Improvising care: Managing experimental animals at a Japanese laboratory

Caring for laboratory animals in Japan

In the contemporary life sciences, laboratory animals are central to how scientists understand disease and human corporality. Because of the rapid progress of genetic modification and gene mapping in recent years, various types of model animals – in particular, rodents – are becoming seen as even more necessary for scientific knowledge-making (Crabbe et al., 1999). Not only are the numbers of genetically altered animals expanding in biomedical research, but new methods for breeding, rearing, testing and evaluating animal bodies and behaviours raise continuous questions about what humane treatment entails. At the same time, growing societal concern with animal welfare has led to demands for universal principles to govern research ethics, as well as more accountability and tighter controls over the bureaucratic management of institutions conducting animal experiments (Davies, 2012; Nelson, 2016). Formal frameworks for regulating laboratory animal welfare have been in place in many Western countries for some time, with some dating from as recently as the 1960s (e.g. the US), and others dating to the 19th century (e.g. the UK). The best known of these frameworks is the 3Rs (Replacement, Reduction and Refinement), which was first proposed in 1959 in the UK and has since become the global standard. This framework, which is based in a cost-benefit analysis, prioritizes the replacement of animals with alternative mechanisms, the reduction of the number of animals used and the refinement of experimental procedures to minimize suffering and improve animal welfare.

This situation asks scholars, as well as scientists and technicians, to consider what constitutes good care for laboratory animals. Scholars in science and technology studies (STS) have explored how animal care in the life sciences today is situated and shaped through international norms, in and across different laboratory, technical, regulatory and affective environments. Most of these studies on animal ethics have been based in North American and European contexts (Davies, 2012; Davies et al., 2018; Greenhough and Roe, 2018; Holmberg, 2008; Lynch, 1988; Nelson, 2016; Svendsen and Koch, 2013), where regulatory systems governing animal testing are relatively well-established and consistent. ¹ Inspired by these studies, I delve into how scientists and technicians deal with the ethical ramifications of caring for but also using, exploiting and killing animals in an ethnographic study of a stem cell laboratory in Japan.

Since the early 2000s, the Japanese scientific community has been integrating global principles such as the 3Rs into law and adopting regulatory systems for animal experiments that are similar to ones in Euro-American countries. However, the integration remains partial and inconsistent. Japanese universities and research institutions have long engaged in other ways of coping with the ethical tensions surrounding animal experimentation, most notably by conducting annual memorial ceremonies for animals killed in the process of research, known as kuyou. Today, these mortuary rites exist alongside other regulatory frameworks, but not always comfortably. The ethics governing animal experimentation in Japan have developed along these parallel trajectories, in part because of religious and cosmological traditions that existed in Japan before the import of Western science, namely Shinto, an animistic tradition indigenous to Japan, as well as influences from Buddhism and neo-Confucianism, which spread from China and other Asian countries. While there were existing empirical traditions of studying nature in Japan, new scientific methods and practices imported from Europe and America during the late 19th century resulted in enduring ethical tensions between ways of knowing (Marcon, 2015). Not surprisingly, these clashes between various norms and values within scientific fields make it difficult to regulate animal experimentation.

In what follows, I look at why this situation has emerged and explore how global standards bump up against ethical understandings of human-animal relationality as practiced in a particular Japanese laboratory. I argue that the inconsistencies of the regulatory system, which are evident in the idiosyncratic ways that Japanese scientists and technicians care for laboratory animals, end up putting more of a burden on individual scientists and technicians to decide what is ethical. In addition to focusing on these tensions, I ask how care for experimental animals is configured materially, through the use of particular technologies, instruments and texts (Mol et al., 2010). Inspired by the framework of ‘ontological choreography’ proposed by Cussins (1996), I shed light on how the coordination of animal care breaks down, as well as the process of how it is reorganized, taking new forms as care moves from the personal to the regulatory level and back again. I pay attention to various kinds of ‘improvisations’ practiced by scientists, technicians and animal ethicists to fill in the gaps between different logics and practices of care. Ultimately, I argue that these are not only tensions between Euro-American and Japanese cultures of care, but between individual people who enact their own practices of care in the course of daily, embodied relationships with specific animals.

I focus on the practices surrounding the care of experimental animals at the Biotechnology Center Japan (BCJ), a public science research centre located in western Japan. (Throughout this article, I use pseudonyms for institutions and individuals.) Since being built in the early 2000s, the BCJ has become known as one of the premier life science research centres in the country. In its efforts to become a ‘global’ research centre, the institute frequently invites foreign scientists to collaborate on projects and officially uses both English and Japanese, which is rare in the Japanese scientific community. Within this centre, I focus specifically on the Tanaka laboratory, one of the largest, most generously funded labs in Japan. This laboratory, which is made up of medical scientists, geneticists, developmental biologists, stem cell scientists and technicians, mainly works on creating new treatments for diseases of the nervous system. Their aim is to connect scientific knowledge with clinical expertise to quickly develop and bring new cell therapies to market.

During my fieldwork in this high-pressure work environment, I found scientists and technicians grappled with questions about how to best care for laboratory animals. Although they relied on these animals to ensure the safety and efficiency of medical treatments, they were also faced with calls to reduce the numbers of animals they used; indeed, the body counts of animals used in Japanese laboratories are relatively high, compared with counts elsewhere. According to one official document, Japanese scientists kill 10 million experimental animals every year. ² Of course, this does not mean that Japanese experimenters do not care about laboratory animals. Rather, my research suggests that they do care, but in ways that differ from the normative framework of animal welfare. Under the loose and eclectic regulations that govern animal testing in Japan, I explore several specific cases of scientists and technicians improvising to deal with the discomforts that come with caring for their nonhuman subjects. By paying attention to how members of this laboratory question the nature of good care for laboratory animals, this case provides an opening for rethinking the cost-benefit logic that underlies global discourses around laboratory animal ethics.

Care, ethics and tinkering in studies of laboratory animals

Since the 1980s, STS scholars have been interested in the relationships between scientists and experimental animals as a topic within laboratory studies. Early on, these studies tended to focus on the justifications made for using animals for experiments. For instance, in his pioneering studies of human-animal relations in a neuroscience laboratory, Lynch (1988) analysed the processes through which living animals are turned into scientific data. He was primarily focused on the material and symbolic transformation of animals into abstractions, highlighting the moment of ‘sacrifice’ when the animal was killed for the sake of science. For my purposes here, it’s more relevant to focus on the intersection between these studies of laboratory animals and feminist STS studies on care, as well as critical explorations of the regulations and principles governing animal research.

Recently, feminist STS scholars have been exploring the bodily, affective and ethical aspects of laboratory experiments, showing that the relationships between animals and experimenters are much more complicated than those described by Lynch (1988) (Greenhough and Roe, 2011; Holmberg, 2008). Such work analyses how mutual bodily and emotional responses emerge between humans and animals, generating particular spaces of care. These feminist STS scholars have also paid attention to the seemingly inconsistent aims that motivate scientific research practices – such as the contradictions between caring for and killing animals, or between protecting animal welfare and producing good scientific data – and how these tensions are negotiated. These studies are part of a broader discussion of the concept of care in technoscience. According to an introduction to a special issue on care in this journal, ‘care is a selective mode of attention: it circumscribes and cherishes some things, lives or phenomena as its objects’, while excluding others (Martin et al., 2015: 3). Similarly, de la Bellacasa (2011) argues that to think about care does not imply a straightforward commitment. She suggests that we consider not only ‘for whom’ we care, but also ‘“Who cares?” “What for?” “Why do ‘we’ care?” and mostly, “How to care?”’ (p. 96, emphasis in original). In short, care includes ethical values – such as questions about what is good care for whom – and sometimes those values create tensions (Mol, 2008).

Therefore, the question of how to care concerns not only the relationship between laboratory animals and the people who handle them directly, as early laboratory studies explored. Rather, recent interventions have emphasized that it is important to ask how care for these animals is situated and shaped by elements outside of the laboratory as well. Thus, when I look at how scientific knowledge-making is entangled with care relations in Japan, my scope is not only focused on a specific laboratory at the BCJ but on broader ethical contexts, including institutional and organizational settings.

In order to do that, I draw on but also depart from the work of other scholars who have tackled the relationship between regulations and practices in animal experimentation. This work asks about the standards of practice that govern how technicians and scientists care for animals, such as the aforementioned 3Rs, within mainly European and American settings (Davies, 2012; Davies et al., 2018; Greenhough and Roe, 2018). Historians of science Druglitrø and Kirk (2014) have shown that in the 1950s and early 1960s various European and North American countries developed national regulatory, bureaucratic and management systems for producing and provisioning laboratory animals, while negotiating with international networks and infrastructures. In this dynamic and mutual process, they argue, transnational rules and techniques were imagined and solidified into law, even though many local differences have remained.

Greenhough and Roe (2018) take up this issue by looking closely at the role of animal technicians who carry out day-to-day animal care in the UK, where the principle of the 3Rs was formalized into law through the Animals (Scientific Procedures) Act 1986. According to Greenhough and Roe, what animal technicians can and should do is clearly defined by the licensing requirements of the Home Office: they must ‘act at all the times in a manner that is consistent with the principles of replacement, reduction and refinement’ (cited in Greenhough and Roe, 2018: 703). At the same time, they argue, slightly ambiguous descriptions in the guidelines make space for animal technicians ‘to not only meet minimum requirements but to strive to exceed them’ (Greenhough and Roe, 2018: 705). Borrowing the concept of ‘practical tinkering’ from the work of Mol and collaborators (Mol et al., 2010: 13), Greenhough and Roe illustrate how animal technicians address the needs of animals by using their intuition, without changing the protocols of scientific experiments. While other STS scholars have emphasized the importance of practices of care as a contrast to universal principles, or implied that rules could erode such practices (Mol et al., 2010), Greenhough and Roe argue that individual practices are able to make up for what the principles cannot cover.

Their understanding of tinkering ³ in relation to a wider system is reminiscent of the famous discussion of classification and standardization in the work of Bowker and Star (2000). They ask, ‘What work do classifications and standards do?’ (p. 9). In particular, they pay attention to cases that do not fit easily into given categories. Bowker and Star (2000) explain that imposed standards cannot account for every local contingency, which produces ways of ‘working around’ (p. 320) formal systems. Similarly, Greenhough and Roe use the term ‘tinkering’ ⁴ to describe how animal technicians adjust to regulations governing animal care. In their argument, people invent how to care for animals on the spot, in order to fit with local situations. But it is the imposed rules that produce tinkering. This understanding of tinkering corresponds well to situations with fairly rigid international standards of practice, such as the UK’s standards.

Law’s (2010) work on the ‘choreography of care’ is also helpful here. Inspired by Cussins’s (1996) concept of ‘ontological choreography’, Law sheds light on the relationships between care practices in laboratories (or, in his case, a farm) and broader ethical contexts, by drawing attention to ‘the intricate organization that goes into the routines of practice’ (p. 67). He describes the care practices of a veterinarian who worked a dairy farm during the 2001 foot and mouth epidemic in the UK. A multiplicity of forms of care emerged during this outbreak, which produced tensions as veterinarians tried to care at once for animals, farmers, neighbours and the nation. Law also puts emphasis on the importance of tinkering, although he uses the term a bit differently from Greenhough and Roe. ⁵ He considers multiple elements of the veterinarian’s care, taking as an example a moment when the vet allows a calf who will be killed soon to be talked to and touched. Law writes, ‘And this is the nature of veterinary care too: it can be understood as an improvised and experimental choreography for holding together and holding apart different and relatively non-coherent versions of care, their objects and their subjectivities’ (Law, 2010: 69). Contrary to Greenhough and Roe, who discuss tinkering as complementary to rules, Law presents tinkering as filling in gaps between heterogenous assemblages.

Law looks at care as ‘multiple’ in Mol’s sense of the term: Care is different depending on the space and time, but it remains mostly arranged – or ‘choreographed’. In his case, how a veterinarian behaves during an epidemic is mostly defined by national policy and his own professional norms; under a relatively consistent organization, the veterinarian cares for animals, farmers and himself. In that sense, both Law and Greenhough and Roe analyse relatively consistent and static cases. After all, ‘tinkering’ originally meant ‘to mend’ or ‘to repair’ in a makeshift way, and thus to return something to its original state. Any actions taken to care for animals are already defined by the original framework of rules and regulations. Therefore, the animal technicians and the vet can respond through what Greenhough and Roe call acts of ‘pragmatic refinement’, without changing the given choreography radically.

Chains of improvisation in a loose choreography of care

While useful, the concept of tinkering does not fully capture the type of activity going on in a space with few institutionalized standards, like Japan. In the case of the BCJ, the lack of consistent regulations for animal experimentation led to more elaborate improvisations than ‘tinkering’ suggests, and for this reason I turn to the idea of ‘choreography’ introduced by Cussins (1996). Cussins focuses on ‘objectification’ and ‘selfhood’ in an infertility clinic, arguing that patients’ selfhood is choreographed by technologies, gametes and interactions with doctors. Drawing on moments when a would-be mother encounters practitioners, procedures and instruments, Cussins (1996) considers the coordination and organization of different kinds of ontological choreography. She (p. 604) writes: ‘I use the word “choreography” in the title of this paper, and as the dominant ontological/political metaphor throughout, to invoke materiality, structural constraint, performativity, discipline, co-dependence of setting and performers and movement’. The concept and the metaphor of choreography points to a much more dynamic process than mere tinkering, one in which different elements of care encounter and sometimes crash into one another. This image of choreography helps to articulate the situation at the BCJ, where scientists and technicians acted in ways that fit with their particular situation, rather than to complement established rules and regulations. Their improvisations ⁶ produced more improvisations, or what I’m calling ‘chains of improvisation’.

Here, thinking about what goes into an actual dance performance is helpful for considering the conceptual contrast between tinkering and chains of improvisation. According to the anthropologist of music and dance Fukuoka (2017: personal communication), the extent to which choreographers define movements and positioning in space before a performance depends on the type of dance and the stage. For instance, for modern ballet, a choreographer often decides most of the floor patterns and the movement of dancers, but sometimes they leave the dancers to figure out some of the more subtle movements on their own. Structures produce small spaces for dancers’ improvisation. Hence, acts of tinkering within set structures, as Law and Greenhough and Roe describe, resemble acts within modern ballet.

On the other hand, chains of improvisation are closer to a form of dance called contact improvisation. It is performed as a duet; dancers support each other’s weight while in motion. Although there are some basic forms that dancers need to learn in the training process, these dances do not have a choreographer. Rather, as the anthropologist of dance Novack (1990: 8) explains, these dancers ‘improvise their movement, inventing or choosing it at the moment of performance’. Novack (1990) describes how dancers improvise the flow of their movements in concert with a ‘partner’s weight, rolling, suspending, lurching together (p. 8). In other words, one movement creates the next one. The sequences and the positions are situational, allowing for spontaneity and open-endedness.

Like contact improvisers, the scientists, technicians and staff at the BCJ were continuously and dynamically changing their movements through encounters. As we will see, these alignments easily collapsed and were rearranged, but they also remained linked together. To focus on these linked encounters, I refer to ‘chains’ of improvisation. In addition, at the BCJ, people brought steps and moves that they had learned in other contexts of their lives to the choreography of care at the centre. In order to better understand what generated some of these sequences, in the next section I provide a brief overview of the history of animal welfare in Japan, focusing on the invention of kuyou and the development of national regulations governing animal experimentation.

Encounters between local cosmologies and Western science

Scholars who discuss Japanese understandings of nature usually focus on the influence of Buddhist and Shinto cosmologies, which emphasize continuities between humans and non-humans. In Japan, animals traditionally have been regarded as threatening spiritual forces but also as resources for human livelihoods. These ways of understanding animals have changed continually throughout history, but one of the most significant transitions, for our purposes, was when Western science and medicine, as well as Western understandings of ‘nature’ and ‘animals’, were imported at the end of the Edo period (1603-1868) and the start of the Meiji period (1868-1912). This period saw the end of the feudal military government known as the Tokugawa shogunate, which ruled Japan for centuries until being overthrown by supporters of the Imperial Court. The Meiji period is often summarized as one of rapid upheaval and ‘modernization’, after more than 200 years of a policy of national seclusion that strictly limited Japanese encounters with the rest of the world.

Before Meiji intellectuals translated the word ‘animal’ from the West as doubutsu, Japanese people had used other concepts and terms to describe and categorize the birds, worms, insects, beasts, humans and sometimes spirit beings that they encountered in everyday life. Most of these terms were derived from Buddhist and neo-Confucian vocabularies (Ambros, 2012). During the early Edo period, these Buddhist and neo-Confucian ethics were brought into the regulatory and political realm. Most notably, in the late 1600s, the fifth shogun, Tokugawa Tsunayoshi, became known as the Dog Shogun when he established a law prohibiting cruelty to animals, and dogs in particular. His efforts followed Buddhist understandings that protecting animals would enable him to build good karma, thus fulfilling his wish to prolong his own life (Ambros, 2012).

