READING GUIDE TO: Latour, B  (1987) Science in Action, Milton Keynes: Open University Press

Dave Harris

 Two useful appendices summarize the approach: (scan of pp 258, 259 of Latour)

APPENDIX 1 Rules of Method

Rule 1 We study science in action and not ready made science or technology; to do so, we either arrive before the facts and machines are blackboxed or we follow the controversies that reopen them. (Introduction)

Rule 2 To determine the objectivity or subjectivity of a claim, the efficiency or perfection of a mechanism, we do not look for their intrinsic qualities but at all the transformations they undergo later in the hands of others. (Chapter 1)

Rule 3 Since the settlement of a controversy is the cause of Nature's representation, not its consequence, we can never use this consequence, Nature, to explain how and why a controversy has been settled. (Chapter 2)

Rule 4 Since the settlement of a controversy is the cause of Society's stability, we cannot use Society to explain how and why a controversy has been settled. We should consider symmetrically the efforts to enrol human and non-human resources. (Chapter 3)

Rule 5 We have to be as undecided as the various actors we follow as to what technoscience is made of; every time an inside/outside divide is built, we should study the two sides simultaneously and make the list, no matter how long and heterogeneous, of those who do the work. (Chapter 4)

Rule 6 Confronted with the accusation of irrationality, we look neither at what rule of logic has been broken, nor at what structure of society could explain the distortion, but to the angle and direction of the observer's displacement, and to the length of the network thus being built. (Chapter 5)

Rule 7 Before attributing any special quality to the mind or to the method of people, let us examine first the many ways through which inscriptions are gathered, combined, tied together and sent back. Only if there is something unexplained once the networks have been studied shall we start to speak of cognitive factors. (Chapter 6)

APPENDIX 2  Principles

First principle The fate of facts and machines is in later users' hands; their qualities are thus a consequence, not a cause, of a collective action. (Chapter 1)

Second principle Scientists and engineers speak in the name of new allies that they have shaped and enrolled; representatives among other represen­tatives, they add these unexpected resources to tip the balance of force in their favour. (Chapter 2)

Third principle We are never confronted with science, technology and society, but with a gamut of weaker and stronger associations; thus understanding what facts and machines are is the same task as understanding who the people are. (Chapter 3)

Fourth principle The more science and technology have an esoteric content the further they extend outside; thus, `science and technology' is only a sub­set of technoscience. (Chapter 4)

Fifth principle Irrationality is always an accusation made by someone building a network over someone else who stands in the way; thus, there is no Great Divide between minds, but only shorter and longer networks; harder facts are not the rule but the exception, since they are needed only in a very few cases to displace others on a large scale out of their usual ways. (Chapter 5)

Sixth principle History of technoscience is in a large part the history of the resources scattered along networks to accelerate the mobility, faithfulness, combination and cohesion of traces that make action at a distance possible. (Chapter 6)

 

My commentary:

 Rules of Method

 Rule one It is science in action that is studied, before things are solidified and black boxed. The idea is to trace the acceptance of initial conclusions and propositions as fact, either by doing history, or by reopening controversies. The examples in Chapter 1 include the tentative steps leading up to the acceptance of the double helix model of DNA. The uncertainties and contingencies included not only the decision to share knowledge with other teams, but also a very interesting case where the alternative model of Linus Pauling was examined. At first it looked as if the Pauling model would triumph, since it emerged from a world-famous scientist and a large laboratory, whereas Watson and Crick were not really supposed to be working on the structure of DNA at all .  Watson and Crick found the confidence from somewhere to examine his model critically and found some errors -- unbelievably. This spurred them on to continue with their model and also to rush to solve the remaining problems before Pauling discovered his mistake. The other example turns on the eventual emergence of a particular kind of personal computer, with an number of competing interests, including rival companies and skeptics within the company, who had to be persuaded that the new model would work. Finally, almost accidentally, the last few major bugs were fixed sufficiently for the new model to play a game, which convinced the skeptics. Modern scientists, working in the 1980s simply took for granted that Watson and Crick's model was an accurate one, and that the computer they were using to display the model was reliable -- both a model and the computer had been black boxed.

Rule two. It is not so much the intrinsic logical or deductive qualities other scientific claim that means it gets accepted, but the 'transformations they undergo later in hands of others' (258)[ same as the specialist term? see below] . The examples in Chapter 2 consist of a marvellously detailed analysis of the argumentative twists and turns associated with original scientific claims, especially one that claims to have discovered a particular hormone associated with human growth. Latour traces the complex interweaving of claims and counter-claims, ways of attacking counter-claims, further ways of gaining support for original claims and so on. Additional references play a major part in this struggle, as do various kinds of presentations of data, which themselves may lead to further controversies, claims and counter-claims.

