What and why is the “Evolutionary Scale” in Swiss Animal Welfare Law? Putting the Platypus in its Place

Abstract

Keywords

Animal sentience, 3R, animal model selection, evolutionary scale, partial replacement

Suggested Citation Style

Persson, Kirsten, Rodriguez Perez, Christian, Hartstang, Sebastian, Kunzmann, Peter, Hiebl, Bernhard and Shaw, David (2025). What and why is the “Evolutionary Scale” in Swiss Animal Welfare Law? Putting the Platypus in its Place. Journal of Animal Law, Ethics and One Health (LEOH), 60-73. DOI: 10.58590/leoh.2025.006

* Institute for Biomedical Ethics, University of Basel, Bernoullistrasse 28, 4056 Basel, Switzerland

** Institute for Animal Hygiene, Animal Welfare and Farm Animal Behaviour, University of Veterinary Medicine Hannover, Foundation, Bischofsholer Damm 15, 30173 Hannover, Germany

*** Care and Public Health Research Institute Maastricht University, Maastricht, Netherlands

^I Corresponding author

Inhaltsverzeichnis

• I. Introduction
• II. What an “evolutionary scale” could mean
• 1. A chronological evolutionary scale
• 2. An evolutionary distance scale
• III. The evolutionary scale in the law
• IV. How to move forward with Article 20 and the Swiss Law on model choice in animal experimentation
• 1. Sentience Scale and its limits
• 2. Case-by-case approach for the choices of animal models
• V. Conclusion
• VI. Publication bibliography

I. Introduction

For long periods in the history of Western thought, the order and structure of life and the place of human beings in the living world were considered as interconnected. When Aristotle, for example, postulated a ladder of beings in his Historia Animalium, often referred to as the scala naturae (Granger 1985), he based it on organisms’ abilities and the difference in their souls. However, it was still a long time before the evolutionary theory emerged. While both approaches elaborate on how the biological realm is organized, they are based on substantially different assumptions and explanations. Current research in evolutionary biology suggests that there is no such thing as a scala naturae in a scientific sense, as all recent living species evolved through the processes of mutation and selection and can therefore only be considered ´to be at the same evolutionary level’ (if there is such a thing). Life evolves in branches, splitting with every property that makes a species different from another, without any hierarchical order. The concept of hierarchical order is nonetheless persuasive in many contexts.

The idea of a hierarchical evolutionary scale is, for example, indirectly used by animal welfare organisations, animal rights activists, and the media, when highlighting their opposition to particular research, e.g., on dogs or primates, instead of arguing against animal experimentation per se (Primaten Zürich 2024; Herszenhorn 2023). Their argument seems to suggest that research is particularly bad because it is performed on presumably higher species.

The concept also became manifest in some legislations, for example, in the Swiss Animal Welfare Act, Article 20 (2), which regulates animal research. When choosing an animal model for an experiment in Switzerland, this article requires selecting the “lowest” species on the evolutionary scale that can achieve the purpose of the experiment. Similarly, for example, the Japanese Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan 6/1/2006) demand replacing laboratory animals “by a phylogenetically lower species” (No. 4, 1) if possible. In this article, we will focus on the case of Switzerland.

Article 20 (2) of the Swiss Animal Welfare Act reads:

“Experiments on animals higher on the evolutionary scale may only be carried out if the purpose of the experiment cannot be achieved in animal species that are lower on the evolutionary scale (…).”

The terminus “evolutionary” scale has only been included in the law in 2008, following several modifications. In earlier versions (see in the older version of the law, Article 16 (3), which can be found at https://www.fedlex.admin.ch/eli/cc/2008/414/de), a distinction was made between “higher” and “lower” animals. Legal commentaries suggest that the underlying scale is somehow related to sensory-physiological development or sentience (see Gerritsen 2024, with further references). The specification “evolutionary scale”, however, must have been introduced for a reason. Indeed, although it has not been definitively clarified which scale Article 20 (2) of the Animal Welfare Act is based on, according to the legal materials, the scala naturae already served as a foundation for the considerations regarding the parliamentary debate for the so-called “Gen-Lex” (see Official Bulletin 2002 National Council).

Thus, it would be too easy to dismiss the wording by pointing to state-of-the-art biological research, as it contains interesting insights into the foundation of the law and its possible interpretation. A closer look into the different meanings and versions can be fruitful. Accordingly, instead of a hasty rejection for scientific reasons, we want to take this reference to a biological theory seriously and investigate what could be referred to when introducing an “evolutionary scale”. From an evolutionary biology perspective, at least two ways of understanding will be discussed here: a chronological evolutionary scale, and an evolutionary distance scale.

In this paper, we would like to discuss the idea of an evolutionary scale for choosing laboratory animal models for experimental purposes in a three-step approach. In a first analytical step, we will shed light on the question of what an “evolutionary scale” could mean. Second, we will critically discuss the different potential meanings and consequences for the decision-making process, and third, present some suggestions for revising the Swiss legal framework on model selection in animal experimentation.

II. What an “evolutionary scale” could mean

1. A chronological evolutionary scale

The evolutionary theory offers an explanation for the development of the diversity of life since the beginning of time. For a basic explanation of the principles and mechanisms of evolution, see, e.g., UC Museum of Palaeontology (2020). The idea of very simple, single-celled structures as the first living beings in contrast to the highly specialised organisms surrounding us nowadays may lead to the idea that, if life evolved to be more complex over time, evolutionarily speaking younger groups differ in morally relevant ways from evolutionary speaking older groups because the latter are “less evolved”. 

Recent animal groups that appeared, for example, already in the Permian age (see figure 1) would be birds and crocodiles (archosauria), whereas mammals and especially the theria, i.e. the non-egg-laying mammals, developed rather recently in the Cretaceous age. If it were true that a younger evolutionary age went hand in hand with the appearance of some morally relevant characteristics, it would make sense to claim that research should, as a rule of thumb, rather be conducted on crocodiles or birds than on mice.

Figure 1: evolutionary tree of amniota, in: DINGUS & ROWE (1998), https://historiadelatierra.com

While on a larger scale, i.e., from the earliest life forms to recent organisms, the complexity of the individual organisms has obviously increased, it is also true that all recent taxa are well adapted to their ecological niche and survived throughout the course of evolution until now. Furthermore, not all incidents of extinction were the result of maladaptation – as it was the case with the mass extinction of dinosaurs in the Cretaceous age that was most likely caused, among other things, by a meteor (Schulte et al. 2010). Accordingly, it is not plausible to differentiate between earlier inferior and more recent superior species. In this sense, there are no “higher” and “lower” recent animals in an evolutionary sense, but only survivors in the process of selection and mutation.

2. An evolutionary distance scale

Besides the time-related explanatory power of the evolutionary theory, it provides the description of pivotal mechanisms of life – mutation and selection – leading to detectable commonalities and differences between taxa of organisms (UC Museum of Palaeontology 2020). Looking at, for example, genetic markers or phenotypic characteristics, science suggests hypothetical models of relatedness between animal (or other life forms’) groups based on evolutionary insights. Those models are called cladograms. The closer two groups are related, the less their separation in the cladogram. Two examples of cladograms of mammals are shown in figure 2 below.

For example, if humans were considered one end of the postulated evolutionary scale, we would have to look for the group “primates” here, which includes humans. Cladograms are based on hypotheses of relatedness that can be argued for – depending on the characteristics that are used as markers for relatedness. According to the cladograms presented in figure 2, for example, primates would be much closer related to bats (chiroptera) than to the animals of the opposite side of the cladogram. However, the platypus (monotremata), right on top of the tree, here, is furthest away in both cases. In fact, most – if not all – cladograms for mammals seem to agree: If we were looking for an only vaguely related mammal, we should check if the experiment can be done with the platypus. As for the chronological scale, a cladogram suggesting relational distance does not lead to a hierarchical understanding or ranking of animal groups, as it is a descriptive, structuring tool. Accordingly, it would not make sense to claim that research should rather be conducted on platypus than on mice or dogs, based on the insights of a cladogram.

Figure 2:Cladograms for mammals, in: Trends in Ecology and Evolution. 19: 430-438

III. The evolutionary scale in the law

Obviously, there is a lack of consistency between the two biological explanations of an “evolutionary scale” and actual research practice. The gold standard for animal model selection is neither a crocodile nor the platypus but rather mice, rats and zebrafish (BLV 2022). These groups do not occupy a distinguished space on the (hypothetical) evolutionary scale. Rather, there are economic, logistic and other pragmatic reasons for sticking with those established model organisms (Dietrich et al. 2020).

Against the background of this inconsistency between the scientific understanding and actual research practice we now explore the meaning of an evolutionary scale in the Swiss Animal Welfare Act. Different legal systems have different methods of interpreting the law. While writing from a bioethicist’s perspective, we do not want to neglect the different angles legal scholars generally apply when interpreting law. Broadly speaking, they range from the focus on the (subjective) will of a law, based on the legislators’ intentions, to the (objective) function of a law, based on its normative purpose, while the wording and the systematic context are also taken into account (for Switzerland, see BGer, 2C 544/2020 of 29 April 2021, consid. 3.3). There is, on the one hand, what is known as “subjective theory” (Walz 2010) providing the option of asking about the legislators’ original intention or volition behind a law. It might be difficult to determine those original arguments and intentions, depending on the law’s age and the documentation of the legal genesis. Ideally, the legislators’ will would be identical with the normative function of a role for society – but that is not always the case (ibid.). On the other hand, there is the “objective theory” of interpreting law which is more commonly used (ibid.). This interpretation is based on the meaning a typical addressee of the legal norm has to follow. A law might change its meaning over time due to, for example, technical or societal changes that the legislator could not foresee when writing the law. Applied to the paragraph on the evolutionary scale, the different ways of interpreting might lead to different understandings and, as a consequence, to different critical discussions of the Animal Welfare Act. Leaving the task of the historic exploration of the legislators’ will to legal scholars, we, as bioethicists, will focus on the functional or “objective” dimension of the law, here, and only briefly speculate about the original intention when introducing the terminology in the Swiss Animal Welfare Act. As our further research interest is in the perception and application of this paragraph by decision-makers in the animal research process, we assume for our purposes that the law must be applicable without thoroughly looking into the history of a legal paragraph.

Bearing in mind that one purpose of animal law is the protection of animal welfare (see Swiss Animal Welfare Act, Article 1), the idea of an evolutionary scale can, accordingly, be considered problematic in several ways:

Although the law presents it as a scientific term, the evolutionary scale is not a concept supported by the current scientific community, as argued in the previous section. The evolutionary theory offers explanations for the development of the diversity of life, but it does not provide any tools to rank life forms in a hierarchical way on a “scale”. The differences that can be pointed out with evolutionary explanations are on a scientific-descriptive, not on a normative level. In a way, the conceptualization of the evolutionary theory precisely led away from the idea of a scale-like order of living organisms, moving from a vertical and towards a horizontal organization of recently living beings. The evolutionary tree of life is not a fir with humans at the tip like a Christmas star pinned to the treetop, but rather an oak tree with branches growing and extending in all directions (Hug et al. 2016) or even an unrooted network (Morrison 2014).

The use of the term “evolutionary scale” in Swiss law can therefore only be considered unscientific, even if it was merely included to give the selecting process a sounder, scientific foundation. This is even more surprising as it was added to Article 20 rather recently, in 2005 (Schweizer Bundesversammlung; Knüsel; Studienkommission des Eidgenössischen Volkswirtschaftsdepartementes), as if new biological insights had resulted in a meaningful precision of the earlier paragraph. However, legal scholarship suggests that Article 20(2) is based on the legislative conviction that sentience correlates with what could be considered the “stage of evolution” (Gerritsen, 2021).

The reason for not using alternative terms like “cognitive complexity” or “sentience capacity” could be the uncertainty about those features in different groups or species, whereas referring to an ”evolutionary scale” leaves room for interpretation without a clear reference to such features. Certainly, the term “evolutionary” suggest a demand for scientific underpinning of the classification and, consequently, the choice of organisms, even though the phrasing was, as demonstrated, rather unfortunate. To put it in Marc Bekoff’s words: ”There aren’t ’lower‘ and ’higher‘ species. We make that differentiation because it serves us well and makes life easier when deciding who lives and who dies. Humans freely construct barriers that include some species and exclude others.” (Bekoff 2013, p. 17)

Even if it were true that life was hierarchically sorted from least to most complex or most developed, a morally relevant criterion for treating animals differently or preferring one species to the other in animal experimentation would still be missing.