Historically, Japanese ways of understanding bodies had been heavily influenced by Chinese medicine, and it was not common to dissect and use animals for scientific and medical knowledge-making (Kuriyama, 1999). ⁷ However, even during the period of national seclusion, the Edo shogunate allowed limited trade relations with the Dutch, and some Western medical books were imported during this period to Japan; medical doctors were becoming interested in anatomy and in different understandings of the body. At the end of the Edo period, some Japanese bacteriologists went to German universities and studied hygiene as part of the government’s efforts to modernize. When they returned to Japan, they brought with them Western scientific knowledge and methods, including animal experiments.

According to historians, it was also around this same time that mortuary rites for animals began to be developed. The oldest Japanese documents referencing animal memorial services are from 1917 (Okada, 2010); they describe a service held by Buddhist medical scientists at Kyushu University. Okada argues, ‘When the method of animal experiments was imported with modern science to Japan, the gap between modern science and a Japanese view, according to which the quality of human life and animal life are essentially equal, led medical students to invent memorial services for animals’ (p. 172, translated by the author). During a kuyou, scientists and technicians honor the lives of the deceased by putting their hands together in a prayer position, thanking the animals and making offerings of flowers and the animals’ favourite foods, such as bananas and oranges. It is thought that by providing designated times and spaces for consoling, comforting and honoring the souls of the dead, Japanese scientists and technicians care for these animals even after they have been killed.

Today, almost every university and institute holds these annual mortuary rites for the animals they kill in the process of research. It was not until the 1970s, however, that practices such as holding kuyou as an annual event and erecting memorial statues on campuses were widely institutionalized. At that time, there were no regulations or laws in modern Japan that mentioned laboratory animals directly. Instead, there was a general law for animals (including pets), called the Law for the Humane Treatment and Management of Animals, which was established in 1973. This law declared that researchers needed to care for experimental animals, but its scope was very limited, which drew criticism from animal welfare advocates outside of Japan. Ceremonial practices honoring the lives of laboratory animals started to spread to universities and institutions across Japan in large part due to this growing concern with animal welfare globally. Hence, the increasing popularity of animal memorial rites throughout the 20th century should be understood not as the continuance or reemergence of an ancient tradition, as it is often presented, but as a response to shifting scientific practices, which viewed animals primarily as resources for knowledge-making.

The widespread practice of kuyou did not provide a full resolution to these tensions, however. The increasing commodification of animals and the traditional ideal of harmonious relations between humans and animals encountered one another, but they were not easily translated into a unified choreography of care for experimental animals. By the early 2000s, the state was forced to acknowledge this problem. In 2006, due to increasing pressure and criticism from foreign countries, and the difficulties faced by Japanese pharmaceutical companies trying to expand overseas, the international guiding principles of the 3Rs were specified in Japanese law for the first time. Three different government ministries, ⁸ as well as the Science Council of Japan, each created some basic policies regarding the treatment of laboratory animals, but they did not stipulate any penalties for failing to follow them. Basically, these were independent guidelines that lacked unity (Kagiyama, 2010; Ooue et al., 2008). Around the same time, other institutions took it upon themselves to create a ‘self- governance system’. According to Kagiyama and Mizushima (2013), the legal framework of self-governance now in place in Japan entails that each research institute, university and pharmaceutical company is obligated to organize an Animal Care and Use Committee, which governs animal experiments based on its own ethical rules.

Around this time, there was also a movement to make a Japanese version of the guidelines concerning the use of animals in science. However, Japanese scientists have tended to strongly emphasize the importance of innovation in science and technology in order to compete with Euro-America countries (Kagiyama, 2009). Since the Meiji period, and especially in the period after World War II, the pressure to ‘catch up’ with Europe and North America to become a modernized nation has been intense (Mizuno, 2008). Even now, though Japan is known as a country of advanced science and technology, this discourse is still very strong. In this context, scientists have expressed concerns that strict laws and regulations might delay the development of science and medicine. These hesitations are no doubt related to the fact that the animal rights movement in Japan is not as active as in Euro-America, and public concern for this issue remains quite low. ⁹

In sum, unlike in Europe and North America, it would seem that animal experiments historically have not been an explicit ‘matter of concern’ (Latour, 2005) in Japan. ¹⁰ However, when we look more closely, we see that the situation is more complicated. There has long been concern about animal welfare and the status of laboratory animals in Japan, but the specific reasons why people have been concerned – and the ways they have expressed that concern – are quite different from the approaches that emerged in Euro-American countries. Reading the history above through the lens of chains of improvisation, laboratory animal care in Japan continues to be choreographed by heterogenous elements, such as encounters with Western science, Buddhist ethics, criticism from overseas, inner requirements to meet global standards and strong objections by scientists towards strict regulations. To reconcile these tensions, new dances were choreographed, such as the creation and institutionalization of kuyou and efforts to integrate global standards in regulations. As I will show in the following sections, Japan’s loose regulations have provided unchoreographed spaces for practitioners to continue improvising how they care for animals.

Improvisation 1: Self-governance at the BCJ

The first example of improvisation that I would like to highlight is the complicated dual system of self-governance at the animal facility at the BCJ. The first system, which I was told was borrowed from the US, was based on the consistent documentation of practices or creating forms that were to be submitted to the Animal Care and Use Committee at the BCJ. To get approval for a research study, scientists were required to first submit a document of several pages that described their goals, their experimental procedures and the number of animals they would use. In addition, the committee asked scientists to include a ‘pain category’, a concept that was imported from the Scientists Center for Animal Welfare in the US. For each proposed experiment, scientists had to write down the ‘potential pain severity of each experimental procedure’, choosing from a list that classified various procedures and pain categories in detail. This requirement was based on the logic of harm-benefit analysis, which entailed that scientists weigh the significance of the experiment (benefit) against the pain it would cause to the animal (harm).

In an interview, a veterinarian I will call Hayashi, one of the members of the Animal Care and Use Committee, told me: ‘Animal experiments certainly involve animals’ pain. Of course, often the animals’ pain and science can be in conflict. If their documents are not sufficient, ‘we require them to submit their application again’. She suggested that, through the process of writing about the pain of animals, it was possible to ‘raise responsible scientists’. Hayashi repeatedly stressed that it was the task of the individual scientist to decide what was ethical.

The other system that the institute had adopted was called a ‘Japanese care system’. At most Japanese universities, scientists must take care of the rats and mice they use in their experiments by themselves, because they do not have a centralized system and lack the funding to hire designated staff to take care of the animals. However, the BCJ was unusual because they did have specialists on staff, called ‘caregivers’, whose role was to clean the cages and feed all of the rodents. Every day, they collected and washed the cages of the animals and brought them food. However, the care they provided was limited; their role was just to provide the minimum environmental and dietary conditions to allow the animals to live without hunger or disease. According to the core principles of the institute, breeding, training, experimenting on and killing animals were jobs for scientists. Kato, the manager of the caregivers, told me:

I’ve heard that in the UK, technicians in animal facilities care for animals and scientists order mice or rats through computers. I don’t think it is a good system. Maybe my way of thinking is a bit traditional, but scientists need to care for the animals as ikimono (living beings). Animals are not mono (non-living beings). They are ikimono. Scientists do not know what is ikimono until they handle and care for them with their own hands. It is only scientists who can say ‘thank you’ to the animals.

Both Hayashi and Kato emphasized the responsibility of scientists as direct caregivers. In addition to that, through the writing of documents, scientists were required to think about the animals’ pain and suffering and to develop their scientific experiments reflectively. By repeating those processes, scientists were supposed to be able to develop an autonomous sense of ethical responsibility.

As Hayashi explained to me further in the context of her own work as an animal ethicist, these two systems were an attempt to reconcile the differences between what she referred to as ‘Japanese customs and science-based animal welfare’. By Japanese customs, she gestured towards ‘Buddhism-ish’ or ‘Shinto-animism-ish’ feelings and practices. Referring to the Buddhist concept of reincarnation, she explained to me, ‘Japanese people think that we are humans by chance; in our next life, however, we might be flies. So we are taught to be kind to small lives, such as insects’. She told me that when she attends international conferences on animal welfare, she sometimes tries to explain to others how Japanese people see and feel animals. However, such ‘customs’ do not always align with the dominant ‘logical’ and science-oriented approach to animals held by other experts. Hayashi’s insights lead to an important point: While the domains of animal welfare and science are often seen to be at odds, they also share an understanding of humans as fundamentally separate from other animals, which does not always align with Japanese cosmologies of life and death.

The Japanese care system at the BCJ was based on an understanding that the lives of all ikimono were interconnected. According to this view, there is no hierarchy between humans and other ikimono; they are interconnected corporeally and spiritually, whether dead or alive. This view of life is in conflict with the framework of a harm-benefit analysis, the foundation for the American-style care system. At the BCJ, these different care systems coexisted under the same core principles of scientific responsibility. However, at the laboratory level, these dual systems could create tensions. The ways that these moments of tension emerged were always contingent on specific situations, actors and experimental settings. In the following sections, I illustrate several of these moments by looking closely at the division of labour within one of the sub-teams in the Tanaka lab, the transplantation team, led by the medical scientist Nakayama. Starting with a particularly eye-opening encounter from my fieldwork at the lab, I consider how one member of this team, Takuro, improvised care for animals, filling in the gaps that were not covered by these two systems.

Improvisation 2: Heads and hands

In general, life science research requires a huge amount of time and effort to prepare for experiments; mixing chemicals, caring for cells and breeding animals all require a significant investment of labor. The Tanaka laboratory had the funding to strategically hire many technicians to help with this heavy workload; in fact, over a third of the members were technicians who supported scientists – an extremely high ratio compared with other laboratories at the BCJ. Experienced scientists did the work of management, such as planning experiments, directing technicians and summarizing and analysing data. Technicians bred animals, conducted experiments that did not require special skills and killed animals during the course of their research. Thus, at the laboratory level, contrary to Kato’s intentions, it was not scientists who did everything, including taking responsibility for the ethical care of animals; in practice, there was a highly developed division of labor.

A specific incident between two of the members of the Nakayama sub-team helps to illustrate the rippling effects of this division of labor. It occurred 1 day when I headed to an experimental room allocated to the Tanaka laboratory, with an experienced technician named Kazuko and a new technician, Takuro, both members of the Nakayama team.

Soon after Kazuko and Takuro arrived at the room, they started to prepare for the experiment. The ‘perfusion’ procedure they would be performing that day was a basic method used to ‘fix’ the tissues of bodies, which was a necessary step in obtaining samples. Collecting these samples was a routine part of the experimental process. As a skilled technician, Kazuko quickly finished the first procedure. After a rat was anesthetized, his blood was removed and a special chemical was inserted to fix the tissues in the body. Then she asked Takuro to harvest the rat’s organs, which were to be sliced and immunostained. Compared to Kazuko, Takuro’s hands were very awkward. He needed to procure the intact organs, but he was struggling with how to use the scissors on the rat’s limp body. After Takuro managed to procure the organs, he closed his own eyes and put his hands together in a prayer position. ‘Itadakimasu’, he said in front of the dead rat.

Takuro’s behaviour surprised me, and I continued to think about this encounter after that day. I knew this word well, but in a very different context. Itadakimasu is difficult to translate into English; it is a fixed phrase that is used during meals, similar to bon appétit in French. Literally, it means, ‘I am receiving, thank you’. In everyday life, Japanese people often say itadakimasu to food before eating it, to show their appreciation. However, for me, the connection between using animals for experiments and the act of eating was not immediately apparent.

