Transgenic organisms are at once completely ordinary and the stuff of science fiction. (Donna J. Haraway)

Transgenic animals are organisms that have been genetically altered on purpose, not by mere chance: genes are knocked out, knocked in or reinforced, and the effects can thus be studied in a living (or dead) whole animal. Above all, it is mice that are being used, and since the early 1990s, the practice has spread like wildfire. Transgenic mice can be mail-ordered from commercial breeders nationally and internationally, or be produced in line with requirements by transgene units at larger universities in what can be called ‘technoscientific bespoking’ (Michael, Reference Michael, Brown, Rappert and Webster2001). There is a whole range of applications, and new ones continuously appear. For example, transgenic mice are used for basic research, in order to study in detail the role of a certain gene, or as disease models. Within the European Union (EU), and also in Sweden, which is the national context of our study, experiments with transgenic animals at present represent about a third of all animal experiments performed (Swedish Board of Agriculture, 2008).

The structural conditions that frame the handling of transgenic mice include the legal context, the international, national and local policy regulations and the ethical review process. First of all, the Animal Welfare Act states that experimental animals should not be subjected to ‘unnecessary suffering’.Footnote ¹ Since 2007, Swedish experiments on animals have been controlled by the Board of Agriculture. Seven local animal ethics committees scrutinize all projects dealing with animal experiments, and the reviewing process is mandatory.Footnote ² Concerning transgenic animals, it is business as usual, with one exception; you have to state in the application whether the project concerns genetically modified animals (Nordgren and Röcklingsberg, Reference Nordgren and Röcklinsberg2005). When the application has gone through the review process—which it usually does, as 99 percent of all applications are approved (but often with certain conditions, Nordgren and Röcklingsberg, Reference Nordgren and Röcklinsberg2005)—it works as a tool for researchers, animal technicians and veterinarians who decide on treatments and procedures. Besides a description of experimental procedures, the application states what anaesthetics should be used, what the ‘endpoint’ of the experiment is, and the degree of suffering that could be expected. These general welfare issues are evidently relevant also for transgenic animals. In addition, the welfare literature describes certain specific problems. Examples of prevalent transgenic dilemmas concern the many ‘surplus animals’ that are used in production and breeding, the severity of burden on the animals used for production (especially on the female donors and surrogates), and the prevalence and risk of unexpected phenotypes (Schuppli et al., Reference Schuppli, Fraser and McDonald2004). Difficulties in assessing the phenotypic and welfare status of GM animals and in establishing central databases with associated data on phenotypes are also mentioned as specific to transgenic animals (Nuffield Council of Bioethics, 2005).Footnote ³

At a somewhat different level, an ethical concern not often mentioned is: what right do we have to genetically modify animals at all? This concern is closely related to the ecological dilemma: what does it mean for an ecological system when natural selection is put aside and nature instead is adapted to the laboratory? This is a well-rehearsed theme in discussions of genetically modified crops, but it does not seem to bear the same weight in relation to laboratory mice, as they are confined to a lab existence. Furthermore, transgenic animals become problematic at a cultural level; they can be understood as boundary walkers (or perhaps ‘crawler’ is a more appropriate noun), constantly balancing on the fine line between nature and culture, organism and innovation, reality and model, science and technology (see Haraway, Reference Haraway1997). These novel creatures constitute forms of techno-scientific hybrids and, as such, simultaneously challenge and confirm cultural categories and dichotomies (Brown et al., Reference Brown2006). Or, as Haraway so aptly puts it: ‘transgenic creatures, which carry genes from “unrelated” organisms, simultaneously fit into well established taxonomic and evolutionary discourses and also blast widely understood senses of natural limit’ (1997: 56).

This article analyses how people who handle transgenic animals in practice—laboratory workers and members of animal ethics committees—talk about and handle the dilemmas exemplified above. We show how a tension between the animals as normal, meaning ‘like any other laboratory mice’, but at the same time highly valued and treasured for bio-medical research, is built up, contributing to a discourse on transgenic mice as ordinary treasures. Towards the end of the article we discuss how this discourse tends to exclude possibilities of discussing the specific dilemmas of genetically modified animals mentioned above. Instead the discourse contributes to certain transgenic silences.

Discursive silences

Research, as discussed above, has pointed out specific problems with genetically modified animals. Science studies, like this one, often articulate transgenic animals as border crossers between culture and nature, inventions and organisms. When we started our project, ‘Dilemmas with transgenic animals’,Footnote ⁴ we expected that this ambiguity would also be articulated by our interviewees and in the field data. Nevertheless, perhaps the most striking result of our research is that people who work with and/or ethically scrutinize research with transgenic mice seldom articulate that there are any ethical issues specific to this branch. Neither has the issue of transgenic animals become an ‘issue’ in mass media, in contrast to similar businesses as genetically modified crops, cloned animals or genetic tests. We would like to probe to see if this silence can be understood as a discursive effect of institutionally produced exclusion procedures (Foucault, Reference Foucault1972) and, if so, how these exclusion procedures are constructed. In our analysis of how people within the business talk about transgenic animals, we depart from a discursive perspective that focuses on how people make use of different interpretative repertoires—culturally available sets of statements. Wetherell and Potter (Reference Wetherell and Potter1987, Reference Wetherell and Potter1993) maintain that people make use of these interpretative repertoires and construct versions of reality in relation to the social context. Speech is viewed as a social act, rhetorically organized so as to bring to the fore a version of reality that fits the context (Potter, Reference Potter1996). Brown and Michael (2001), for example, have shown how scientists and medical practitioners routinely switch between different repertoires in their legitimizing of porcine donors of organs for human transplantation. Depending on the comparison point, human or non-human primates, different cultural-ethical or natural-scientific repertoires are used—all with the discursive effect that pigs are (scientifically speaking) similar enough to humans to be used as organ donors, but (morally speaking) different enough to be used as organ banks. By switching between these two repertoires, the pig is constructed as the perfect donor—similar and different at the same time.

In the present article, we use the concept of repertoire to understand how informants talk about transgenic mice as ordinary on the one hand, and as a means of scientific breakthroughs, on the other. Within the transgenic animal discourse, these repertoires, which are paradoxical on the surface, seem to be the most available ones. They are so strong that other repertoires, dealing with, for example, welfare, ecological and cultural problems become unavailable, almost ‘unspeakable’ (see Billig, Reference Billig1999; Kulick, Reference Kulick2005). The highlighting of some aspects of transgenic practices has the discursive effect that other aspects become hidden or silent. We would like to stress that we are talking about discursive effects, rather than about intentional strategies.

Fieldwork

The project the present article builds on consists of two parts, and the empirical data were collected through ethnographic methods, i.e. observations and interviews. One case study (in quotes referred to as Laboratory Workers; LW), conducted by sociologist Tora Holmberg, derives from research practice and concerns how researchers, laboratory assistants and animal technicians handle dilemmas in talk and practice. The data are diverse and consist of observations from a two-week course for researchers (Holmberg, Reference Holmberg2008), fieldwork with several research groups, including observations of animal experiments in practice, and fieldwork at two different animal houses (where the animals are bred and kept). In addition, a total of 20 semi-structured interviews were conducted with researchers at different levels, laboratory assistants and animal technicians from two different Swedish universities.

In another case study, ethnologist Malin Ideland investigated how members of animal ethics committees talk about transgenic dilemmas, both in committee meetings and in individual interviews (referred to in quotes as Ethics Committee Members: ECM). Twenty members of animal ethics committees were interviewed during 2006 and 2007. Of these 20 members, three represent animal welfare or animal rights organizations, six are representatives of political parties, ten are scientific experts from different disciplines and one is an animal technician. The interviews were semi-structured. Twelve meetings, both in preparation groups and plenary meetings, in six different committees were observed.

Interviews as well as observations have focused on how researchers and committee members handle ethical issues associated with animal experiments in general and transgenic animals in particular. These legitimizing processes go on at different levels. Holmberg shows, for example how laboratory staff primarily focus on one rather technical dilemma: where to take the biopsy for genotyping, thus back-grounding other, perhaps more urgent, dilemmas (Holmberg, forthcoming). Elsewhere, Ideland (2009) points out how members of ethics committees talk about possible future scientific breakthroughs rather than current ethical problems. In the present article, we focus on how dilemmas associated with transgenic animals become back-grounded through rhetorical comparisons with ‘something else’. Through comparisons, transgenic animals are discursively constructed as ‘normal’, ordinary and thereby unproblematic, and as valuable treasures in which hopes and expectations of future medical treatments are embedded. Based on the empirical material, we have constructed three categories of rhetorical comparisons used by the interviewees. These categories are:

• comparisons with other laboratory mice

• comparisons with other animal enterprises

• comparisons with humans.

The categories are sometimes overlapping, and they are not the only ones possible. However, because they are so prevalent in the data, they deserve a thorough analysis. The empirical sections deal with how informants use these different points of comparison as discursive resources in their narration of the transgenic mouse as an ordinary treasure. We also want to highlight how the embedded uncertainty and the potential of unexpected risks associated with genetic modifications are turned into something positive, a scientific advantage. The unexpected and uncertain add to the hopes of finding something new and untried, something that ‘perhaps’ can contribute to a scientific revolution. However, if this is to be perceived as ethically acceptable, risks for individual animals must be rhetorically minimized. At the end of the article, we also discuss the consequences of discursive silences.

Comparisons with other laboratory mice

In the introduction, we mentioned that transgenic animals can be ordered according to needs: some research teams use established models, i.e. developed transgenic mice that are transported from commercial breeders to the animal house, and can be used as soon as they are out of quarantine. Others develop and produce their strains themselves, with the help of their university’s transgene unit. This process is complicated and time-consuming; it takes several generations to get the required results. In this first empirical section, we will illustrate how possible problems concerning production and breeding become articulated in comparisons to other, ordinary laboratory mice.

An issue connected to the origin of the mouse (established model or locally produced) is what the phenotypic consequences of the genetic modification may be. In applications for ethical review, possible welfare problems are expected to be discussed. Sometimes these welfare problems are easy to specify when the model is already characterized, for example by having a greater risk of developing tumours. It is common to use transgenic mice as disease models, and phenotypic consequences like disease symptoms are then already known.

According to this quotation, it all depends on the alteration you make, but the consequences seem to be quite easy to predict. At other times, the prediction is less easy to make. When the model has been locally produced, phenotypic consequences can be rather tricky to predict, and different tests to characterize early behavioural and other deviations have been developed. Most of the interviewees agree that dramatic and unexpected effects seldom occur, which in itself can become a problem:

What is good for the animal—in this case lack of unexpected effects—is not always good for the experiment, which confronts the researcher with a dilemma. On the one hand, you aim at certain effects, and on the other, you might have to break off the experiment if they occur. The degree of awareness of this specific dilemma associated with transgenic animals varies significantly among animal ethics committee members. Several of the members talk about the problem of undesired phenotypes. One politically elected member, who at the committee meetings criticizes most experiments reviewed, describes and reflects on the silence on this matter:

Uncertainty over phenotypic consequences is a dilemma some of the interviewed ethics committee members reflect on mainly when responding to a direct question. They talk about juvenile mice that are considerably smaller than their wild-type siblings, mice with too short legs and without a lower jaw. But most often they emphasize the transgenic normality: ‘they act and look like any other mice’ (Interview, animal technician Pia, ECM). Instead of problematizing the differences from wild types, similarities become the focus. Many of the interviewees accentuate visual similarities:Footnote ⁵

Having little to do with transgenic affairs, the ‘earmouse’—a widespread picture of a mouse carrying a human ear on its back—has become an iconic reference through which people articulate their different views on animals in bioscience (Brown, Reference Brown2006). Here, it is used with the effect of distancing GM animals from something far worse, stating that the animals the researcher uses look just like ordinary mice. It is also worth noting that when the experiment is said to cause the animal suffering, it is the suffering in itself, not the technology, that is described as a problem. The majority of interviewees mean that all laboratory animals suffer to varying degrees, and that this is nothing special to transgenic technologies. It is thus expressed as being one kind of suffering among others, and should be treated as such. Not even Petra, one of the few politically elected laypersons who often notes reservations and is oppositional at the meetings, considers transgenic mice to be something special: ‘Well, I don’t like gene manipulation, but I don’t think it is more unethical in relation to the animals than other experiments are. I guess’ (Interview, Petra, ECM).

The application sheet for ethical review offers a box to be ticked off if the application embraces the breeding or use of genetically modified animals. This box can be interpreted as indicating that the committee will give extra attention to the consequences of the genetic modification. But the question remains of whether and how the researchers can predict and express unexpected suffering. Anders (researcher, ECM) claims that a tick in the ‘GM-animal box’ is the same thing as expressing awareness of unexpected suffering. The tick mark becomes, in his view, a confession, and thereby the researcher has abjured his or her responsibility, something that Anders criticizes:

Comparisons with ordinary mice also mean that the transgenic animal becomes naturalized—it is just another mutated mouse. It may seem paradoxical to describe transgenic mice—a symbol of high-tech innovation—as something essentially natural. However, the interviewees portray modification of the genome as one end of a continuum. According to them, this is nothing new. ‘Nature’ has always made similar changes through spontaneous mutations, and nowadays it is done in the laboratories, in an accelerated process:

The prevalence of spontaneous mutations is often brought up as an argument in defence of genetic modification. The transgene technique is legitimized because natural processes can entail results similar to those of conscious biotechnological interferences. Talk about naturalness plays down and normalizes gene technology (see Butler, Reference Butler1990). It is also obvious that referring to nature ‘hides’ human agency, it just happens.

A related legitimizing instrument is history: humans have ‘always’ made these kinds of changes, not least through breeding:

Referring to breeding is a common technique to legitimize genetic modifications, both when it comes to animals and to plants. Once again it is a matter of placing the technology in a historical context. By using history, it is possible to avoid the fear and insecurity that often follows new technologies (Marwin, Reference Marwin1988). But history can also be used in the opposite way: to create fear of new techniques. For instance, gene technology has been compared to eugenics and Nazi’s experiments. Critical voices compare contemporary animal experiments with the consequences of pregnant women’s use of thalidomide. The side effects were, in that case, impossible to predict through animal models (Forsman, Reference Forsman1993). History can accordingly be used to either criticize or legitimize a technique. In this case, the history of transgenic animals is mostly used to point out the normality and naturalness of these mice. In sum, we have shown how ordinary laboratory mice, used as a discursive resource, function. The outcome is that the transgenic affair is constructed as not very problematic at all.

Comparisons with other animal enterprises

An apparently opposite discursive strategy is to distinguish a new technology from an already established one. This is done in particular through choice of expressions. A closely related example is the Swedish media debate on embryonic stem cell research. In an analysis of that debate, it was obvious that the stem cell technique was being established as something completely different from earlier controversial biotechnologies concerning gene technology and cloning. Instead of talking about ‘therapeutic cloning’, the concept of ‘nuclear transfer’ was used and the concept of ‘gene technology’ was only spelled out once during six months of debate (Ideland, Reference Ideland2002b). Favouring the use of some concepts, and not others, contributes to dissociations from old technologies that are more controversial in the public debate. Similar processes can be perceived in the public discourse on transgenic mice. Expressions like ‘gene manipulated’ or even ‘genetically modified’ are rare in the mass media, as well as in discussions and interviews. Instead expressions like ‘knock-in’, ‘knock-out’, ‘transgenic’ and—after the 2007 Nobel Prize in medicine—‘chimeras’ are used. These concepts do not have the same symbolic meaning, because they are ‘new’. Nelkin and Lindee (1995) write about the symbolic capital of the concepts ‘gene’ and ‘DNA’. By using these words, decades of cultural baggage—in the shape of hybrid-DNA, GMO, cloned sheep and dystopic science fiction—enters the scene. Thus, the normalization process of transgenic animals—the construction of them as ordinary—also means that transgenic mice are described as something more ordinary than other—more or less debated—phenomena.

Several researchers oppose the ongoing discussion of transgenic animals’ possible welfare problems by employing comparisons with other activities:

In defending the use of animals in research, it is customary to talk about how the situation actually is worse in other sectors, especially in the meat industry (Birke et al., Reference Birke, Arluke and Michael2007). In several interviews, the cattle breed Belgian Blue is used as a deterrent example of problematic phenotypes caused by ‘ordinary’ breeding. It is easy to understand this contrast as a discursive strategy used to handle dilemmas associated with the use of animals in research; like the researcher quoted above who does not want the mouse to be affected by unpleasant effects, but at the same time wants to be able to study these effects in his research.

Laypersons on the ethics committees also use the meat industry as an example of poorer practice. Petra is an otherwise critical committee member, but when she reflects in the interview over the use of different species for research, she changes her opinion:

Through this comparison, the laboratory life of experimental animals appears as nice and civilized, and easier to accept. The same strategy takes other, closely related, forms in talk about pets.

Pets become a comparing matrix for comparison in the legitimizing of animal experimentation. Dogs that are left at home alone for hours and hours, neglected guinea pigs and abandoned cats are used as examples of animals whose lives are far worse than the lives of lab animals. The interviewees point out that when it comes to pets, no ethical committees assess their lives, demand enrichment of the cages, daily supervision and careful health examinations. In other words, their needs are not fulfilled. Similarly, wild mice or house mice—among the most popular figures for comparison—are brought forward as a means to point out what a good life laboratory animals have, and the detailed ethical regulations that surround their existence in the lab. ‘Our mice’ are euthanized in a humane way, while a mouse in the house is beaten to death without the slightest hesitation or tormented to death in mousetraps (Field notes, LW).

In this comparison with other activities, we can see an interesting shift in the repertoires. Almost all interviewees, even representatives of animal rights and welfare organizations, maintain that a world without animal experiments is an impossible scenario:

The right to use animals for human purposes is, in other words, relatively unproblematized (with some exceptions). At the same time, the same people bring up problems with pet-keeping as an animal rights issue, particularly since pets do not serve a medical purpose. This is an obvious example of how one can shift from a medical discourse to an animal rights discourse in the same conversation—a shift that has specific effects in terms of dissociation from the ethical dilemmas associated with transgenic animals.

Another example of shifts in repertoires is when the researcher Lennart (ECM) justifies the use of dogs in experiments by comparing laboratory dogs’ lives to the life of a family dog. He means that the laboratory dogs live a more ‘natural’ life. They live in groups, are not housetrained and so on. In another context, Lennart highlights the problems associated with using dogs, as they are social beings. Many of the interviewees admit having a non-rational approach to the use of different species—they have more feelings for larger animals than for rodents, even if they cannot justify this at an abstract ethical level. Helen, a researcher in veterinary medicine, is one exception, however. She even told us that she has donated her own, severely ill, animals to the department, which must be seen as a strong argument for a good laboratory life.

Establishing similarities to house pets is also evident. For example, the researcher Hans makes a connection when he talks about euthanizing lab animals:

However, even though it is stressed that you, as a researcher, are responsible for your research animals just as for your pets at home, there are also some differences. Experimental animals are there for a certain purpose, to perform certain tasks. This is perhaps mostly evident in an interview with the researcher Joakim, who talks about all the positive aspects of enrichment separately; the animals are not there to have fun, but to be run in experiments (Field notes, LW).

Genetically modified crops are viewed as a much bigger problem than GM animals, according to our informants.

Maria talks about the risks associated with dispersion, the consequences for ecological systems and human health, and insecurities. A shift in repertoires takes place, from one based on welfare to an ecological one. The ecological risks associated with transgenic animals are portrayed as non-existent, because the animals are kept in a closed laboratory, and—as many of the interviewees stress—if they were to escape, they simply would not survive. The laboratory is the natural environment for the transgenes, an environment in which an enclosed evolution takes place (Haraway, Reference Haraway2004: 72). The animals are bred for this particular end, and thus additional possibilities to justify the experiments are constructed.

To sum up this section, comparisons with other animal enterprises/lives enable talk about the good life of transgenic mice—there is nothing special about them in general. It is stated that other animals, whether living in the wild, as pets or on farms, often live a much worse life than does the laboratory transgenic mouse, the rhetorical effect here being to back-ground any specific transgenic dilemmas.

Humans as comparison points

As stated above, in comparisons with other animals and crops, it is transgenic animals’ usualness and normality that is emphasized. But, when the interviewees instead use humans as a comparison point, another perspective appears. The transgenes then become transformed, through a shift in repertoires, from ordinary, common and safe, to cherished treasures. Thus a discursive shift takes place, but the meaning remains: There are no specific dilemmas associated with transgenic mice—they are like other laboratory mice—except for the hope that they can ‘perhaps’ open doors for future breakthroughs. They are ordinary, but still fantastic.

The future is central to talk about transgenics. It is—just like history—a major element in the legitimizing process, as it has been in many similar debates about new biotechnologies (Borup et al., Reference Borup, Brown, Konrad and Van Lente2006; Brown and Webster, Reference Brown and Webster2004). Hope for medical solutions becomes embodied through experimental animals, but also through talk about future patients—they become an important comparison point in the transgenic discourse. Transgenic animals and humans are not compared the same way as, for example, transgenic animals are compared to other lab animals. But in the discourse on transgenic animals humans are often present, as a reference point for understanding their behaviour and suffering (see below) or as legitimating animal experiments in the name of future patients. The research leader in the quotation above emphasizes the future possibilities by referring to earlier research. Another researcher brings forward the potential possibilities of the uncertain searching that sometimes goes on in basic research.

In this uncertainty scenario, the utility aspect that is supposed to be included in the cost–benefit analysis becomes extremely plastic. ‘Perhaps’ is a key term, a recurring word in the discourse on transgenic animals, not least in comparisons with humans. The animals will ‘perhaps’ develop a problematic phenotype. But they can ‘perhaps’ lead to a cure for millions of patients. The uncertainty is both good and bad, but it is the prospects of the good outcomes that most often win the battle, especially because there are patients who embody these expectations (Allen, Reference Allen2002). The following quotation is a description of an incident at a meeting of one of the animal ethics committees. The incident nicely highlights the expectations embedded in the transgenic animals:

The discussion goes on about how disease symptoms and deformations should be defined. In the application it is written: ‘Animals that show signs of disease (poor fur, behaviour disorder, loss of weight or poor weight increase) will be euthanized for tissue testing.’ But what happens with animals that show other symptoms? And what is disease and deviation? What can you expect and what is impossible to predict? Peter points out that he has knocked out genes in ‘a tremendous number of mice’ and is surprised nine out of ten times. Maja makes her first statement that day:

Hopes and expectations are central to discourses on new medical technologies. Mike Michael (2000) points out that these expectations can be viewed as performative. Through different discursive forms such as metaphors, narratives and other signs of promises of scientific progress and better (healthier) lives, they conjure up the future in the present (Brown et al., Reference Brown2006: 2). Hopes of medical breakthroughs are built into discourses on transgenic animals. ‘We cannot close those doors’ is an effective argument to overshadow fears of the unexpected and the potential suffering of the animals. The unexpected is to be expected (Borup et al., Reference Borup, Brown, Konrad and Van Lente2006: 295), it is what can open doors hitherto closed, further our understanding and, ultimately, help us develop new treatment methods.

The rhetoric of comparisons—constructing transgenic silences?

In the introduction, we gave some examples of dilemmas involved in the handling of transgenic animals. We asked how the use of comparisons as a discursive strategy diverts from these dilemmas. In the section ‘Comparisons with other laboratory mice’, we investigated how comparisons with ordinary, inbred laboratory mice contribute to the construction of ‘normal’, non-deviating transgenic animals, in relation to which problems of unexpected phenotypes, for example, are minimized. By naturalizing the transgene production, any welfare consequences are projected on the animal, not on the transgenic procedures. In the next section, comparisons with other animals and other animal enterprises/lives were investigated. Here, we showed how the shift in repertoires enables a portrayal of the ‘good life’ of transgenic mice and minimization of the risks of genetically modified laboratory animals. In comparisons with other lab mice, as well as with other animals and fields of GM applications, suffering appears as a problem associated with the mouse, not as something caused by technologies. Moreover, human agency is veiled by the repertoires used. Perhaps the most obvious example of this is the comparison with random mutants, ‘nature has always done this’. But, even technologically speaking, there are differences between breeding on random mutations and genetically modifying animals. Experiments with mutations are categorized as classical forward genetics, where the ‘abnormal’ phenotype is the starting point for studying a genotype. When using reverse genetics, as in transgenic technologies, it is the phenotype coming out from the genotype that is in focus after a controlled genetic change. Human agency has a more than obvious place in the transgenic processes, but by using comparisons of different kinds it is hidden behind naturalness, history and future.

This non-articulating of human agency could be understood as part of the black-boxing process, in which the production of scientific results is made invisible (Latour, Reference Latour1999). In this process, the notion that the human factor is decisive for valid scientific data is erased; matters of fact instead become regarded as a ‘mirror of nature’ (Shapin and Schaffer, Reference Shapin and Schaffer1989; see also Kruse, Reference Kruse2006). In a similar way, successful—or unsuccessful—transgenic animal models are not represented as the results of human agency. Instead, they are constructed as the results of scientific progress and, in the black-boxing process, human involvement, and thus responsibility, vanishes. Perhaps the notion of responsibility is decisive for the use of animals in research. Schuppli et al. (2004) consider that responsibility should be added as a fourth R, the aim being to enhance the quality of life of animals in human care, and to reflect real costs in analyses of costs and benefits. Haraway (2008) discusses the possibility of transitioning from an instrumentalist to a responsible, or response-able, relation between human and non-human animals in the laboratory. One step in changing this relation might be to make human agency visible in the production and use of transgenic animals. Without articulated discussions about human involvement, human responsibility does not become either explicit or problematized.

In the last empirical section, ‘Humans as comparison points’, we discussed how the use of human patients as a comparison point contributes to an economy of hope and scientific breakthroughs. By talking about medical progress, discussions of potential and actual ethical dilemmas are left out. These discussions belong to unavailable repertoires. Furthermore, the transgenic mice are talked about as tools, used for (perhaps) creating solid bio-medical treatments. In the public discourse, biotechnologies are almost synonymous with the language and imagery of futuristic breakthroughs. In this discourse, a medical reason is always a good enough reason to use animals, as well as—for example—embryos, for the purpose of improving human health (Brown, Reference Brown2003). The hope of biotechnological salvation leads to an instrumentalization of animals. The animals lose their own inherent value and become dispensers of hope for medical solutions, for ‘the good life’ (Brown et al., Reference Brown2006; Michael, Reference Brown and Michael2000).

Ordinary treasures

Taken together, what we have done here is to investigate the discursive effects of comparing the use of transgenic animals to other phenomena. Different rhetorical comparisons are used, legitimizing the transgenic enterprise, and the message is that transgenic animals are simultaneously ordinary stuff and treasured rarities. With the help of comparisons with ordinary laboratory mice, other animals and GM fields of application, natural processes and humans, the increased use of genetically modified animals comes out as altogether unproblematic. Rather it appears as an obvious continuation of both natural and historical processes, leading towards better treatments for severely ill people. The handling of experimental animals appears to be good in light of the treatment of other animals. The ecological risks associated with genetically modified animals stand out as minimal in comparison with GM crops. The flexible use of—and frequent shifts between—interpretative repertoires highlights cultural norms concerning ecology, human health and animal suffering in an interesting mix. Through an emphasis on the laboratory as a natural environment—a specific ecological setting run by technology, research funding and a wealth of innovations—animal welfare arguments concerning the ‘unnatural’ environment of experimental animals seem weak. The laboratory as a closed system is also put forward as a guarantee against negative ecological effects. By promoting the potential medical benefits for humans, the interviewees’ statements add to a hegemonic discourse on biotechnological salvation (Brown, Reference Brown2003; Haraway, Reference Haraway1997).

But then again, is there ‘actually’ anything special about transgenic animals? Are they not just like ordinary mice? Or are they just the contemporary endpoint of a beaten track of domestication of other species (see Cassidy and Mullin, Reference Cassidy and Mullin2007)? To help us discuss this issue, we look at the work of Sarah Franklin and her analyses of Dolly the cloned sheep (Franklin, Reference Franklin1998, Reference Franklin2007). Franklin concludes that there is something rather special about Dolly, not because of her iconic status in terms of scientific breakthrough and progress, but because her mere existence calls into question many scientific as well as cultural assumptions. This is a matter of, for example, sexuality and reproduction, life and mortality, kinship and relatedness, and, not least, origin and genealogy (Franklin, Reference Franklin2007). Paradoxical as it may seem, what makes Dolly a successful clone is that she is unique. But, in the public presentation of Dolly, her commonness also came out strongly. She was not described as a monster, just as an ordinary sheep (Ideland, Reference Ideland2002a). In a similar vein, we state that there is something special about the transgenic mice, both on a welfare level and a cultural level. They challenge our system for cultural categorizations in their crossing over borders between, for example, biology and technology. Haraway writes that ‘trans’ is the product of the crossing over of nature/culture borders (1997). We argue that one of the problems with the discourse is precisely the disappearance, or silencing, of the trans-thing, and thus we would like to reinstate ‘trans’ as a troubling tool, the aim being to counteract silences in institutionalized routinization For example, there are suffering and thus welfare issues connected to the transgenic technology, like expected and unexpected phenotypes, and surplus animals. ‘Trans’ can thus be viewed as a special kind of suffering. But this specialness is hidden under a rhetoric of commonness, which contributes to how actors within the apparatus manage cultural and ethical dilemmas. We want to stress that shifts between some repertoires, and exclusions of others, contribute to the construction of transgenic silences. In the discourse, issues about welfare problems, uncertainty in transgenic research, and human agency and responsibility become ‘ethical non-issues’. We argue that the debate needs more available repertoires. For a democratic, responsible and public debate on science that can contribute and raise new concerns, anything but silences are needed.