The Reflexive Turn: ‘Constructing’ Nature and Society

By the early 1980s the followers of the ‘strong programme’ in the sociology of science, and related approaches, were facing criticisms from a new direction. Previously enmeshed in debate with realists and rationalists over their relativism about scientific knowledge, they were now criticized for their (supposedly) unquestioning realist assumptions about society.

However, a more fundamental line of criticism noted the inconsistency involved in the strong programme’s systematic scepticism about the reality claims made by natural scientists, while accepting as ‘real’ and ‘given’ the sociological variables in terms of which scientific controversies were to be explained. Consistency could be restored only by ‘reflexively’ extending scepticism to the sociologist’s own explanatory concepts and methods. Sociologists were now not only to see nature as a socio-cultural construct but also to ‘problematize’ the individual and collective actors, their interests and power relations, alliances and so on which had so far been drawn upon as explanatory variables (see, for examples, Gilbert and Mulkay 1984; Woolgar 1988; Latour and Woolgar 1979, 1986; Latour 1987, 1993). Both social and natural worlds were now to be treated as ‘constructs’. (‘But of what and by whom?’ one might ask.)

This reflexive turn in the sociology of science was connected to a wider debate in social theory between advocates of post-modernism (see Chapter 10) and others about how to characterize contemporary institutions and social processes.

The work of Ulrich Beck in Germany shared this sense of epochal change, but he understood it as a shift within ‘modernity’, from an earlier ‘simple’ modernity towards a ‘reflexively’ modern future. For Beck, ‘reflexivity’ referred to the way modern institutions, most especially science and technology, were, through their own development, putting themselves in jeopardy. Beck’s Risk Society (not published in English until 1992) emphasized the significance of a new order of risks associated with new, large-scale technologies such as nuclear power, the chemical industry and genetic manipulation. These risks were literally incalculable, potentially universal in their scope, and exceeded the capacities of private and public sector to insure or guarantee security. In the face of these risks, the scepticism about particular knowledge-claims which had always been central to scientific method was now being generalized to science itself. A new ‘sub­politics’ was emerging, in which the authority of science was under question, with the hopeful prospect of wider democratic participation in scientific and technical decision­making (for evaluations of Beck’s Risk Society, see Rustin 1994; Benton 1997).

However, for all Beck’s use of the term ‘reflexivity’, his work remains realist about the ecological risks he describes. The work of other leading advocates of the reflexive turn has been strongly anti-realist (see, for example, Wynne 1996), or deeply ambiguous. The very influential work of the French philosopher-sociologist Bruno Latour and his associates comes into this latter category. This work, known as ‘actor-network’ theory, shares with the feminist approaches we discussed above an attempt to overcome rigid dichotomies between nature and human agency, and subject and object of knowledge.

At the heart of actor-network theory is this intention to break free of the dualism which separates off, or ‘purifies, society from nature. The ethnographic study of laboratory life (see Latour and Woolgar 1979, 1986) is what enables this break, because it is here more than anywhere that we can observe the production, or generation, of ever more ‘objects’: chemical substances, micro-organisms, instruments, devices and so on.

Essentially, R. & D. is an industrial affair (three quarters is carried out inside firms) financed out of tax money (amounting to 47 per cent in the US). This is the first massive transfer of interest: scientists have succeeded only insofar as they have coupled their fate with industry, and/or that industry has coupled its fate to the state’s... by and large, technoscience is part of a war machine and should be studied as such.

(p. 170)

Latour's attempt to overcome the dualistic separation of nature and culture focuses on the way modern science and technology (‘technoscience', as he calls it) is constantly proliferating new ‘objects', as candidates for subsequent ratification, or ‘stabilization', as accepted, taken-for-granted realities, pieces of equipment, measuring devices and so on.

Successful science and engineering consists in deploying the whole gamut of persuasive, rhetorical and material resources to extend and maintain alliances beyond the laboratory and the scientific journals. Crucially, these alliances are composed out of previously unconnected elements which cut across conventional divisions - they are heterogeneous ‘networks of actants', which include human groups, animals, pieces of equipment, synthetic materials, measuring devices and so on. The term ‘actant' is used to include the full range of elements which play a part in the establishment and consolidation of the network, irrespective of our subsequent wish to allocate them to one side or the other of the nature-society dichotomy.

It is only by virtue of a subsequent practice of what Latour calls ‘purification' that nature and society are constructed as separate, unconnected domains. At this point, Latour is establishing a double distance between his approach and that of more orthodox sociology of science. On the one hand, he brings non-human beings and objects into the account of the processes of science and technology. To this extent (and with some ambiguities) his position is realist about the objects of scientific investigation: ‘It is hard to reduce the entire cosmos to a grand narrative, the physics of subatomic particles to a text, subway systems to rhetorical devices, all social structures to discourse' (Latour 1993: 64).

But this leads to a second departure from more traditional sociologies of science. If our lives are lived through participation in collectives, which are, in turn, made up of heterogeneous networks of human and non-human elements, and if society and nature are merely secondary, ‘purified' constructs, it follows that we can't use ‘society' to explain science. What we have constructed as ‘pure' society is a consequence of scientific and technical practices, and so cannot be used to explain them. This conclusion he calls, following Michel Callon, ‘the principle of generalized symmetry' (Latour 1993: 94-6; Callon 1986; see also the exchanges between Collins and Yearley (1992, b) and Callon and Latour (1992) in Pickering 1992):

Thus it [the strong programme - TB] is asymmetrical not because it separates ideology and science, as epistemologists do, but because it brackets off Nature and makes the ‘Society' pole carry the full weight of explanation. Constructivist where Nature is concerned, it is realistic about society.

