Drowning out the scientific approach to animal welfare


Switzerland has one of the most restrictive animal welfare laws in the world, stemming from the fact that since 1992, the Swiss constitution imposes the ‘protection of the dignity of animals’ on its citizens. The dignity concept has introduced a lack of consistency and comprehensibility into Swiss animal research, leading to confusion between scientific arguments and political views. As a result, animal rights activists are offered an increasing playground for obstructing science, which is illustrated by my personal story about licenses for animal research.

Since the year 2003, I have been researching the neural mechanisms of song production and learning in zebra finches at the University of Zurich and the ETH Zurich. By analogy, songbird research provides insights into human speech development and its pathologies, as well as into general mechanisms of skill learning and knowledge acquisition.

To perform an animal experiment, even just to observe an animal in the wild, researchers need a license. At my institution, the Veterinary Office of the Canton of Zurich (Vet Office) is responsible for authorizing animal experiments by consultation of the cantonal Animal Experiments Commission. This Commission, whose members are appointed by the cantonal government and confirmed by the cantonal council, evaluates applications and recommends them to the Vet Office for acceptance or rejection. Several Commission members are representatives of animal protection organizations. According to a Federal Court ruling of 2009, the Vet Office is bound by the recommendation of the Commission and may deviate from it only in justified exceptional cases. My story is about two such deviations and their opaque justifications following a moving goalpost.

In past years, my group has been researching how knowledge acquisition depends on the cues an animal learns from. By pairing animals in an environment and letting one of them learn about the auditory world via air puffs against its body while the other learns via observing the airpuff’s effects on its partner, we discovered that the animal that experienced air puffs could generalize its acquired knowledge well to a new situation whereas the observer could not. It seemed to make a large difference whether knowledge was derived from a direct or indirect source of information.

Back in 2012, these air puffs had gotten me into trouble. When we first used them as a means of behavioral training, we had applied for a license but the license had not been awarded yet, which is why after an inspection of our activities, I had to report to the cantonal police of Zurich and was fined 1850 CHF for violation of the animal welfare act. I had not been fully aware that testing this manipulation was strictly forbidden. We had been using air puffs routinely in the past to determine  the onset of anesthesia. Naively, I assumed that researchers have some freedom of exploration, which is why we had not applied for animals for fine tuning our setup. I learned my lesson and have since been working diligently to avoid any illegal activities, which I managed but was not spared of trouble.

Every three years, we need to renew our licenses. In August 2016, we submitted two renewal applications for our ongoing work. Several months later, we received an unusual long list of questions from the cantonal Commission about the prospective stress of animals. We answered the questions as well as we could and received new questions, which I was invited to clarify during a meeting. In April 2017, more than one month after my old licenses had expired and my lab had in essence been shut down, the Vet Office forwarded me yet another list of questions about stress in zebra finches, requesting me to comment on the questions. The Commission planned to forward these questions to two German Veterinarians who were working at University clinics for reptiles and birds. I requested to have the questions also forwarded to a US-based Veterinarian, who unlike the two German clinicians had direct experience with most of the inquired manipulations such as animal tethering and head fixation, two common procedures in neuroscience. My request was turned down by the Vet Office with the explanation that ‘a decisive reason for our choice was the independence of the experts from research on zebra finches’ and that ‘it is not a necessity that a reviewer has made stress measurements on zebra finches to provide a technical expertise’ (translation from German using deepL

At the same time, my group had been analyzing available data on stress from our past experiments and we were about to publish a manuscript. I asked the Vet Office whether I should forward the URL of our imminent publication to the two German evaluators or whether the Vet Office would prefer to do so. The Vet Office did not reply, so I sent the URL of our 71-page publication (1) to the two evaluators, along with a hardcopy of our work. Thereafter, the Vet Office wrote to me that ‘this correspondence (with the evaluators) is not part of their mandate and cannot be taken into consideration at this point in time’ and that ‘the evaluators have been informed’. My interpretation was that the Vet Office invited the two clinicians to disregard our publication from the mandate.

Several months later, I received the evaluation by the two clinicians, one of which is a recipient of the Felix-Wankel animal welfare research prize.  In the experts’ evaluation, I found no evidence that they had read our publication. Instead, they had a simple line of argument, which was that whatever birds do in the wild is good for their welfare, and, if they are deprived from doing the same in the lab, it is bad for them. In their evaluation, the clinicians cited the European Convention for the Protection of Animals kept for Farming Purposes as a guideline to assess welfare. In my reply, I pointed out that the relevant Convention in our context is the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes. Most importantly, I pointed out also that according to Swiss law (Tierschutzgesetz, Art 3b), one main indicator of positive welfare is given when an animal is not overtaxed in its biological adaptability (when it can adapt to a burden). My main argument was that welfare cannot be assessed simply on normative grounds as the evaluators did, but that instead, welfare is linked with the subjective experience of animals:  pleasure as in positive welfare and suffering as in negative welfare. The most important point I argued is that precise measurements are key to make informed decisions about welfare.

I was given two weeks by the Vet Office to respond, which I did. In November 2017, more than 460 days after our initial submission, I received the verdict on our pending application. The commission decided that it cannot evaluate the trade-off between the prospective gain in knowledge against the burden of animals because ‘the complete and comprehensible stress assessment for the many requested experiments is not possible.’ In the same letter, they invited me to submit ‘several smaller and clearer license applications’. In other words, after 15 months of evaluating my application, they put me back to square one.

