitating this aim.

3. The concept of integrity – a two dimensional model

There are several definitions of ‘integrity’. We suggest they can be subdivided into four perspectives according to two variables. The first variable distinguishes one mainly scientific and one mainly phenomenological perspective. The second variable divides these into individual organism level and species level considerations. ⁵⁸ By ‘scientific’, we mean a way of understanding integrity that is based primarily on the biology of the animal – especially its genotype. Thus, focus is on the actual genetic change to the animal rather than on the genome it would have had, had it not been genetically modified. By ‘phenomenological’, we mean a way of understanding integrity where the everyday experience of the animal–human relationship (the Gestalt of the animal) is in focus. This notion ties in with philosophical and sociological approaches termed ‘phenomenological’, which are concerned with the overall structure of experience, our access to reality, and the nature of emotions and feelings (cf. Section 3.2).

By making these distinctions we seek to achieve two goals: (1) to clarify the concept of integrity to make it more operational; and (2) to present a philosophical interpretation of the non-empirical (i.e. not relying directly on data-based evidence) skepticism that many people express towards the use of biotechnology on animals.^3–6 Please note that ethical issues are discussed both under the heading of ‘science’ and ‘phenomenology’, as the notion of integrity is used normatively within all four perspectives.

Despite many attempts to define animal integrity, the concept remains vague and elusive. This, as we see it, is due to the complexity of a concept that seeks to express the normative independence of animals from humans as it is experienced before the animal is reduced to being an object for consumption or science^29,59 – a process also known as ‘the de-animalization of the animal’. ⁶⁰ Understood in this manner, integrity should be viewed as a concept that opens up a discussion of human experiences with animals, rather than a concept that seeks to close the discussion in a narrow definition of its content.

The word integrity is derived from the Latin integer meaning wholeness, completed, untouched, unharmed. In the literature, the definition by Rutgers & Heeger ³⁴ is often seen as a point of departure. Here integrity is understood as the ‘wholeness and intactness of the animal and its species-specific balance, as well as the capacity to sustain itself in an environment suitable to the species’. To a large extent, violations of animal integrity will therefore also violate the welfare of the animal, if animal welfare is understood to also include mental experiences. ³¹ Other practices can also be seen as a violation of integrity, such as breeding animals that fit production systems better, rather than developing housing systems that satisfy the behavioral needs of the animals. ⁶¹ Insertion of genes into pigs that enable them to mimic human diseases such as Alzheimer's disease ⁶² is another example.

Other important notions brought into play in trying to elucidate the wider discussion of integrity include telos ³² and a notion of inherent worth.^34,37 The literature in the field thus often ends up evaluating whether the close link to biological and genetic processes that most definitions rely on makes sense (see also examples^22,31,32).

The non-empirical character of integrity is often discussed in relation to the applied science of animal behavior. Bovenkerk et al. ³⁵ have found this concept of integrity too vague to be helpful, since it does not ‘refer to a state of affairs that can be assessed empirically’ (see also examples^63,64), but they see it as bridging ‘a gap between moral theory and moral experience’. We agree it is exactly at this point that the concept of integrity has a role to play. Even though we reject the Bovenkerkian assumption that integrity cannot be used in moral theory, as it is not open to empirical studies, ³⁵ we find that integrity is a concept that can bridge what is empirically assessable and what is ethically relevant and experienced at a phenomenological level.

3.1. The species level from a scientific perspective

Violation of integrity at the scientific species level is often seen as being caused by changing the species through the genetic modification of one or more individuals. However, as Sandøe & Holtug ³³ have pointed out, this implies that traditional ways of changing the genome of a species (such as selective breeding) would also be wrong. Thus, it is argued that traditional breeding may bring about just as dramatic changes in the appearance of an animal as genetic modification in a laboratory. As such, both practices can be seen as tools, which give rise to the same ethical issues concerning integrity. This way of interpreting integrity also leads to a claim that the extinction of a species is a violation of their integrity, since this wipes out its genome.

The idea that changing the genetic make-up of a species is a violation of its integrity in and of itself also seems rather difficult to defend within a biological framework, because the genes are always in flux – even without human interference. The biological changes in the genomes of species happen far too rapidly for integrity to be present in a way that requires protection and may hence not be used as a reference point.

The idea that integrity, at the scientific species level, can be violated can be understood in a different way, however. Here it is specifically linked to situations where genes are transferred across insurmountable species barriers in nature (e.g. inserting a gene from a jellyfish into a pig). The genome of the species is thus changed in a way that could not happen in nature. There are, however, two objections to considering this a violation of species integrity. The first relates to the unclear ontological status of the species concept. ⁶⁵ The other relates to the fact that the genetic change at the genetic level can be seen as any other change caused by humans or naturally occurring mutations, as it is simply that a number of chemical bases are altered.

3.2. The individual level from a scientific perspective

If one takes the concept of integrity to be linked to physiological and empirically discernible phenomena in the individual animal, it seems that some socially-accepted practices become ethically controversial. Breeding, through genetic modification or traditional techniques, to produce individuals with reduced capacity for suffering in intensive production systems, is a frequently used example of a controversial use of biotechnology on animals.^32,61

Moreover, it could be argued that people might be apprehensive about the production of ‘hybrids’ between species. This worry could be explained in terms of an ontological insecurity, i.e. drawing upon Gidden's ⁶⁶ notion of ‘ontological security’ as a sense of continuity and order in people's lives. While this worry seems to capture something which many might feel, the question is, however, whether this can actually be seen as a worry about integrity as such, or maybe rather, as a worry of the consequences that stem from living in a society producing such animals or, from having them in society.

Nevertheless, it is hard to find an ethically relevant difference between this and traditional breeding goals when perceived from the perspective of physiological integrity. As Bovenkerk et al. ³⁵ have noted ‘though we have a prima facie duty not to interfere with an animal’s interests, there is nothing wrong with changing these interests’. The same argument can be found with Rollin, who argues that it is ethically important to allow the animal to perform its species-specific behavior, but ethically acceptable to change this behavior through the means available. ⁶⁷ The integrity of the animal can be said to be intact – it is simply a different kind of animal compared with its predecessors – and therefore it has a non-violated integrity.

Here, however, seems to be a situation where there is a gap between moral theory and moral experience. ³⁵ This gap can be described as the difference between how our moral obligations are experienced in everyday life, compared with what makes sense from the perspective of a moral theory which is tied to a scientific perspective on individual animal welfare. We make moral choices throughout our lives. The question of how we make those choices has been subject to, to say the least, intense philosophical debate. According to the British philosopher, Mary Midgley, human experience should be central to all ethical decision-making. ⁶⁸ Instead of only a systematic evaluation of ethical dilemmas or considerations on how to justify certain acts morally, emphasis is on the immediate experience of being confronted with an ethical issue when told that, for example, animals are changed radically to fit production systems.

Bovenkerk et al. ³⁵ argue that the more integrity is removed from a scientific, empirical framework, the less it can be used in ethical debates. We will claim, however, that integrity becomes better at containing the moral experiences that initially gave rise to it when removed from this framework and thereby better adapted to bridging the gap between moral experience and empirical physical impacts on animals. To get to that point it is necessary to employ a different kind of moral epistemology and see the issue from a phenomenological perspective.

3.3. The species level from a phenomenological perspective

If integrity is taken to mean ‘whole’, ‘untouched’, and ‘complete’, then violation of integrity happens when humans interfere in the process of evolutionary adaptation (whether for the benefit of the animals or themselves). This is because an animal is considered ‘complete’ in relation to the entire environment to which it is adapted. This implies a moral duty of non-interference in the development of living beings and their communities. The integrity of a species is thus basically an expression of its freedom to develop without human intervention.

Obviously this is far from current practice, and good reasons can be given for interfering with the integrity of animals anyway. It may, for instance, be necessary to do so in order to protect other morally relevant qualities. However, it is important to note that integrity does not ‘disappear’ when violated. It is not so that domesticated or otherwise manipulated species lack integrity. It merely ‘presents’ itself in a different way than in wild, untouched, or less altered animals. Just as the dignity of a human is not taken away when tortured even though the act itself constitutes a violation of that person's dignity.

This understanding of integrity at the species level is far from unproblematic. One central issue is that humans are seen as inherently ‘unnatural’. This goes against the common experience that evolution is marked by interdependence. All species are shaped by the communities of which they are part of and the environment in which they unfold their existence. So how then is human participation in this mutual influence especially problematic? The short response to this is to point out that humans are in one sense clearly nature, but to the extent we have managed to free ourselves from nature and become culture (understood as beings who have the ability to choose how to influence the world), our influence on the development of other species has taken on an ethical dimension, compared with say how a lion's speed affects the speed of its prey and vice versa. Thus, from a phenomenological point it does make sense to speak about integrity at the species level, although it is not unproblematic.

3.4. The individual level from a phenomenological perspective

The real advantage of cutting the ties to a scientific framework of integrity seems to be on the individual level though. It is here we find most moral experiences pointing beyond welfare. And the scientific framework cannot incorporate the widely held moral experience that integrity can be violated, even when the animal does not experience any negative mental states because of it.^33,36

As an example, the term ‘integrity’ is used in this way if an animal is bred to be less capable of experiencing suffering, so as to better fit the production systems of modern agriculture, although it might be necessary to do so to ensure the welfare of the animals. ⁶⁹

According to Nussbaum, ⁷⁰ each individual animal is a complex form of life, which opens the possibility of ‘wonder-inspiring’ experiences. One example of this is the phenomenologically-based experience that the animal is ‘complete’ even before we might intervene to change the animals into something that can satisfy our needs, such as for experimental medical purposes. ‘Integrity’ can thus be interpreted as the experience that an animal is whole, complete, and finished when we encounter it. ⁵⁹ Consequently, we cannot add integrity, but only take away from it.^71,72 The integrity of the animal is thus violated when the animal is designed to serve human needs instead of being left to develop or fulfil its own species-specific goals. We have reduced it to an expression of our intentions and needs – we have de-animalized the animal. ⁶⁰ Something has been denied this animal. It is not aware of this, but it has nonetheless had its integrity violated, because it could have been allowed to come into existence, free to follow its own agenda, not ours.

If the concept of integrity is developed along these lines, as something that shows up when there is a threat of or an actual violation of it (cf. Ref. 73), it will be possible to encompass many of the moral experiences that are often placed (or even dismissed) under the heading of ‘other moral concerns’ in academic discussions of the ethics of animal biotechnology. By interpreting integrity as a basic ‘zone of untouchableness’ resulting from the very existence of the animal, the moral concerns about interfering in the lives of animals can be seen to make sense, even if they go beyond the welfare interests of the animal.

This does not mean that integrity should always be protected, but that possible violations should always be justified, e.g. in relation to animal experimentation. We propose that the notion of animal integrity and the ethical concerns that follow may be used to explain why some think that there is more to life than welfare, especially from a phenomenological perspective on the individual level.

4. Using integrity as a means for democratization in an AEC?

Our point of departure has been that public concerns are to be respected and included in agencies and boards within democratic societies. As shown above, although half the members of the Swedish AECs must be lay members in order to ensure that public concerns are taken into account,^39,41 there is often a discrepancy between this ideal and the actual assessment process. This leads to the question of how better to make use of the democratic structure to improve the project authorization process? Having presented different ways of understanding and discussing the notion of integrity we return to the two core questions in this paper: How may integrity improve the work of AECs to make discussions and decisions reflect publicly held concerns related to animal experimentation and the use of modern biotechnologies on animals; and how may the use of ideals beyond animal welfare, such as the concept of integrity outlined here, contribute to ‘democratization’ in the project authorization process?

Integrity, as seen from a phenomenological perspective on an individual level, where the pre-scientific experience of the animal and the human–animal relationship are in focus, may contribute to a wider discussion of ethically relevant issues than mere ‘welfare’, when considering what it is ethically acceptable to do with a research animal. This would make it possible to critically discuss the production of animals which are mentally or physically better suited to live in solitary containment in a lab than current mice. This view would also take seriously the moral experience that non-recovery studies (performed under anesthesia followed by killing the animal) are ethically problematic. To us this could lead to a more ‘humble’ use of research animals, as the potential benefits would generally need to be more evident than today to justify their use. Secondly, inclusion of integrity in the ethical evaluation process may facilitate a ‘democratization’ of discussion about the use of animals as it includes citizen perspectives as well as intuitions that are widespread in the ethical literature – although expressed in very different ways. This does not mean that the concept of integrity becomes the decisive factor in evaluating different uses of animals. Other morally relevant concerns might require the violation of animal integrity in certain situations. Rather, by inclusion of the concept in project authorization, discussions would better reflect the ethical concerns of society and meet the claim of respecting public interests.