In the end, a successful claim is buttressed by whole arrays of supporting resources, and a complex stratified structure of argument. Some excellent examples are given of processes referred to as 'stacking' - an interesting practical version of induction, where some observations of biological specimens are stacked to lead to conclusions about entire species (see 51); 'staging and framing' -- where ideal readers and their possible objections are addressed and persuaded;  'captation'--  a way of steering obstinate readers towards particular conclusions by leading them down various argumentative routes.

The whole process might well be described as 'fact writing', essentially the same process as fiction writing but with more limited options. Ultimately, only a few people are left to thoroughly arbitrate the claims, most possible critics or dissenters would have given up or gone along with the argument. Of course, lots of texts are never even read at all. Scientific writing is another kind of rhetoric, more admired because of its dispassionate style, and able to 'mobilize on one spot more resources than  [its rivals]' (61). Because facticity and objectivity depend on connections with other resources, Latour argues that successful science is 'extremely' social  (62). It is these connections to authorities, experiments, definitions and previous work that makes it almost impossible to disagree with a successful claim. [Is there a pragmatic way of limiting the long list we are supposed to draw up as in Rule 5? Let scientists draw one up until they are convinced of the objectivity of the claim? Otherwise the list would be endless – tracing all the references and background to a single Sports Science article would be prohibitive! Why should we want to draw up any such list that is longer than what the scientists themselves would draw anyway?]

Rule three. It is not that easy to argue from what 'nature' is really like to settle any controversy. In practice, what seems natural and real is what emerges at the end of long debates and disputes, when some consensus has been arrived at, or, rather, dissenters have been driven off or given up. Incidentally, this forgotten [repressed?] creative process is what Latour calls the 'tacit knowledge' of the scientist. This is a long process, and in natural science, much depends on what goes on in laboratories. Latour's examples turn on controversies that went very deep indeed into the organization of laboratories, the calibration of instruments, the purity of samples, very detailed attempts to control variables and so on. He insists that this should really be seen as an element in the resolution of disputes or trials of strength, where laboratories serve to discourage dissenters: the only way to respond to laboratory evidence is to develop bigger and better laboratories. An instrument here is defined as something that produces data connected to displays. The relation between instruments and their 'spokesmen'  (those who interpret the data and attempt to make it as objective as possible) is crucial. Scientists interpreting data strive to be objective spokespersons of instruments, while the dissenters attempt to argue that some subjective matters have been introduced into the interpretation. Anyone shown to be subjective loses the strength of support gathered by a scientific writing and laboratory experiments. There are many interesting examples of how particular laboratory demonstrations or experiments or even canonical pieces of work became decisive allies of particular point of view, including Mead's work on Samoan adolescents, which became a decisive issue for those arguing for cultural versus biological determinism of sexual behaviour. There are also political struggles where spokesmen claim to be speaking objectively for their constituents.

This is actually an example of Latour means by the term 'actant' --  'whoever and whatever is represented [by a spokesman]' (84) and it can be a machine as well as a person. [This is unexceptional, but is not the root and branch challenge to human subjectivity that I had expected. I can see that Latour wants to represent science as wholly and irredeemably rhetorical, and thus he has to take on what looks like the objective elements. So he denies their objectivity by saying they have no meaning until they are interpreted and woven into controversies. But this is not saying, so far at least, that objects and machines are fully subjective as humans are. Machines cannot be spokespersons for and on behalf of themselves.

Of course, people also need spokespersons to become active in controversies - much work is simply ignored. This is also illustrated by 'publication bias', where only positive results tend to be reported, both by the companies that sponsor the research and by the scientific journals, even the Lancet]

Occasionally instruments and their readings are interpreted to suggest that a whole new object has been discovered. Unusual readings or findings raise a tentative claim to have discovered a new object, and again these claims are black boxed in due course. One example was the discovery of polonium, which originally appeared as a set of interesting findings in the laboratory of the Curies. In this early stage, 'a "thing" is a score list for a series of trials' (89).

The chapter ends with the points summarized in rule three about nature. Scientists sometimes want to claim that nature will resolve disputes in the end. At the same time though, they face the other way [Janus figures appear throughout], because much of their activity is designed to find out what actually is natural. Only after a great deal of activity does this become apparent. In this way, scientists face both ways, claiming to be both realists and relativists. Latour does not want to argue for one option rather than the other, but to agree with the scientists themselves -- that once a controversy is settled, we can call the result natural and real. However, all the time controversies persist, what is natural and real is in doubt. Nor is it any good to use current science to arrange past scientific experiments in terms of their correspondence with what is now thought of as real and natural. This is merely 'Whig history, that is, a history that crowns the winners, calling them the best... and... the losers... simply... wrong' (100).  [In other words, it is teleological, or historicist]. The double stance toward nature is also used to fight off critics, invoking the realist stance against relativists, and occasionally vice versa  [possibly when accused of being naive positivists . This might happen with sports science -- on one hand, approaches are simply either right or wrong, but on the other, sports scientists congratulate themselves on finding ingenious ways to define and operationalize particular matters, and maybe even to manage controversies].