After intense discussions with my superiors and the animal welfare officer at my University, we decided to submit two small applications to extend and conclude our research effort on zebra finch welfare, before submitting other applications focused on basic neurobiology research. In January and February 2018, we submitted two applications on stress measurements in either head-fixed or tethered animals. Our applications were framed in the context of 3R measures required by the Swiss Confederation. The 3R measures are the reduction of the number of animals (reduce), the replacement of the animal experiment by other non-animal methods (replace), and the improvement and refinement of the experimental set-up (refine). My applications were intended to further improve and refine our procedures and thus were aimed at reducing animal exposure in future experiments.

After one round of question answering, the Commission recommended at its meeting that the Vet Office approve my applications. On May 30th, the Vet Office issued the approval of my applications. On the 29th of June (one day before the end of the legal deadline), four members of the Commission filed an appeal to the Health Department of the Canton of Zurich and applied for the revocation of the licenses and non-approval of the experiments. The Canton of Zurich - as the only canton in Switzerland - assigns a right of appeal to the defeated minority of the Commission against granted licenses that have not yet become legally binding, provided that at least three members of the Commission support this appeal.  On July 19th, the Vet Office informed me that after reconsideration, my applications have been rejected with the justification that ‘refinements focused on very specific experimental situations are necessary but always refer to the least burdensome possible experiment design and cannot be used as an interest in the use of an animal experiment of its own.’ Thus, the Vet Office in effect questions that a pure 3R experiment is in accordance with the gain of scientific knowledge required by Art. 137 para. 1 of the Swiss Animal Protection Ordinance, raising the question as to whether pure 3R studies are in principle approvable at all. The Health Department charged the Vet Office for the appellants’ lawyer costs of CHF 34'641.70. In their lengthy letter, the Vet Office wrote that ‘animals (in my experiments) are subjected to experimental situations which can unquestionably be classified as very onerous’. In other words, the Vet Office responded to my 2018 applications for clarifying and reducing the burden of animals by saying that the burden is unquestionable. ‘Unquestionable’ not being a scientific criterion, I had to ask myself on what basis the Vet Office performs welfare assessments.

In summary, I note that the Vet Office and the commission interrupted my in-vivo brain research in 2016 on the grounds that the animal burden was unclear for the requested manipulations and that the weighing of interests was not feasible. But once I started to research animal welfare specifically, the Vet Office judges the burden to be unquestionable and its investigation as a research goal cannot be justified. The Vet Office and commission seek to assess welfare by prize decorated people who are independent from research and who have no experience with stress measurements themselves. And, the Vet Office instructs them to ignore published scientific evidence.

It is very difficult for an outsider to discern the higher principles governing the decision-making process in the Vet Office. To an outsider, it appears that the license approval process in cases like mine is governed by political interests rather than by interest in animal behavior and physiology. If the aim of animal protection organizations were to hinder researchers with tactics of slowing and eventually blocking research in disregard of welfare considerations, they would have been very effective in my case. Is animal welfare in Switzerland increasingly becoming a matter of politics rather than science?

I am not arguing against regulating or inspecting animal research. We are interested in the wellbeing of the animals used in our lab. We try to understand how our birds think and feel. We try to assess their wellbeing and suffering from their physiology and the behavioral choices they make. We try to refrain from projecting antropomorphic views onto them. Yet, I feel my story is only a tree in the forest of a seeming Swiss-wide political strategy of using non-scientific views on animal welfare to push against science.

The latest worrisome disregard for scientific rigor is the guideline for severity degrees, released by the Federal Food Safety and Veterinary Office on September 1st 2018. In line with past tradition, the document provides a definition of Severity Degree 3, the highest severity degree of an animal experiment: The document defines Severe constraints as procedures and actions performed on animals for experimental purposes that cause medium to long-term moderate pain or severe pain, medium to long-term moderate harm or severe harm, long-term severe fear or a severe impairment of general well-being. These criteria are very reasonable, because they are centered on an animal’s subjective suffering. However, the guide then provides rather strange severity degree examples, completely disconnected from subjective welfare. For example, under Neurology and behavioral biology, the reader finds under severity degree 3 the following procedure:  Models with chronically implanted catheters/electrodes (including wireless technology) in the skull, with additional strain (for example head fixation and water deprivation). Is it really true that a person with implanted deep-brain stimulation electrodes to treat depression is suffering from the highest severity degree when he goes to the doctor for a blood sample test (food deprivation) and an MR scan (head fixation)?

As mentioned in the outline, Swiss animal protection laws are not only concerned with welfare, but also with dignity, which relates to non-pathocentric impairments such as the visual appearance of an animal and its instrumentalization. However, it is a mistake to let dignity considerations interfere with assessment of welfare. Welfare and dignity can be opposing concepts, for example when a monkey experiences improved welfare thanks to the increase in social status deriving from its head implant and the associated attention from experimenters. To incorporate non-pathocentric (non measurable) scales into scientific definitions of welfare is an insidious effort to overrule science with politics. Dignity is a normative concept, it is what we project into animals, but animals may or may not agree.

Where is all of this leading to? Is Switzerland heading towards banning invasive animal experiments and pushing them abroad to benefit from the medical progress made there? Would this not be schizophrenic? Modern neuroscience is about 200 years old and advances rapidly. To me it seems clear that researchers will stop invasive experiments when there is no more need for them. So why let politics infiltrate scientific definitions in an attempt to outpace progress of research? There is something that matters more than what we project into animals and how we deal with them, namely, it is how we deal with people and how we justify our actions.



  1. Welfare of zebra finches used in research. Homare Yamahachi, Anja T. Zai, Ryosuke O. Tachibana, Anna E. Stepien, Diana I. Rodrigues, Sophie Cavé-Lopez, Gagan Narula, Juneseung Lee, Ziqiang Huang, Heiko Hörster, Daniel Düring, Richard H. R. Hahnloser. bioRxiv 154567; doi: