Karin Knorr Cetina. The Sage Handbook of Sociology. Editor: Craig Calhoun, Chris Rojek, Bryan Turner. Sage Publication. 2005.
There is a widespread consensus today that contemporary Western societies are in one sense or another ruled by scientific knowledge and expertise. Science and technology were a driving force in the transition from traditional to modern societies and they are central in explaining the great socio-economic transformations early in the industrial revolution and the later progress of industrialization. Today at the beginning of the twenty-first century many believe another new epoch is in the making, and science and technology are again deeply implicated in the changes under way. Most of the concepts that have been suggested to refer to the new system, including labels such as ‘postindustrial society’ (Bell, 1973; Hage and Powers, 1992), ‘postcapitalist’ society (Drucker, 1993), ‘technological society’ (see, for example, Berger et al., 1974), ‘information society’ (see, for example, Beniger, 1986; Castells, 1996), ‘risk society’ (Beck, 1992) and ‘knowledge society’ (Stehr, 1994), embody this view. One major line of reasoning holds that the new order is postin-dustrialist by virtue of moving from a system based on heavy industry where capital and labor are the engines of economic growth to one where knowledge is the main productive force. One recent source of this awareness is Daniel Bell (1973), for whom knowledge in the form of scientific theory has become an axial principle that accounts for changes in the division of labor, the development of specialized occupations, the emergence of new enterprises and sustained growth. A second line of reasoning centers more on postmodernity as an era of skepticism toward any absolute foundations of knowledge, as one in which a plurality of heterogeneous claims to knowledge has emerged instead and information technologies have become a dominant social force in shaping social life (Lyotard, 1984). This argument has been pushed furthest by Castells (2001: 3), who sees the ‘network’ as the message of the information technology revolution which, in his view, drives contemporary social transformations. Other assessments have highlighted further aspects of this knowledge and information transformation. For example, Habermas’s argument about the ‘technicization’ of the life-world attempts to understand the spread of abstract systems to everyday life (1981). Giddens (1990), arguing that we live in a world of increased reflexivity mediated by expert systems, extends the impact of knowledge to the self, pointing out that today’s individuals engage with the wider environment and with themselves through information produced by specialists which they routinely interpret and act on in everyday life. Most accounts see knowledge and technology from a social impact perspective: knowledge and technology are the independent variables that have a profound effect on the character of social and economic life. In these accounts, knowledge is sometimes formulated to fit longstanding beliefs about science (an example is Bell’s attempt to explicate knowledge in terms of theory 1973: 44), but it is in effect the last thing to be explained, having no reality independent of analyst’s models.
A Brief History of Science and Technology Studies
The specialty that attempts to break open such notions as science and technology, or scientific knowledge and information, is the sociology of science and technology. The sociology of science dates back to the late 1930s, when Robert K. Merton (1970 ) displaced the then-existing Marxist perspectives (see, for example, Bernal, 1939; Hessen, 1931) with a genuinely sociological approach to knowledge. In what came to be called the ‘institutional approach to science,’ some of the important questions concerned the social organization of science and the ‘institutional imperatives,’ or norms and values, that sustained the scientific attitude from within (the norms were universalism, disinterestedness, organized skepticism and ‘communism’—the collective ownership of scientific results). The generation which followed turned its back on the functionalist mode of reasoning which is evident in these concerns and which Merton, like Parsons, had adopted. It collectively moved away from the focus on social-structural and institutional processes characterizing scientific groups and organizations, arguing that science cannot be understood if the cognitive content of science and technology, and the processes of knowledge and technology creation, are not included in the analysis. This attitude was encouraged by the work of Kuhn (1970 ) and Feyerabend (1975), who espoused a philosophy and historiography of science in which they traced the interdependence of cognitive and social factors in the history of physics and other disciplines. The outcome of this collective change of mind was that the sociology of science and technology turned into a sociology of knowledge, also called ‘the new sociology of science’ and ‘science and technology studies.’ Unlike Mannheim (1936) and Scheler (1924), who had proposed a sociology of knowledge in the 1930s but excluded from it the natural sciences, the new generation insisted that the natural sciences must be at the center of attention of social studies of science. Mannheim and Scheler’s central thesis was that human thought was socially conditioned, as was knowledge in economics, and more generally in the human sciences; but they assumed, as did most scholars after them, that the natural sciences were exempted from such influence. The common assumption explicitly endorsed by the earlier institutional school had been that sociologists had nothing to say on the practice and content of science; the content of science was, at best, the concern of philosophers of science, who investigated the structure of scientific theories and the logic of discovery and justification. In contrast, what unfolded with the new sociology of scientific knowledge were research programs that took it upon themselves to demonstrate the relevance of empirical sociological analysis in the hardest possible cases, those of the natural sciences and mathematics. The resulting studies showed that the technical core of science could be studied empirically with social science methods to great advantage. Doing science and technology was a form of constructive action in collective contexts; though the processes observed were intricate and complex, they were open to examination to no lesser degree than those in other areas of social life. This new sociology of science was later expanded into a new sociology of technology (Pinch and Bijker, 1984) and both proved tremendously successful, conceptually as well as empirically, yielding a number of new programs and results discussed in the next section.