When discussing the earlier version of the Swiss Animal Welfare Act in the 1970s, legal experts and counselors referred to biological differences between “lower” and “higher” animals, such as “low” animals only having a very basic perception of their environment, in contrast to “higher” animals who are certainly sentient and have a conscious perception of pain, harm, and suffering (see e.g. Schweizer Bundesversammlung; Knüsel; Studienkommission des Eidgenössischen Volkswirtschaftsdepartementes). Those criteria are not elaborated in the law, but they might be presented as morally relevant factors. However, the separation between “higher” and “lower” animals turns out to be superfluous if the important difference was the degree of consciousness or sentience. Additionally, the terminology “higher/lower” conceals the non-existence of a clear boundary. Which taxa do, after all, belong to the group of “higher” animals? Or is the “scale” to be understood as a continuum where each animal taxon can be ranked as comparatively “higher” or “lower”, which could be considered in a decision-making process for animal experimentation?

A closer look at the Swiss Animal Welfare Act suggests that there is a relatively clear cut between vertebrates and invertebrates. In Article 2, the law is explicitly limited to vertebrates with the option to include certain invertebrate species based on new insights regarding their sentience. That option has already been put into action by, for example, including certain crustaceans and cephalopods in the Animal Welfare Ordinance (see Article 1, Swiss Animal Welfare Ordinance). It can be assumed that the “vertebrate boundary”, which is known from other countries’ animal welfare legislation, too (see, e.g., the German Animal Welfare Act, Article 4), is rather drawn for practical than for ethical reasons. If the same laws that consider vertebrates were applied to invertebrates, human-animal interaction, and in particular animal experimentation, would become a practical and bureaucratic challenge. Including particular invertebrate taxa in the law still allows for practices like the use of insecticides in agriculture and private gardening without questioning animal welfare aspects (related to the insects), and, accordingly, permits animal research on insects without ethical scrutiny. 

However, the law provides us with a reason in Article 2 of the Animal Welfare Act, namely, that scientific findings on the sentience of invertebrates might lead to the application of the law on those groups, too. Thus, sentience is named the pivotal criterion for legal consideration, even though other legally defined concepts such as dignity, welfare, or value could have been applied on them, too, regardless of their sentience.

Here, it might make sense to set aside the idea of “scientific findings on” and focus on “sentience” in another way. Specifically referring to different layers of protection for different animals in the UK Law, where cats, for example, are classified as a “specially protected species” in the context of animal experimentation (meaning that laboratories must prove no other species is suitable before they are granted permission to use them) Hobson-West and Davies (2018) conclude that legal differences “cannot be just a question of ‘biological’ sentience. For example, a ferret might be said to have a similar capacity for suffering as a cat but is not included in this additional layer of protection. The answer [i.e., to the question why groups of animals are protected differently] relates to a particular imagining of societal concerns, where more concern is assumed for certain mammalian species.” (pp. 682-683) In a next step, they ask what such societal concerns might be based on and if and where the respective evidence could be found. They identify two sources that might be involved in establishing, what they call, “societal sentience”, which are firstly, the lay public representatives on UK animal ethics committees and, secondly, opinion polls on the use of different animal groups for medical research. That way, the public view on the acceptability of research on animal groups, the “public sensitivity” (Hobson-West and Davies 2018, p. 683), is roughly mirrored by the actual use of those groups in research. However, including consideration of the feelings of the public with regard to the relevance of certain animals in the animal welfare law, which is referred here as “societal sentience” reflects a decision to shift the purpose of the law from only protecting the animals’ welfare to also considering (human) societal wellbeing.

In our view, it is highly problematic to blur the distinction between genuinely scientific concepts like sentience and other concepts like respect to “societal sentience”. The latter suggests a criterion that can be described in scientific terms rather than a societal construct that is based on “concerns”. If the law was in any way referring to a criterion based on societal concerns it should make this explicit, rather than seeking refuge in pseudo-scientific terms.

The idea of an evolutionary scale is closely connected to the biological concept of “species”. “Species” is the unit that is (or would be) sorted on the scale and that is considered to carry the criteria for decision-making. Besides being an ambiguous and vividly discussed concept in biology (Harrison and Larson 2014) “species” might turn out inappropriate for selecting organisms for animal experimentation for at least three reasons.

First, it is pivotal – and also implemented in the Swiss Animal Welfare Legislation on animal experimentation (Article 112 Swiss Animal Welfare Ordinance) – to differentiate between different developmental stages of animals according to their (presumed) sensitivity to pain, harm, and suffering. A chicken egg or a mouse embryo during the first two thirds of their ontogenetic development are much less protected than a hatched chick or a newborn mouse. In fact, the way that the legislation on animal experimentation is phrased, those early stages are not considered as if they were animals at all. From an evolutionary perspective, however, the early developmental stages belong at the same step of the scale as fully-grown animals. Additionally, sentience can be dynamic over the course of an individual’s lifetime. In mammals, sensory experiences can influence the degree of future sentience (Snyder et al. 2015). Thus, importantly, two individuals of the same species might have a different degree of sentience, based on their past experiences.