What did saying itadakimasu after killing in the laboratory mean? In observing how scientists and technicians behaved, I found that it was not uncommon for people working in labs to put their hands together in prayer and say ‘thank you’ to dead animals. However, saying itadakimasu as Takuro did was quite unique, as compared to simply saying thanks. Considering that Takuro was working at a meat company before he started at the laboratory, I wondered if he connected his experiences of selling meat to people to eat and using animals in scientific experiments. Perhaps this practice reflected his idiosyncratic behaviour and ethics more than anything else. As I considered his care practices further, however, I came to see that his response to this unfamiliar situation was also shaped at least in part by tensions between the combined care systems of the BCJ, as well as the particular division of labour in the Tanaka laboratory.

In this case, it was Nakayama, a senior medical scientist, who planned and managed the experiments, while technicians and PhDs did the rest of the work. Nakayama’s team’s overall aim was to develop new treatments for diseases by transplanting tissues to regenerate nerve cells in bodies. In a typical experiment, when the mice reached 6–8 weeks old, cultured tissues made of stem cells were grafted onto their bodies. Then, team members conducted behavioural analysis for about 240 days to identify whether the transplanted tissues improved bodily function. Finally, technicians killed these animals, extracted their organs, created samples and checked the tissue integration through a microscope. Over the course of the 8 months of the experiment, all of the procedures, the image data collected of cell conditions and the results of behavioural tests were carefully recorded.

Throughout this process, Nakayama offered instructions, but she did not always explain the experiment fully to technicians. Hence, sometimes technicians were not sure about the precise mechanisms at work or what a particular procedure meant for the whole experiment. This revealed one aspect of the relationship between scientists and technicians, which was sometimes explained using the analogy of ‘heads and hands’. Nakayama, the head, engaged in some parts of the experiments – for example, the transplantation surgery – which required her expert medical knowledge and skills. Takuro and other technicians, the hands, took care of the rest. Often, Takuro was the one who held the mice and anesthetized them for transplantation. He conducted the behavioural tests on the animals for several months. And, when that phase of the experiment was finished, he was also the one who killed and dissected the animals.

Thus, contrary to the principal values espoused by the institution, which claimed that scientists should take primary responsibility for the animals, in the Tanaka lab the lead scientists planned the experiments, while technicians like Takuro were tasked with the ethical labour of caring for and killing these animals. Taking a further look at how Takuro negotiated these encounters reveals how these divisions created other kinds of tensions, requiring further variations on these improvisational practices.

Improvisation 2, variation a: Being mushin

Even though several mice might have the same genetic makeup, for Takuro individual mice had distinct personalities. Since he was observing and touching the mice almost every day, he recognized each animal’s character, as well as emotions such as pain, fear, anger and pleasure and dealt with them accordingly. Engaging in 8 months of daily experimentation enabled him to cultivate a sense of attachment and responsibility for the individual animals.

Importantly, however, Takuro also maintained some sense of detachment throughout the experiments. He knew that for the sake of the scientific data, every single animal would be killed in the end. Takuro tried to understand and imagine how the rodents suffered and how humans could offer better ‘working conditions’ for the animals in their care. This was because he understood the animals as ‘co-workers’, who contributed to the experiments in the same way as human technicians and scientists.

Even though Takuro maintained a sense of detachment, it was not easy to kill the co-worker who had contributed to the experiment with him. Even more, killing an ikimono created an inherent tension that was not easily resolvable. However, at the end of the experiments, Takuro still needed to take his co-workers’ lives in order to obtain tissue samples. Once every several weeks, Nakayama asked him to ‘sakku’ – a shortened version of the English word ‘sacrifice’ – the animals who had been in his care. Sometimes, Takuro ‘sacrificed’ dozens of animals a day. When it was time for him to kill an animal, he separated his mind and his hands in order to release the tension he felt. Some technicians expressed this state as a feeling of ‘being mushin’, or ‘being of no mind’. If they hesitated, the animals would suffer more.

To become a ‘good’ technician, then, it was not enough to have training in technical skills; these psychological skills were also important. This was a part of how Takuro (and other laboratory members) coped with the ethical tensions they felt. However, Takuro admitted, ‘Although I have experienced it a lot, I have not become used to it. Every time I take their life, I say itadakimasu to animals’.

Improvisation 2, variation b: Saying itadakimasu and the web of life relations

After the perfusion experiment I described earlier, I asked Takuro if I could interview him. During our conversation, Takuro started off by talking about his career trajectory. After he got his Masters, he started to work at a meat company, which involved cooking and selling meat at a supermarket. But soon after, he developed a chronic illness and stopped working at the company. While resting up and dealing with this illness, he saw an opening for a technician position in the Tanaka laboratory and decided to apply.

After he started to work in the lab, he began to deal with mice and rats. I asked how he felt when saying itadakimasu to the animals and why and how he started using this phrase. He explained:

For today’s lunch, I ate katsudon [a bowl of rice topped with a deep-fried pork cutlet and egg that is popular in Japan], and I said itadakimasu to the rice, pork and egg. It means we receive their life. I think making knowledge and eating are the same. For scientific experiments, I take their life and make it into better knowledge. For food, I take their life and make it better for our bodies. Humans cannot live without receiving the lives of other ikimono.

Then, he vividly depicted his memories of working with cows for his previous research in the field of genetic engineering, describing in detail the building where the cows had lived, the facial expressions of the cows before killing them and the movements of their death throes. He had strongly willed himself not to forget about them. I was struck by how he repeated phrases like ‘I take their life seriously’ again and again in our interview.