The seemingly deliberate effort to extend the perspective of the sociology of knowledge to the natural sciences, the very domain which had acquired a monopoly on defining what counts as knowledge from the Enlightenment onward—while at the same time establishing itself as being in some sense outside society when it came to its own internal processes—was the result of the convergence of several independent endeavors originating in the early and mid-1970s (Barnes, 1977; Bloor, 1976; Collins, 1975, 1985; Knorr Cetina, 1977, 1981; Latour and Woolgar, 1979; Lynch, 1985; MacKenzie, 1981; Traweek, 1988; Zenzen and Restivo, 1982). These efforts became enhanced and related to one another by simultaneous early formulations of the ‘sociology of knowledge turn in science studies.’ The most notable early formulations of the goals and musts of the new research program included Bloor’s ‘symmetry thesis,’ Collins’s methodological relativism (1985) and Knorr Cetinas and Latour and Woolgar’s assessment that scientific activities had to be seen as constructive rather than descriptive and that they are bound to particular sites, scientific laboratories. The symmetry thesis was part of four principles Bloor set out in his 1976 book and which circumscribed the strong program in science studies: the program was to be causal, that is, it should attempt to determine the factors that account for the convictions and knowledge-beliefs of scientists; it should be neutral in regard to true and false knowledge, successful and unsuccessful beliefs, ‘rational’ and ‘irrational’ outcomes—meaning both sides of these dichotomies should be explained, not just one; it was to be symmetric, meaning one should be able to make reference to the same causes as explanations of true and false knowledge; and it should be reflexive, meaning the patterns of explanation should be applicable to social studies of science and the results it comes up with. The main thrust of these principles, and notably of the second and the third, has been the symmetric treatment of true and false knowledge in empirical research. These principles declared normal, proper, successful, rational natural science research whose results counted as true or potentially true to be subject to the same scrutiny and explanatory variables as false, irrational, shaky, or politically mandated scientific results—the kind of pseudoscience that had been subjected to social explanations before. Bloor’s principles formulated as legitimate and long overdue what he himself and others in science studies had set out to do: make inquiries into the nature of normal science and ‘good,’ established knowledge processes. The new sociology of science and technology is characterized by a methodological relativism that follows from these ideas. In essence it holds that we must bracket any presumption of the rationality of science and our deeply entrenched beliefs in scientific authority. Only then will it be possible to study scientific practices and beliefs on a par with other beliefs and practices (Rouse, 1996: 5-7). Methodological relativism needs to be distinguished from judgmental relativism that takes all knowledge claims to be equally valid—which is not what the new sociology of science proposed. The constructionist program in the new sociology of science and technology extends the tenets of the symmetry thesis and of methodological relativism by adding an empirical strategy of making sense of science that pays attention to the (humanly) made character of knowledge: ‘As we come to recognize the conventional and artifactual status of our forms of knowing, we put ourselves in a position to realize that it is ourselves and not reality that is responsible for what we know’ (Shapin and Schaffer, 1985: 344). We shall consider this in more detail, in connection with a host of research results and perspectives that have since emerged in the area under examination.
Selected Results of the New Sociology of Science and Technology
Constructionism and Laboratory Studies
The source of the phrase ‘laboratory studies’ is a number of early on-site observation studies of knowledge processes in natural scientific laboratories (Knorr Cetina, 1977, 1981; Latour and Woolgar, 1979; Lynch, 1985; Traweek, 1988). For the first time in the history of science studies, these authors made a full-scale attempt to study the process of knowledge production in its natural setting, the laboratory. Perhaps coincidentally, the first studies were all done in California, a context that proved conducive to the intrusions into the ‘cathedrals of knowledge’ these authors attempted. The most radical outcome of the two earliest studies was an assessment which has itself developed into a dominating approach in recent science studies: the assessment that science was constructive rather than merely descriptive of the nature’ it addressed. There are weaker and stronger readings of ‘constructionism’ in science studies. One of the strongest is that the world as described by science ought to be seen as a consequence rather than a cause of scientific representations.