(Latour 1993: 94)

Latour's proposal, then, is to treat both society and nature as constructs which stand in need of explanation, and to place the weight of explanation on the centre ‘between' these poles - that is, on the networks of actants, and the processes by which they get put together, extended and stabilized. However, it is not at all clear that these concepts are up to such a demanding task. For one thing, Latour's key concepts for defining ‘actants' are those of ‘hybridity', ‘quasi-object’ ‘quasi-subject' and so on. These terms get such meaning as they have only in terms of the prior understanding of what ‘subjects’ ‘objects' and the ‘pure' elements of the ‘hybrid' are. Latour contravenes his own methodology in the very act of defining his most basic ideas.

Moreover, Latour's desire to find terms which link humans to non-humans into seamless networks leads to metaphorical excesses which carry little conviction. An example is his account of the way the windmill succeeded in forming a more powerful network than the pestle and mortar (Latour 1987: 129), binding together corn, machinery, bread and the wind. He writes of ‘translation' of the ‘interest' of the wind, and ‘complicated negotiations' which have to be conducted with it to secure its alliance. Far from representing the windmill as something irreducible either to nature or society, Latour's language reduces the wind to a conscious quasi-human interlocutor, with its own interests and capacity for compromise. What is of value in the account is the recognition of the independent causal power of the wind, and its role in shaping the range of possible human responses to this, but this insight is put at risk by the resort to anthropomorphism.

Similarly, if we follow his argument about the extension and stabilization of ‘networks' beyond the laboratory, we find Latour, as in the above quotation on R&D from Science in Action, explaining successful science in terms of the ability of scientists to establish alliances with states, businesses, the military and so on. As he puts it, scientists and engineers can succeed only in so far as they interest

more powerful groups that have already solved the same problem on a larger scale.

That is, groups that have learned how to interest everyone in some issues, to keep them in line, to discipline them, to make them obey; groups for which money is not a problem and that are constantly on the look-out for new unexpected allies that can make a difference in their own struggle.

(Latour 1987: 169)

So, scientific success, which for him includes the power to define reality itself, depends on powerful groups, money, alliances, states, military institutions, businesses... This certainly looks very much like an explanation in terms of the forbidden ‘society'!

So, despite his rhetoric, it seems that in practice Latour has not moved far from the sociologists of science whose approach he criticizes. Even what seems most distinctive - his inclusion of the activity of non-human ‘things' in the processes of science - is compromised by ambiguity and contradiction. What exactly is the status of the ‘objects' (‘quasi-objects’ ‘hybrids' and so on) which science ‘generates'? In his Pasteurization of France (1988) we read: ‘Did the microbe exist before Pasteur? From the practical point of view - I say practical, not theoretical - it did not' (p. 80). In Science in Action (1987) he writes of the way laboratories ‘generate’ ‘new’ objects, but that at first these ‘new objects' are no more than their responses to certain laboratory ‘trials’. He goes on:

This situation, however, does not last. New objects become things: ‘somatostatin’, ‘polonium’, ‘anaerobic microbes’, ‘transfinite numbers’, ‘double helix’, or ‘Eagle computers’, things isolated from the laboratory conditions that shaped them, things with a name that now seem independent from the trials in which they proved their mettle.

(PP. 90-1)

In some places he expresses scepticism about the realism of science when networks are consolidated, and controversy ended, but elsewhere he follows scientists themselves in shifting ‘modality’ from relativist non-realism in periods of controversy to conformist realism when controversy is settled. But can any coherent sense be made of a world in which ‘objects’ simply come into existence and depart again in a ballet perfectly choreographed with the power struggles of scientists? If success in defining reality is just a matter of superior strategic skills, owing nothing to either rational argument or what the world is actually like, why should we meekly accept the outcome? After all, a new power might be installed tomorrow!

Latour’s account of technoscience presents it as a struggle to recruit powerful allies and material resources, as ‘part of a war machine’, and as an ‘intolerable source of pathology’:

[W]e find the myth of reason and science unacceptable, intolerable, even immoral. We are no longer, alas, at the end of the nineteenth century, that most beautiful of centuries, but at the end of the twentieth, and a major source of pathology and mortality is reason itself - its works, its pomps, and its armaments.

(Latour 1988: 149)

Much of today’s ‘big science’ conforms well to this description, but even as a description it is selective and one-sided. Not only have many courageous scientists sacrificed their careers out of a sense of social responsibility, but also there is a long tradition of scientific dissent, while much of what has come to be called ‘environmental science’ has been open to mobilization by oppositional social movements in their resistance to the destructive and exploitative impositions of the wealthy and powerful. Most seriously, however, the cynicism of Latour’s demonic view of contemporary technoscience leads him to an outright rejection of reason itself. But if we abandon the means of criticizing power exercised in the name of science, or military force posing as reason, if we have no vision of an alternative practice of knowledge creation, what then does Latour’s strong language amount to? ‘Yes - it’s terrible, but that’s just the way things are’? The positive visions of an alternative science, respectful of its objects, democratically accountable to its wider citizenry, and open to full participation on the part of previously excluded or marginalized human groups, which are offered by the feminist and other radical critics of science, provide an alternative to this apparent defeatism.