As I noted, before eating many Japanese people today put their hands in a prayer position and say itadakimasu. The phrase includes nuances of ‘respect’ ¹¹ and ‘appreciation’ for the humans and non-humans involved in making the meal. In contemporary Japan, saying itadakimasu plays an important role in moral education around the topic of food and eating. But according to the historian Kumakura (1991), saying itadakimasu is a relatively recent invention, which began with the homogenization of education in the 1910s and 1920s (from the end of the Meiji period to the Taisho period). By teaching children to say the phrase, educators encouraged them to imagine the ecological relations embedded in eating – that is, to think about where their food comes from and to appreciate the animals, vegetables and farms and the people who raise and cook food. In other words, children learn about the web of relations among various ikimono’s lives. The emphasis is not on hierarchical relations, with humans at the top and the rest below; rather, the notion of itadakimasu implies that humans should respect and appreciate those who have given their bodies and efforts as a part of an interconnected web. Appreciating animals even after their death and thinking about broader relations means caring for ikimono, whether at the dinner table or in the laboratory.

During my conversations with people at the BCJ and other institutions that engaged in animal experiments, I found that it was not unusual to connect eating with killing animals for science. When I interviewed a technical staff member at one university about holding kuyou at their animal facility, he also drew this connection. His gray hair told me that he had been working at the facility for a long time. He explained:

We eat fish from the sea as food. We receive their ‘life’. For instance, a chef at a restaurant cooks a spicy lobster and serves us. It is the same as a laboratory animal. I think scientists and technicians should not have a feeling of guilt. What we should do is care for the animals and reduce their pain and suffering. Of course, we shouldn’t waste their life. We cannot live without eating other ikimono. Laboratory animals are essential for humans to make medicine and have a better life.

His explanation was also based on the web of life relations concept. However, his tone sounded defensive, as if he was protecting his job and his colleagues from criticism and stigma.

Historically, in feudal Japan, jobs related to killing animals, such as those in meat production and leather processing, were done by a discriminated caste called burakumin (Amos, 2011). Killing animals was seen as impure according to Buddhist ethics, and only those at the bottom of the strict feudal caste system would do this work. This system was abolished in the Meiji era, but there are still taboos around killing animals. Even today, though working in the fields of science and medicine is considered a respectable career, many scientists and technicians feel uncomfortable around this aspect of their work. During my fieldwork, I found that most people working in Japanese laboratories did not openly speak about and share their senses of personal ethics. The very awkward responses of my interviewees made me aware that it was considered inappropriate to ask people about how they felt regarding animal experiments. However, scientists and technicians also knew that they must kill animals as part of their job, without complaining or giving up. In order to do this difficult work, they had to care for themselves.

Thus, for Takuro, itadakimasu was not only about care for the animals but also about care for himself, to encourage him to continue his work. It was also about caring for the data produced by experimental science and medicine. By obtaining reliable scientific knowledge and getting results, he would ultimately contribute to the creation of new treatments for patients who suffered from nerve diseases and other chronic conditions, which he himself faced. Hence, saying itadakimasu and putting hands together in prayer was an acknowledgment of care for animals, for the self, for the data and for fellow patients, now and in the future – just as people, before they eat, appreciate the other lives and systems that produced the food on their plates.

By connecting eating and killing for science, Takuro did not borrow explicitly from the concept of the web of life relations; rather, he arranged the story by himself and reflected on both his actions and his narrative again and again. There was some historical basis to his improvisations; his prior experiences in the meat industry, his memory of cows and his suffering from disease all shaped his future improvisation of care at the Tanaka laboratory, but in loose and unpredictable ways. ¹²

Improvisation 3: Barbecue versus kuyou?

Another chain of improvisation followed from the self-governance care system at the BCJ. As we have seen, one of the coping strategies that Japanese scientific and medical institutions have created to deal with the tensions surrounding animal experimentation is to hold kuyou every year. Noticeably, however, the BCJ did not hold such a ceremony. Instead, they offered lectures every September. These lectures were obligatory for all staff members who used experimental animals. The BCJ also held an annual barbecue after the lectures, where the members of the institute gathered, ate and talked. When I attended one of these parties, I found that the absence of kuyou generated odd rumors around the significance of the barbecue party. This chain of improvisation revealed another connection between eating and killing.

Before I delve into rumors around the barbecue party, I would like to first consider why many institutions continue to hold kuyou every year. In Japan, kuyou are held for many purposes, not only for humans who have passed away but also for pets and even inanimate objects like dolls, needles or tools that are no longer useful. Within the agricultural and industrial sectors, companies often hold kuyou for animals and fish killed for food. The Japanese ecological anthropologist Akimichi (2010) argues that the reasons people observe kuyou vary. They may be showing appreciation, or acting out of fear and attempting to purify and pacify evil spirits, as exemplified by the kuyou for whales that are held in many places across Japan. Similarly, in my interviews, the purposes of these kuyou for laboratory animals were explained in different ways. Some told me that, during these ceremonies, ‘I express appreciation and say sorry to animals’. Or, ‘It reminds me that I use their life, that they are not mono (non-living beings). Otherwise we tend to forget about it’. On the other hand, going to a kuyou could not solve everything. A PhD student divulged, ‘Even though I go to kuyou, I cannot forget what I did. This is because I kill them for my curiosity. I’ve put this responsibility on myself. I will live with karma until I die’. Participants interpreted and used the opportunity of kuyou for multiple purposes, including showing appreciation, apologizing, purifying, memorializing and encouraging themselves.

When I was at the BCJ, I realized that there was little consensus about the absence of kuyou there, which led to speculation surrounding the annual barbecue event and its significance for the research centre. At the start of the event that I attended, in the auditorium of the BCJ, scientists and technicians gathered to listen to lectures. The veterinarian Hayashi, whose work as an expert in animal welfare I mentioned earlier, gave a talk about how scientists could avoid infections from animals. Another guest scientist also spoke about his research. I had expected to hear more of a ‘cultural’ commentary, such as ‘scientists should thank and console animal spirits according to Japanese tradition’, which was often said during mortuary rites. However, during these talks, the veterinarian and scientist kept to scientific explanations, as if purposefully excluding such discourse.