This claim becomes less radical if one understands it as the outgrowth of an ontological pragmatism characteristic of science and technology studies. Accordingly, one needs to distinguish between the pre-existence of a material world that is granted by every constructionist, and the concrete phenomena of experience that science comes up with and that have specific characteristics. These are thought to begin to exist as entities that can be reliably picked out and encountered only after science has articulated and defined them. This is perhaps most evident when the phenomena at stake are remote and invisible—examples are the top quark or the Higgs mechanism, TRF (a hormone releasing factor) and cell mechanisms and viruses. All these were not part of our life before they were named, described and otherwise designated by science. What exists reliably beyond such phenomena are everyday entities, which are always culturally defined and shaped; the classifications and characteristics associated with these entities vary between cultures, and they frequently conflict with the classifications and qualities produced by science. We have no access to any reality independent of such universes of meaning and practice. Hence not only must we see reality (as concretely defined in terms of encounterable entities with specific characteristics) as an outgrowth of everyday cultural or scientific practice, but we can also observe the construction of these objects in terms of changing cultural definitions and articulations, or, in the case of science, in terms of the activities and accomplishments observable in scientific labs. Note that this interpretation does not imply the sort of judgmental relativism that has been politically suspect to many because it leaves us with no tools to evaluate certain accounts as better than others. The advantage of ‘good’ scientific accounts, in terms of this notion, is pragmatic and rests with their difference from everyday accounts—good scientific accounts are backed by research programs and instrumental practices that contrast with everyday instrumental practices and backings. Instead of working with realist assumptions about the relationship between representations and nature, which it considers intractable, this sort of epistemic constructionism is based on a logic of differences between universes of knowing which it believes are the only tractable elements available to us. It should also be noted that truth itself is a historical notion, as is ‘evidence,’ ‘objectivity,’ ‘experiment’ and all other terms used in epistemology. Historical studies conducted in the spirit of the sociology of knowledge have demonstrated the cultural and temporal embeddedness and frequent transformations of these notions, thus making us acutely aware of the fragility and path dependence of their current status and epistemic authority (see, for example, Daston, 1991; Shapin, 1994; Shapin and Schaffer, 1985).
Laboratory studies have been the breeding ground for the notion that science is constructive, and they have sustained the ontological pragmatism just spelled out—among other things by pointing out the artificiality of laboratories and the preconstructed and artificial character of the ‘nature’ within them. In addition, the laboratory itself has come to be seen as a theoretical notion rather than simply a place where science gets done: the lab, in this view, is a knowledge tool that rests on the reconfiguration of the knower and the known and their relationship. For example, in laboratories scientists no longer confront nature-in-the-raw (like weather conditions, seasonal and temporal constraints on plant growth, the problems of observation in field-astronomy), but a nature miniaturized and remodeled in other ways such that it can be processed in a rationalized and accelerated manner. In the laboratory, nature is subject to ‘social overhauls’ that prepare it for inquiry. According to this perspective, it is the conventions embodied in laboratories and in the sequential succession of laboratory set-ups—rather than methodological principles of experimentation—that account for some of the successes of science. Laboratory studies have also shown the ambiguities, the ‘slack’ and the missing elements in research outcomes. Research outcomes are rarely clear, definitive and complete; laboratory studies have traced the negotiations and techniques of persuasion adopted to eliminate slack. Here construction takes on the concrete sense of social negotiation; it points to the interpretative leaps and the creation of ‘surplus’ meanings that compensate for gaps and ambiguities in scientific data. The constructionist approach is clearly less controversial in the case of technology, to which it has also been applied (see, for example, Pinch and Bijker, 1984; Bijker et al, 1989). Since technologies are artifacts, claiming that they are constructed causes no outrage, though the adoption of a constructionist methodology and a sociology of knowledge perspective by some authors (see, for example, Bijker, 1995; Law, 2002; MacKenzie, 1990) has helped to move technology studies away from a mere history of inventions and to bring into focus the cultural, political, social and other factors and the multiple associations through which technologies are built from within.
The categories introduced by the constructionist research program have been critically examined under the heading of reflexivity (see, for example, Woolgar, 1988; Ashmore, 1989). Taking seriously the reflexivity which Bloor’s strong program demanded, and indeed extending its range, Woolgar and Ashmore argued that constructionism is incoherent if it does not interrogate its own constructions of representations of science and technology. In other words, constructionism should not be understood as a program that attempts to improve representations of science, since its own categories and distinctions are equally constructed. Constructionists’ texts must be open to reflexive criticism of their own cultural practices and should express this awareness and evince attention to the constructedness of their own texts in their writings.
Studies of Experiments
In a post-Kuhnian move away from the hegemony of theory in science, not only sociologists but also social historians of science have given scientific practice more weight in the study of science and technology. In addition to the laboratory, experiments have become a focus of study. This work has produced a number of interesting results, some of which fundamentally challenge traditional beliefs about experiments while others highlight the role experiments play in particular natural sciences. Natural scientific experiments are commonly seen as the most important venue for settling knowledge claims and lending credibility to scientific results. But, as Collins’s and Pinch’s work has shown (see, for example, Collins, 1985), this assumption is problematic. As described by Collins and Pinch (1998: 11, 25, 98), whether or not scientific results can be replicated and what counts as a valid replication depends on agreements about what the important variables in an experiment are. Such agreements are normally taken for granted but are made explicit during controversies. As a controversy develops, more variables that potentially affect an experiment come to the fore. From the critics’ point of view, these can be used as excuses by the proponents of a particular experimental outcome when results cannot be replicated. For the proponents, they are reasons for why the unpracticed may have difficulty with the replication. Whether the experiment is flawed and the sought-after signal is really there or whether the signal is not there and the experiment is valid can only be decided if one knows the correct outcome, which in original research one does not, since finding out the correct outcome is the very point at issue. What the authors call ‘the experimenter’s regress’ is the circle implied by a situation where one has to build a good experimental apparatus to detect a signal, but cannot decide whether one has built a good apparatus until one has tried it and obtained the correct result, which one cannot determine until one has built the apparatus—and so on. Reaching experimental closure, then, in the natural sciences, is not a mere matter of setting up decisive experiments and of experimental replication. Some scientific results become discredited not because there is any published disproof that rests on decisive evidence, but because the field tires of them, a principal investigator dies or loses credit, or because more interesting problems come along. For the sociology of science, this means a large opening for studies that unearth the real venues of consensus formation and result stabilization in science and technology.
A second project centers around experimental systems, a term Rheinberger (1997: 27-30) proposed for a series of experiments connected to one another, and forming the ‘smallest integral working units of research.’ Rheinberger defines experimental systems as vehicles for materializing experimental questions that center around particular research objects; the notion is also used by scientists themselves to characterize the scope of their activities. Its usefulness lies in its pointing away from the notion of single experiments as the ultimate arbiter of truth. Single experiments can prove little and carry little conviction in scientific controversies. Yet as Rheinberger argues, even the argument about the experimenter’s regress embraces, in its very rejection of their decisiveness, the focus on single experiments. Rheinberger uses the notion of an experimental system to examine the history of molecular biology, which he finds to be neither the outgrowth of a unifying theory focused on the notion of information, nor that of the work of a few research groups led by prominent scientists, but of a number of scattered, differently embedded and only loosely connected heterogeneous experimental systems that sought to characterize living beings down to the level of biologically relevant macromolecules. Without explicitly using the term, Galison also examined an experimental system: a series of instrumental technologies defining a sequence of high energy physics experiments which, in this field, take many years and involve large international collaborations of scientists. Galison’s work (1997) situates itself more broadly with respect to philosophical understandings of science and with respect to cultural debates over modernity. Like others, he attempts to set the theoretically inclined philosophy of science right by demanding equal rights for theory, experiment and the material technologies of instrumentation. As a result of his studies of high energy physics’ experimental technologies in the last decades, he proposes a model of intercalation that splits apart these components of scientific paradigms, arguing that they may develop and change at their own pace relatively independently of other components. Thus the Kuhnian model, according to which science changes via the wholesale replacement of scientific paradigms during scientific revolutions, is not substantiated. Different components of experimental work follow their own logic and dynamics. In high energy physics, they are in the care of specialized and often fragmented epistemic communities (Brown and Duguid, 1991; Saxenian, 1996) that pursue these developments within their own distinctive frames of reference and meaning.
The laboratory study perspective has been combined with a network approach to yield what has come to be called ‘actor-network theory,’ an influential approach for analyzing the power struggles embedded in and, according to this view, defining science and technology (Callon, 1986; Callon and Latour, 1981; Latour, 1987, 1988, 1993; Law, 2002; Law and Hassard, 1999). Scientists, engineers and others build heterogeneous networks consisting of non-human objects (like microbes, scallops, or machines), colleagues, financial resources, publications, organizations/corporations and other elements to make their findings successful and unassailable. The emphasis here is on heterogeneity; the defining characteristic of this particular network theory is that it is not limited to human agents, but explicitly recognizes non-human entities as nodes in the network and as actors in any technoscientific game: in fact, non-human agency is a key term in this theory. A second key element is that the network is seen as a stabilizing arrangement. For example, technologies become more widely implemented and more successful as the networks expand; actor-network theory is the approach that has devoted most attention to technology, perhaps because both original proponents, Latour and Callon, work at a technological university (the Ecole des Mines in Paris). But the argument has been extended to science, where it claims, more importantly and more controversially than in the case of technology, that scientific knowledge claims also become more factual as the networks grow. Latour and Callon propose what might be called a network theory of truth, a view according to which the outcome of scientific and technical conflicts is largely determined by an ability to get others to ‘align themselves’ with a knowledge claim. Fuller alignment is equivalent to more ‘stabilized’ (more held to be true) knowledge claims that are difficult to call into question; deconstructing them requires significant resources that match those of the existing network. There is no provision in the theory for a master-mind or master-actor orchestrating the alignment (but see Latour’s study of Louis Pasteur, 1988). Instead, the non-human actors in a network are as important as the human ones, ‘co-constructing’ outcomes and their stabilization as well as network expansion. Despite the label ‘actor-network theory,’ which suggests an emphasis on individual agency, the approach tends toward seeing agency as distributed in a network.
If one wanted to read this approach as a causal theory there would be several problems. On the one hand, there is no indication of the conditions under which network construction and expansion are successful. Actor-network theorists appear to consider success a contingent outcome that can only be empirically determined on a case-by-case basis. Second, anything can build a network link; for example one’s holding a cup of tea establishes a link between a human actor and an object. Here the approach tends toward tautology and ignores the question of network boundaries: if anything is a link, nearly everything is already interrelated by virtue of anything going on in the world at all, and the theory can neither grasp new links nor explain success in establishing them. Furthermore, equating network expansion with stabilization and success must appear problematic; as we know from political theory, military expansions, colonialism and the like, networks can also become too large to be manageable and effective. Finally, the approach defines science as politics and bases truth on power, that is, on the alliances forged in heterogeneous networks. Yet at the same time, the edge is taken off this audacious hypothesis when objects are brought back into the picture as agents that resist alignment and that co-define network outcomes and scientific representations. In other words, since non-human actors (‘nature’) determine at least part of the events, realist assumptions creep back in, and, given the lack of a social theory of network construction and expansion, carry some weight in explaining fact-stabilization. Perhaps against the will of the authors, the approach appears to work best when human actors are foregrounded, when the analysts describe how powerful scientists such as Pasteur or Diesel built and shifted allegiances, recruited resources, rhetorically persuaded other parties of their success, and got things to work (‘forged alliances’) in the laboratory.
The Cultural Turn in Science and Technology Studies
The turn to the laboratory as the place of science had brought into view a whole universe of cultural activities implicated in research. The shift to analyzing science and technology as process rather than product proved tantamount to seeing science as culture defined in terms of particular sets of practices (Pickering, 1992). Kuhns views about the role of paradigms as holistic sets of methodological preferences, theoretical beliefs and sample cases that ground normal science had already suggested what Collins (1985) later called an enculturation model—the socialization of scientists into shared and implicit background knowledge that provides the basis for routine work. Since its beginning, the new sociology of science and technology has made references to the cultural make-up of what it was about. None the less, what took precedence in the early period was registering the social processes of formation from which science had been exempted for so long. Perhaps this explains why more focused works on the ‘cultural’ aspects of science and technology only came later. They look, for example, at the influence the cultural environment exerts on disciplinary, ‘culture-free’ science and on transnational science. And they have begun to analyze differences in epistemic practices between sciences, leading to the term ‘epistemic cultures.’
Consider first the question of whether scientific disciplines ought to be understood as transnational, culture-free traditions or whether these traditions are reworked and transformed as they are implemented in different cultures (Hess, 1995: 39-53,49). According to Hess, only a small number of studies have addressed this issue, but these studies provide evidence of such transformations in the social and natural sciences. One of the first studies of this kind was Sherry Turkle’s comparison of psychoanalysis in France and in the United States (1992). Turkle described how Freudian psychoanalysis became reoriented in the US ‘to a message of hope in a culture of pragmatic self-improvement’ and accepted by large populations, whereas in France existing research traditions in dynamic psychology led to a rejection of psychoanalysis as a general cure outside small circles of intellectuals, artists and writers. This picture holds until after 1968, when psychoanalysis gained the prominence in French culture that it had long occupied in American culture, although it was also rewritten by authors such as Lacan. While Turkle’s study can be interpreted as referring mainly to questions of national reception and to the human sciences, Sharon Traweek’s continuing comparisons of American and Japanese professional and organizational cultures in high energy physics have a broader focus (1988). Traweek showed how the funding systems and, as a consequence, detector designs varied across these settings. In the US, the detector design allowed for continuous detector rebuilding and surprise data, while the Japanese design emphasized reliability and precision at the expense of new data (see also Hess, 1995: 50). Similarly, Haraway’s study of primatology in Japan, India and the United States (1989) shows how different national beliefs play themselves out in different formulations of primatological methods. As Hess sums up, the Japanese ‘tend to view nature as something to be cultivated by hand,’ and hence tended to feed primates, whereas Western beliefs in a sharp division between nature and culture led to a rejection of provisioning of animals that are located in natural surroundings seen as wild. Then again, the cultural meaning of monkeys in India as belonging to a sacred supernature and as interacting with humans may have prompted Indian primatologists to investigate the interaction of monkeys and humans, whereas Western scientists focused on studying primates in their original (natural, wild) state (Hess, 1995: 50-1). A particular genre of studies best exemplified by Joan Fujimura’s work on transnational genomics (see, for example, 2000) focuses on both transnational science and cultural variations in the implementation of a particular project (see also Fujimura, 1996).
Most accounts of science and technology implicitly or explicitly assume the ‘unity of science,’ a notion associated with positivism and the Vienna Circle of philosophy of more than fifty years ago. Science connoted an attitude of tolerant skepticism and a particular method that promised the accumulation of findings and technical capability in the service of humanity (Rouse, 1996: 51). Postpositivist philosophy (see, for example, Feyerabend, 1975; Hacking, 1983; Kuhn, 1970 ) turned against positivism’s formalism and challenged the notions of linear progress and the assumption of a specific rationality of science. But it continued to talk about science as if it was somehow unified and all part of one bloc; questions of the epistemic heterogeneity of the natural sciences were simply not addressed, nor were they, until recently, raised by the new sociology of science and technology. For example, the debates raging over realist, skepticist, feminist and other interpretations of science all tend to assume that science is a unitary enterprise to which these interpretations can be applied across the board. This picture has now changed (see, for example, Galison and Stump, 1996), and the second cultural approach to be mentioned can be associated with the change. It brings into view cultural differences within science in regard to the understanding of measurement, the meaning of ‘empirical,’ the configuration of the objects investigated and of what counts as real, the emphasis placed on errors and failures of knowledge, the role and construction of laboratories, and the organizational practices implemented. This work sees different machineries of knowing at work in different scientific fields—composed of different empirical systems, different logics of instrumentation and different systems of epistemic authority and organization. The machineries add up to different ‘epistemic cultures.’ The first full-scale study of two epistemic cultures compares high energy physics and molecular biology (Knorr Cetina, 1999). Among other things, the study illustrates how high energy physics cultivates a kind of negative knowledge, which is not non-knowledge, but knowledge of the limits of knowing, of the mistakes we make in trying to know, of the things that interfere with our knowing, of what we are not interested in but still have to confront in empirical research. In a sense, high energy physics has forged a coalition with the evil that bars knowledge, by turning these barriers into a principle of knowing. The liminal things high energy physics focuses on are neither the objects of positive knowledge nor effects in the formless region of the unknowable, but something in between: examples are ‘limits,’ systematic errors, efficiencies, acceptances and so on, whose investigation and presentation consume a large part of experimental time. Other aspects of high energy physics experiments include their reflexive turn toward self-understanding, toward replacing the ‘care of objects’ with the ‘care of the self’ (Knorr Cetina, 1999: 55ff, 63 ff.). High energy physics experiments may seem remote from particular social interests, but they are not. Apart from the knowledge strategies they exemplify, they illustrate what it might mean to organize global cooperations in a knowledge society outside the realm of large corporations, and to make them work. At the beginning of the twenty-first century, high energy physics experiments are conducted by large global collaborations of up to 2000 physicists and up to 200 physics institutes that work together for approximately 20-30 years, or the better part of the lifetime of a scientist. These collaborations now involve the largest, longest-lasting and presumably best-integrated epistemic groups. As these groups struggle through the stages of the birth and lifetime of a new experiment, they create and illustrate for us many organizational innovations that are relevant to other areas within (the genome project, for example) and outside science. Another field which may seem equally remote but which has been studied is mathematics, the distinct practices of which differ not only from physics but seem unique among all sciences (Heintz, 2000; Merz, 1999).
Finally, a number of authors have turned to studying scientific texts or to studying scientific theories and research results as narratives and discourse structures (see, for example, Bazerman, 1999). This enables them to follow the metaphors and images that scientists use in their theories and descriptions, and import, or translate from, other cultural orders (Martin (1991) and Haraway (1989) can be interpreted in this way). The cultural turn is exemplified here in the choice of data and approach (scientific texts are analyzed as narratives) as well as in the attempt to trace the flow of cultural symbols (metaphors) through different social domains (science on the one hand and other socio-cultural domains on the other).
Standpoint Theory Approaches and Feminist Science Studies
The field has not only experienced a cultural turn, but attracted strong feminist scholars who have developed their own viewpoints. One of the first perspectives that developed was a feminist standpoint theory that was continuous with the field’s earliest formulations of a sociology of knowledge and was also loosely based on Lukács’ theories of class consciousness and reification (1971). What the feminist version of a standpoint theory added, then and now, to the earlier formulations, is a critical edge derived from exploring Western science as a realm of endeavor dominated by white males from privileged backgrounds. Standpoint theory made plausible that science is marked by and perhaps locked into the perspectives of the categories of people who produce it—white males, who are also the oppressors in the sex/gender structure of Western societies. Feminists argued for the privileged perspective of women in recognizing gender bias. Women, the oppressed, feel the constraints of the structures which dominate them, and hence are more likely to identify these constraining forces; while those who are not oppressed (white males) do not notice such constraints and are blind to their own perspectival biases (Hartsock, 1983; Sismondo, 1996). Later feminist authors have upheld perspectivalism, but have backed away from interpretations of standpoint theory that suggest that there can be a single true story about reality. In her book The Science Question in Feminism (1986) Harding acknowledges that there are many forms of oppression, and hence there must be many privileged perspectives.
Haraway (1988) goes a step further in attempting to find some ground for political action while acknowledging the radical contingency of all knowledge claims. She also rejects the possibility of an all-encompassing, objective standpoint, but advocates partial perspectives’ as a positive tool in producing knowledge: the resulting ‘knowledges’ are limited and historically contingent, but they are ‘about’ something. In other words, Haraway advocates a sort of perspectival realism as a remedy for relativism and objectivism, a realism which allows for not just one but many true stories. In a recent book, Harding (1998) also takes a step beyond her earlier works by attempting to work out an anti-essentialist reading of standpoint epistemology. She launches further complaints about the claims to absolute cognitive superiority made by modern technoscience, and discusses the possibility of multicultural knowledges that do not devalue non-Western types of ‘science.’ Going even further in the direction of approaches that are not in themselves feminist, Longino (2002) points the way for a compromise between constructionist ideas, more traditional philosophical concepts and feminist thinking on science.
It should be noted that feminists have produced a wide range of results revealing gender bias in the content of scientific conceptions and theories, going far beyond the questions of the representation of women in particular positions, access to resources and so on (Fox Keller, 1985; Haraway, 1989; Martin, 1991). For example, Martin has shown how masculine vocabularies and conceptions of heroic conquest are inscribed into biologists’ descriptions of sperm and egg activity, and how new views reiterate such trends while ostentatiously undermining an earlier conceptual sexism. While earlier biological conceptions construed the egg as passive and powerless, the newer active egg models transform the passive egg into its opposite, a dangerous female that ‘captures and tethers’ the helpless male sperm—drawing upon cultural models of the female as a witch or a whore (Hess, 1995: 30; Martin, 1991). A different but important line of research focuses on reproductive technologies and on the development of reproductive science (Clarke, 1998). This work shows, for example, how amniocentesis opens up different choices, opportunities and dilemmas for different groups, while the construction of the technology also shifts dramatically between groups (Rapp, 1990).
The new sociology of science and technology is a young field: it started off with a bang in the 1970s when it was reworked and revamped as a sociology of scientific knowledge. At that point, the field had a point of intense focus: the question of the social conditioning of knowledge in the natural sciences and mathematics. Since then, the field has expanded in the various directions just discussed, and more. It is now diverse, but it still retains a particular orientation that distinguishes it from other specialties that address knowledge, notably knowledge management, the sociology of information and transformation theories that point toward a knowledge society. The sociology of science and technology retains its orientation towards studying knowledge not just externally, but internally; not just from the outside with respect to its social structural, occupational and other implications, but from the inside, by including in the studies conducted a level of content of knowledge, of knowledge practices and epistemic relations. The vigorous growth and vitality of science and technology studies in the last decades may have something to do with this insistence on getting inside the domains of expertise examined. Accordingly, when students of science and technology and others who took over a ‘science studies approach’ (identified with the research programs listed) have addressed larger social questions, they have done so in distinctive ways. They have applied science studies concepts and findings to legal battles, pursuing issues of the constructed, networked and translated nature of evidence and techniques in this context; they have also extended the approach to financial markets and economics, to politics, to management and to general questions of classification. They have transferred concepts like that of the laboratory to non-laboratory settings, and they have given their own assessments of what postmodernity might mean. Here are some examples.
One of the strongest consequences of recent studies of science and technology is that it has raised the awareness for the role of non-human objects (animals, other natural objects and technologies) in society, students of science and technology attributing a more active role to such objects than sociologists in general. For example, the actor-network approach treats them as ‘actors’ according to a semiotic and grammatical (rather than Weberian or phenomenological) definition of action. Presumably, any object can fill the subject’s role in a linguistic sentence structure. Analogously, non-human actors (technologies, viruses, scallops etc.) can have agency (for example, have power, provoke effects, create resistance) in scientific and technological settings (for examples of such treatments of non-human actors see Callon, 1986; Latour, 1993; Latour and Johnson, 1988; Pickering, 1995). The discussion resonates strongly with certain traditions of sociological and philosophical research, for example with claims in ecology for the need to reintroduce nature into the social contract (Beck, 1992; Merchant, 1983; Serres, 1990). It also overlaps with a line of work that has grown out of science and technology studies and problem-solving and has come to be known as ‘work place studies,’ that is, investigations of the usage of (electronic information) technology by workers at their place of work (Suchman, 1987). A further point of contact is given in studies of ‘virtual society,’ electronically connected communities and their relationship to technologies (Pels et al., 2002). But as a rule, sociologists have defined the social world as the domain of human interaction, human institutions, human rationality, human life. As Luckmann pointed out in 1970, we take it for granted that social reality is the world of human affairs, exclusively. Yet why should we take this for granted? As Luckmann also argued, the boundary we assume between the human, social and the non-human, non-social is not an essential structure of the life-world. Latour (1993) has even proposed that what we are taking to be modernity is characterized by a systematic refusal to recognize and incorporate in our thinking the hybrid forms of organization that are neither purely subject nor object. It is the hidden working of these hybrid forms behind our back that makes it possible for us to continue to conceive of ourselves as pure subjects and of society as purely human. Knorr Cetina (1997) has maintained that if the social is not limited to the human, we can begin to develop an analysis of the ways in which major classes of individuals (for example, scientists) have tied themselves to object worlds which situate and stabilize selves, define individual identity just as much as communities and families used to do, and which promote forms of sociality (forms of binding self and other) that supplement the human forms of sociality studied by social scientists. Objects may also be the risk winners of the relationship risks which many authors find inherent in contemporary human relations. The strongest claim is that what lies ahead is a ‘postsocial’ environment where objects displace human beings as relationship partners and embedding contexts, or increasingly mediate human relationships, making the latter dependent on the former. The ‘question of objects’ poses a challenge to sociology comparable to that which globalization does. To develop an understanding of global society we need new concepts that point away from conceptions of nation-state societies with which sociological thinking has been bound up in the past. To learn to understand the role of objects in contemporary society we need new theoretical frameworks and conceptual tools, and we must let go of the exclusively human concept of social reality that Luckmann attacked.