Second, there are different breeding lines of model organisms – of the same species – for several experimental purposes. If a breeding line is, for example, genetically modified to develop certain pathological changes, the associated pain and suffering are relevant for ethical considerations. Again, from an evolutionary perspective, there is no difference on the scale between those groups of animals.

While these arguments do not prove the process of choosing species from an (imaginary) scale wrong, they demonstrate that it is at least not sufficient for the fine-grained options of animal experimental model selection processes.

Third, in the age of biotechnology, morally relevant criteria in different animal species can be modified by what has become known as animal disenhancement (Thompson 2008), i.e., the modification of properties that decrease some animal ability. Two animals of the same species might have very different capacities for welfare, sentience, or harm, if one of them was disenhanced, for example, concerning their pain perception abilities (i.e., they have a decreased ability to feel pain). The idea that morally relevant properties are strictly linked to the membership to a certain species is, thereby, fundamentally questioned for yet another reason. Accordingly, if species were the target unit of Article 20 (2) and if scientists can manipulate members of the same species to differ in morally relevant properties, a decision-making process concerning the animal model could be simultaneously legally compliant and ethically unjustified.

IV. How to move forward with Article 20 and the Swiss Law on model choice in animal experimentation

1. Sentience Scale and its limits

One way of reading Article 20 (2) would be less literal and more interpretative. What else could the legislators have had intended when formulating and reformulating this part of the Animal Welfare Act? Moreover, what could be a way of rephrasing the law so that either a) the actual and meaningful intention could be correctly understood and put into practice and/or b) a meaningful phrasing promoting animal welfare would be included? After all, it is to be expected that those who must adhere to the law consider the paragraph as it is phrased now at least slightly problematic, given that those persons are generally experts in biology and aware of the non-existence of an “evolutionary scale”.

As suggested above, it can be assumed that the legislators aimed at a gradation of properties like sentience, consciousness or sensitivity to pain. These aspects are widely implemented in other animal welfare laws, first and foremost by the EU Directive 2010/63, e.g., in Article 13, 2.b. Despite that, there are arguments to avoid such scales in legislation on animal experimentation.

It is highly anthropocentric to introduce a ladder-shaped order of animals with humans at the top, suggesting an attribution of different values to life forms according to the place on that ladder, and conveying the idea that humans might be entitled to make use of the “lower” steps of the ladder for their own purposes. If it was possible to identify characteristics that are “most developed” in humans, that is certainly true for other characteristics in other species, too. It is obvious that different groups of animals – being shaped by selection pressure – are adapted to their respective environments, including quite specific cognitive capacities.

Furthermore, the assumption that the relevant property is expressed to the greatest extent in humans can have a confusing effect on the harm-benefit analysis, which is an important tool in the decision-making process (Ethics Committee for Animal Experimentation 2022). In Switzerland, like in most states, to verify if an animal experiment should be approved, a balancing of the harms to the animals and the benefits, often to humans, must be performed. If the respective property – for example, the ability to feel pain – were deemed greater in humans than in any given nonhuman animal, a potential to decrease human pain would, in most cases, outweigh the presumed animal pain of an experiment, leading the weighing process ad absurdum. (This is in addition to the typical pro-human bias of harm-benefit analysis in general, where it is generally assumed that animal pain is justified if it meets some human end.)

The Swiss Animal Welfare Act has previously faced criticism for Article 20 (2), along with a suggestion to interpret “lower at the evolutionary scale” as “less sentient” or “less subject to strain as a result of the proposed experiment” (Ethics Committee for Animal Experimentation 2022). If, indeed, it is time to revise this part of the animal welfare legislation in Switzerland, though, there is the opportunity for a substantial and meaningful change.

Sentience or the ability to experience pain, harm, and distress are criteria frequently used in other animal welfare laws. However, these alternatives present challenges, as well:

• 1. The problem of accessibility of animal sentience remains, especially for those types of animals that would have been placed “lower on the evolutionary scale”. For some species that are commonly used in laboratories like mice, grimace scales to assess pain do exist, but for other common laboratory animals like zebrafish, there is no such tool – most likely because fish do not express pain facially (Sneddon 2011; Gibbons et al. 2022). However, there is already promising research on, e.g., invertebrate nociception and pain (Gibbons et al. 2022) providing more detailed information for future decision-making processes.
• 2. Besides pain, there are other criteria contributing to animal welfare, such as the status of being captive, being bored, being handled by humans, being kept in isolation, not having opportunities to perform the whole behavioural spectrum, a lack of choices of partners and roommates etc. It is unclear whether suffering from these factors is expressed in animal taxa proportionally to their ability of feeling pain, and the weighing of these factors would also require legal regulation.
• 3. Even if there was a sentience scale, choosing the presumably less sentient species would not always cause less harm. When it comes to distress due to handling, for example, mice – being prey animals – are much more at risk of suffering than pigs or dogs who enjoy (positive) interactions with humans. At the same time, stress due to handling is already tackled in laboratory practice by improved handling practices and corresponding staff training (3RC 2025). The underlying problem seems to be acknowledged even if it is not clearly stated in the law.

As long as the main idea is based on a scale, it is most likely rooted in the anthropo-relational conception that the relevant criterion (e.g., being sentient, being in distress, dignity, etc.) is most fully expressed in humans and to a somewhat lesser extent in other creatures.

For example, according to the UK Animal Welfare Law (ASPA), studies must “involve animals with the lowest capacity to experience pain, suffering, distress or lasting harm” (Part 3, 181 (b)). Here, the criterion is not linked to species membership, but the phrasing still implies an inherent property such as the capacity to experience pain.

2. Case-by-case approach for the choices of animal models

Given that constraints on animal research can be based on, e.g., economic, logistic, and scientific aspects (Dietrich et al. 2020), ethical criteria may play only a subordinate role in the selection of animal models for research. Dietrich et al. (2020) list them as one of twenty (groups of) selection criteria among, for example, availability of methods and techniques, institutional support, financial constraints, and translational potential. Thus, if we as bioethicists want ethical criteria to be meaningfully considered by decision-makers in animal research, and if their implementation relies on animal welfare legislation, these criteria must be formulated in a clear, pragmatic, and plausible way while also being easily applicable within the context of animal experimentation. We set out above why Article 20 (2) of the Swiss Animal Welfare Act is not a good example, neither in this sense nor from a theoretical perspective, and, more generally, why an anthropo-relational scale might not be an appropriate way of ethically steering the animal model selection process. 

In contrast, the German Animal Welfare Act demands that experiments on “animals whose species-specific capacity to suffer from the effects of the experiment is more developed may only be carried out if animals whose such capacity is less developed are not sufficient for the purpose pursued” (Part 5, 7(a), (2), translated from German by the authors). The difference between the two laws might seem subtle, here, but in the German version, the small addition “from the effects of the experiment” takes the criterion from a general to a specific consideration, namely, the precise case of the planned experiment. A similar focus on the planned experiment has already been recommended in the Swiss context. As said above, the Swiss Ethics Committee for Animal Experimentation (2022) suggested that “lower on the evolutionary scale” could be taken to mean “less subject to strain as a result of the proposed experiment”.

If taken seriously, this wording could lead to consideration of case-specific conditions that make a certain group of animals – not generally but within the context of a specific study – more susceptible to harm and prompting careful avoidance of those animal groups for the particular project. In accordance with this, the effects of the experiment would have to be carved out and considered thoroughly, leading away from the animals’ species-specific capabilities like “sentience”, “consciousness”, or “sensitivity to pain” to more detailed questions about the respective animals. A meaningful question, for example, might be, “Does this animal suffer if kept in isolation for more than 24 hours?”. Mice and rats, the animal species used for most animal research worldwide (Mikulic 2023), are social animals and suffer from being kept in isolation. As an example, solitary housing of mice and rats is considered as a mild harm (i.e., degree of severity 1) under Swiss law (Federal Food Safety and Veterinary Office 2022). From an ethical point of view, it would make sense to switch to solitary animals – like hamsters – instead, if possible, for those studies that require isolation. However, according to many of the 20 criteria listed by Dietrich et al. (2020), hamsters might not be a good choice. Compared to mice, for example, far fewer markers, antibodies, and tools may be available for hamsters, making the results less comparable to other research (on mice) and posing challenges if laboratory staff are not trained to work with hamsters. It is crucial to examine the significance of ethical argumentation in animal model selection, particularly in relation to more pragmatic or science-oriented criteria.

Further questions that point out the ethical dimension of the selection process in line with a case-by-case approach might be: “Does this animal prefer a type of food that is in conflict with the experimental design?”, “Will the animal be wet (and cold) for some time?”, “Is it necessary to work with embryos and, if so, will the mother be affected in that process?”, “Is this animal usually active at dawn but the experiment needs to take place during daylight time?”, and “Is the life expectancy of this animal 15 years but the experiment demands killing her or him at 2 months old?”. This particular aspect – the varying sacrifice of animal life years – is particularly neglected in the debate on animal harm in research. However, integrating both the quality and quantity of animal life years affected by the choice of the animal model could become an important factor in the weighing process, especially considering that animals with considerably longer lifespans than mice or rats are occasionally used (Mikulic 2024), such as dogs, different species of birds or ferrets. A way of taking the affected animal life years into the evaluation process has been suggested elsewhere (Persson and Hartstang 2024).

Beyond species selection, there may be a reciprocal process of adapting both the research process and the animal model selection if there are strong (non-ethical) arguments favouring the choice of a certain animal model that would be harmed more than another by the effects of the given experiment. Finally, it is worth noting that the German animal welfare law merely refers to the capacity “to suffer from the effects”. In this context, as seen elsewhere in Swiss and German legislation, it would be reasonable to add "to be harmed" to account for aspects such as the loss of life years, as pointed out above.

V. Conclusion

In this article, we demonstrated why the evolutionary scale is not a useful tool for animal model choice for several reasons: a) there is no clear scientific meaning of “evolutionary scale”; b) there is no morally relevant criterion linked to what could be understood as evolutionary scale; and c) evolutionary criteria are limited to the level “species” and do not allow for other morally relevant differences between animals such as their developmental stage. Additionally, we argue that the use of a sentience scale instead might not serve the purpose (of providing criteria for a decision making process guided by ethical criteria), either, due to: a) a lack of accessibility of animal pain; b) the importance of other welfare-related factors such as boredom and captivity; c) the importance of specifics of the experiment that might not be linked to the animal’s degree of sentience, such as the stress from being handled by humans. Accordingly, we suggest a case-based approach to the decision-making process.

If the selection of an animal model for an experiment should align with the 3R principle and its pathocentric foundations by contributing to harm reduction in animal research, the challenge remains to define case-based decision-making criteria. These criteria must assess the potential harm of the entire procedure – including training, housing, the experiment itself, and the animal’s fate after the experiment – across different animal groups.

However, the fact that animal models are often chosen for pragmatic reasons, such as the availability of markers, tools and techniques for procedures, the experience of staff members and equipment that is established and available at a given institution, or incentives to use animal models considered “standard” for comparability, replicability, or publication purposes highlights potential conflicts in the application of that law. However, the shortcomings in implementing an animal welfare-based legal norm cannot be solved by modifying the law to accommodate more pragmatic arguments. Rather, the focus should be on identifying and overcoming the barriers that hinder the effective application of ethical criteria.

Furthermore, the blurry concept of “societal sentience” should remain open for critical discussion, acknowledging that the public perception of “sentience” goes beyond scientific evidence. If the law reflects societal values, it may also account for different concerns for certain animal groups. However, it should provide – or at least be grounded in – rational arguments for treating creatures differently without morally relevant different properties.

Critically examining the potential impact of changing the current wording of the Swiss Animal Welfare Act implies that the law, in its current form, is already implemented into the scientific process of animal model selection. However, empirical data on this question are still lacking. It is possible that the above-mentioned practical criteria play a more important role than ethical criteria like those outlined in this article. This argument leads to a broader discussion on the potential impact of the law – and potential amendments – on scientific practice, a topic that extends beyond the scope of this article.

The Swiss Animal Welfare Act should incorporate the diverse arguments presented here and strive to become a true pioneer in laboratory animal welfare legislation.

Acknowledgements

We are very grateful to Nicole Lüthi who provided us with material and insights concerning the history of the evolutionary scale in the Swiss Animal Welfare Act!

Copyright (c) 2025 Kirsten Persson, Christian Rodriguez Perez, Sebastian Hartstang, Peter Kunzmann, Bernhard Hiebl, David Shaw

This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.