By extension, we argue that there can be an obligation to protect the integrity of an animal and an obligation towards someone else that makes it necessary to violate that integrity, thus introducing an ethical dilemma. From our point of view, although ethical systems must be context-sensitive, they cannot be judged by whether they are able to present action-guiding rules that always make it possible to do what is right, good, or just. It could be that we are caught up in a tragic existence where mutually exclusive ethical obligations exist. There is not sufficient space to argue this case here, rather it should be noted that it is possible to interpret integrity as an absolute while recognizing situations that require violation of the principle. If AECs are established as a consequence of the new Directive – which ideally includes taking the intrinsic value of animals and the ‘ethical concerns of the general public’ (preamble 12) into account – these ethical deliberations could be improved by inclusion of the notion of integrity at species as well as individual levels, as seen from a phenomenological perspective. Indeed, this is also the case for existing AECs.

Moreover, the case of animal experimentation indicates a need for critical reflection on the role of scientific knowledge in ethical discussions. Reducing the discussion of animal ethics to animal welfare is to misunderstand the role of science in this regard. While science can be used to heighten aspects of the discussion, as it has been shown in the case of integrity, there are issues and concerns that cannot be explained from a scientific (in the sense of relying on empirical data) point of view, which are nevertheless relevant. ⁷⁴ Rejecting the importance or even existence of issues other than those that can be examined by scientific methodologies is simply not viable. Not only is it philosophically unfounded, ⁷⁵ it also undermines the democratic process of the AECs as it leaves out several areas of concern that AECs are explicitly required to take into account.

5. Conclusions

Animal experimentation is a practice that invokes strong feelings among both proponents and critics. Bodies performing project authorization, such as AECs, have therefore been established in many countries to weigh the importance of potential human benefit against harm caused to animals – with a view to broader societal concerns. From January 2013, all European member states are required to ensure a project authorization of research using animals. However, there are differences in how the concept of ‘ethics’ is defined and the way in which the purpose of an ethical evaluation among committee members in the individual countries is perceived^45,53 – and which hence have a bearing on the process project authorization.

Building on the EU Directive and Swedish Animal Welfare Act, we argue that an AEC or its equivalent must widen the aspects included in the decision-making process, if it is to fulfil the intent of the legislation. Furthermore, democratic methods must be embodied in the deliberation. This must be the case in order to ensure that findings are representative of the diverse views and concerns in modern society. Based on a review of how AECs work, we conclude that these diverse concerns are not reflected sufficiently and that animal welfare is not always considered with sufficient consistency or depth to balance animal suffering with potential benefits. Moreover, other ethical issues pertaining to the animals, such as animal integrity, receive even less consideration.

The question thus becomes one of how to ensure a project authorization process which is capable of mirroring some set of societal concerns about research animals and the use of biotechnologies. This is where integrity comes in. We suggest using a phenomenologically founded concept of integrity at both species and individual levels. This is a way of interpreting integrity from everyday experience, which qualifies and broadens ethical discussions within AECs in order to mirror concerns in the public sphere. Furthermore, we suggest that the concept of integrity may be employed in a way that explains why there is more to life than welfare. To us, the concept of integrity captures highly relevant ethical aspects of the human–animal relationship and of animal life per se that go beyond the concept of animal welfare, and which are often left out of AEC assessments. Although we have stated that it might not always be possible to respect animal integrity in each situation, we argue that, by interpreting integrity as a basic ‘zone of untouchableness’, moral concerns about interfering in the lives of animals can make sense and will improve the assessment process when balancing a number of ethically relevant factors. Hence, animal integrity is crucial, whether or not it goes beyond the welfare interests of the animal.