There are other, less controversial extensions of science and technology studies to the larger social context. For example, laboratory studies have developed into a laboratory studies perspective that takes its lead from family resemblances between laboratories and other physical and virtual spaces. In this case the notion of a laboratory, and the concepts emerging from laboratory studies, are transferred to areas that are not literally laboratories but can be seen as spaces of knowledge. As a perspective, the laboratory studies approach simply brings into view matters with which students of laboratories have concerned themselves in other areas. For example, processes of reality construction are ongoing occurrences in a variety of settings. This is the sense in which the industrial factory itself begins to resemble a laboratory as a site for invention and intervention in which new realities are created (Miller and O’Leary, 1994). We can illustrate this use of the laboratory approach by an example from transsexual research and one pertaining to the factory. In the first case the new sex of a person who desires to have a different sex from the one he or she is born with is seen as a ‘fact’ that is being constructed in a laboratory that is constituted by the different locations and stages of the treatment that transsexuals undergo (Hirschauer, 1991). The laboratory approach sheds light on the multiple arenas that make up the lab, on the constructive and transformative work involved, and on the heterogeneity of the frameworks of knowledge applied to transsexuality. In the second case, that of the factory, Japan is brought to Illinois, as it were, in the attempt to reconfigure the American factory in the image of global factory modernization, by redesigning shop floors, recalculating new spatial orderings of production, and molding the worker according to the ideal of a ‘new economic citizenship’ for plant personnel. Here the factory itself becomes a laboratory for acting upon its own assemblage of locales and internal relations and for refashioning the person who participates in the assemblage (Miller and O’Leary, 1994).
A third research focus where the science and technology studies approach is extended to the wider society concerns the relationship between science and law. Lynch has studied the Iran Contra Hearings and Lynch, Jasanoff and others have examined the OJ Simpson trial in relation to DNA fingerprinting and the failure of science to convince the jury in the trial (see the works collected in Lynch and Jasanoff, 1998). For example, Lynch (1998: 853, 855) reviewed arguments about DNA profiling between the prosecution and the ‘dream team’ of defense lawyers in the OJ Simpson trial, arguing that the defense lawyers’ motions were based on rebuttals to the prosecution’s ‘realist’ claims about the inherent workings of natural and technological processes that paralleled those of constructionists’ rebuttals to realist claims about science in general. The motions provided an impressive inventory of uncertainties and contingencies, relating them to political, financial, ideological and career interests of those who collected and analyzed the evidence. The rebuttals rested on exposing sources of uncertainty and contingencies that remained unexamined by official inquiries and hidden from the defense. Jasanoff’s work supplements these findings by conceptualizing trials as arenas in which visual authority has to be created and defended. Jasanoff (1998: 713) argues that it is part of the judge’s role to construct whose vision will be authorized in trials as expert, and in what circumstances lay vision can take precedence over expert sight.
As these examples show, the sociology of science and technology of today brings its approach to a variety of societal questions where science, technology and technical expertise, or what is sometimes perceived to lie at the center of these domains, the non-human and objectual, are key elements. It also at times brings its approach to bear on questions that do not focus on these elements, but are of fundamental theoretical concern, for example the question of organization (Vaughan, 1996), or that of classification and its consequences (see, for example, Bowker and Star, 1999). Over the past few decades, the field has transformed itself slowly from the pure sociology of science Merton had in mind to the discipline of science and technology studies, which seems at times on the verge of transforming itself further into what is purely an approach (or perhaps a confederation of approaches) applicable to many of the dazzling facets of contemporary life. Since science, technology and knowledge are always ‘just around the corner’ in contemporary societies, and questions pertaining to research, analysis and information appear to be implicated in most social institutions in the environment in which we live, the field is not likely to lose its footing in the process.