After the lecture, we went to the institute’s annual barbecue. Outside of the building, just next to the auditorium, scientists and technical staff members enjoyed beer and other beverages in the hot summer evening. In the garden area, members cooked meat and vegetables on a barbecue grill and served up food. Officially, the barbecue was a party for promoting interactions between animal caregivers and scientists, something that could be difficult during normal work, for the very practical reason that they all had to wear masks, caps, gloves and white clothes. According to Kato, this situation made mutual recognition difficult. However, this explanation was not totally convincing to me. I knew that every autumn, kuyou were held at almost every other university and institution across Japan. Why did the BCJ hold a barbecue party instead, while other scientists were thanking and consoling animal spirits?

Some of the scientists told me that they thought that the centre actually did have a kuyou every year, even though they had not attended before. For instance, one of the technicians in the Tanaka laboratory told me that she believed that Kato’s group held their own kuyou. ‘I’ve heard that Kato’s group members who are responsible for animal management offer the animal ceremony on behalf of all of us before the lectures and barbecue. I believe that every year the research centre offers this ceremony’, she said. However, when I interviewed an assistant of Kato’s group, she denied this story and explained:

It is strange. We don’t offer the ceremony. Well, I don’t know the details and it might be a private affair. I heard that maybe ten years ago, soon after the research center was built, the barbecue was offered as a substitute of the ceremony. But someone thought that eating meat as kuyou was not suitable, so it became a party for scientists and caregivers without our noticing it.

However, another senior scientist disagreed with this explanation. ‘No. We haven’t made the barbecue as a substitute for the ceremony. We always have said the barbecue is a party for scientists and caregivers.’ Clearly, there were differing stories about when and how the tradition of having a barbecue had been established, as well as conflicting opinions on whether the barbecue was connected with the (absence of) kuyou.

During this party, I asked one of the senior scientists who managed the animal facility why there was no kuyou at the BCJ. He told me: ‘I don’t like the ceremony because it is a mere formality and ritual. We need to say “thank you” to animals not annually but every day. So I decided not to offer the ceremony.’ Then he exclaimed, ‘Oh! That reminds me that only Japan offers this animal ceremony. This centre aims at having a “global standard.” That’s why we don’t have the ceremony.’ Later, I heard that when the institute was built it was the decision of this senior scientist not to have mortuary rites for animals. Leading members of the institute had studied abroad, and they were encouraged from these experiences to have a more ‘global style’ in terms of management. Evidently, the lack of a kuyou was also part of this ‘global’ identity. Moreover, the casual style of the barbecue, inspired by Euro-American institutions, was intended to encourage a lack of hierarchy and promote open communication between members of different ages and positions.

Reading this situation through the lens of the chain of improvisation, the absence of kuyou and the implementation of the barbecue party in the case of the BCJ could be seen as a disturbance of the way that care was usually choreographed in similar settings. It was easy to see how, for experimenters, this absence might cause tensions. In everyday life, however, scientists and technicians at the institute did not expressly mention or confirm who was in charge of such ceremonies or how they were held. Therefore, some people ambiguously thought of or imagined the barbeque party itself as some form of kuyou or assumed that someone else at the institute was holding a proper kuyou on their behalf. The technicians and scientists improvised these different stories, which allowed them to maintain some sense of alignment.

Conclusion

I have asked how scientists and technicians care for laboratory animals under the loose and eclectic regulations that govern scientific experimentation in Japan. To answer this question, I have sketched out the broader ethical contexts, in their affective, institutional and regulatory settings. With the concept of chains of improvisation, I have showed that the choreography of care easily collapses when different, incommensurable logics encounter one another, compelling people to improvise how to care for animals on the spot. These improvisations are more eclectic and unpredictable than the ‘tinkering’ that other STS scholars have found in similar settings with more consistent ethical norms, such as the UK.

As with the open-ended flow of movement in contact improvisation, in order to avoid collapsing the choreography of the work of animal experimentation, scientists and technicians at the Tanaka laboratory improvised how to care and for whom to care on the spot. Nobody forced technicians and scientists to say something or behave in particular way after their experiments were completed. Some said itadakimasu. Some said ‘I’m sorry’ or ‘thank you’. Others said nothing. Each person coped with this ambivalence differently. In other Japanese institutions, kuyou worked as one form of relatively stable choreography – a space where people could tinker within a given form. When kuyou was absent, as it was at the BCJ, scientists and technicians started to improvise. Their ways of choreographing care depended on their own ethical stances, specific laboratory settings, the patterning of the divisions of labour in this space, how and what animals they used and many other factors. Furthermore, as Takuro’s story shows, care extended not only to animals but to themselves, to the data and to colleagues and patients.

In exploring what it means to provide ‘good care’, laboratory members improvised how to care by bringing in steps and moves that they had learned in other contexts of their lives, such as everyday eating or spiritual practices. However, as I’ve shown, Takuro’s relationship with the animals in his care was not necessarily representative of broader cultural values, such as a general Japanese attitude towards animal experimentation. By considering these idiosyncratic actions within the context of Japanese histories of human-animal relationships, the lack of national regulations and the fraught practice of animal experimentation more generally, we can see how such personal attitudes and cultural concepts of gratitude tend to become intertwined or conflated in contemporary scientific settings. Looking closely at these improvisational relationships that Japanese scientists and technicians have with the experimental animals in their care provides openings for STS scholars to continue rethinking existing narratives of research ethics and questioning what ‘good care’ for laboratory animals entails.