Rule 4 It is a problem to decide who actually qualifies for the title of scientist -- is it those few who work in leading laboratories, or the much wider group of people who keep them there, including those senior scientists who travel the world getting funds and licences?  [This is going to be one of those distinctions that Latour simply sidesteps, by refusing to prioritize or organize them into a hierarchy instead of a network]. The latter group is important to permit the former to operate. When we look at where scientists or actually employed and who pays for them, we find only a few are engaged in experimental laboratory work, and that the biggest budgets are found in defence and health. Latour says that these tend to have limitless budgets so it is a good idea for any scientific programme to get connected to them if it can. Scientists themselves are not interested in these links, and don't understand anything about society, certainly not in the way that sociologists understand it  -- hence the absence from the book of any terms like class or gender. What 'society' amounts to is a pragmatic matter for the scientists themselves, and they don't really understand its extent until they can point to a stabilized network of contacts. Thus both society and nature are sometimes seen as the causes of scientific activity, but they are really the results of extensive activity stabilized into a network.

Rule 5 Chapter 5 has a very interesting discussion on relativism with some excellent examples including the Azande and the Trobriand Islanders. It is easy to demonstrate that the simple opposition between Western thought as rational, and everyone else's as limited and partly irrational, will not hold water. Western thinking and practices are every bit as irrational as Azande witchcraft, and vice versa. Both systems of belief and science result from a network of claims, problem-solving and activity developed according to an everyday logic - a 'sociologic'.  [Here and elsewhere we find echoes of American pragmatism both as a way of organizing ordinary beliefs and of developing scientific knowledge]. There are important differences between science and belief, however, seen best when scientific networks encounter nonscientific ones. There are a number of ways in which this can happen, and Latour uses terms like the angle, direction or scale of the encounter. [There are also hints of Schutz's work on different notions of reality between 'strangers' and residents, and between scientists and others.] Latour points out that scientists go off to investigate other social lives, not to settle among them and maintain their network of belief, but to return back to their home countries and strengthen and extend scientific networks. There is also a hint, developed fully in Chapter 6 of the greater explanatory power of scientific networks.

Rules 6 and 7 There is no difference between the way that ‘savages’ think and the way that scientists think, no ‘Great Divide’. Scientific networks and belief systems are elaborated and extended in the same way, buttressing claims, managing crises and refuting rivals using the same pragmatic procedures – ‘sociologic’. Science is organized differently though, since it operates with much more data, gathered systematically and then processed, abstracted and rationalized into systems. Science operates just like bureaucracies do. Back at the metropolitan centres in Europe, people are despatched to discover data about other people and bring it back for processing and addition (this is like capital accumulation but Latour says the term capitalism is confusing and ambiguous). A large interlocking network is required, of scientists but also politicians, explorers, rulers etc. Once collected or imported, data is subject to further processing, abstraction (like the ‘stacking’ mentioned earlier), and is ideally mathematicised. Science then becomes specialist and powerful, capable of considerable application – but only if the real world is transformed first (tidied, rationalized, made to resemble the data of science). This process ‘hardens’ facts (surrounds them with so much support that they are irrefutable).

 

 Overall, this is not quite as strange and alien as it looks. I think Latour mystifies it a bit, possibly for good reasons: he is wary of using conventional terms such as  'abstract' to describe scientific theory, although he is content to rely on that word as a verb in order to describe the explanatory power of scientific theory. The same goes for words like 'transformation', which is not much more than the process of objectifying complex reality so as to make it compatible with the formulae of science. I suppose both of these words indicate that scientific theory doesn't just describe worlds, but colonizes them (although Latour would not like the political and critical sense of that term).

 As for the famous bits about agencies and networks, I think this is not much more than saying that individual action always takes place within collective action. Obviously, the efforts of others are crucial for understanding the activities of individual scientists. Those others include not only instrument makers and technicians, but collective activity going on in universities, funding bodies, government departments and the like. Instruments and machines become actants because they condense a large network inside them, so to speak? These networks are so interconnected that it makes no sense to describe events as  'theoretical',  'political',  'geographical', or whatever, since all of these aspects are involved  (the example here is a rather good one -- initial explorations of Pacific islands by the French government in order to map and possibly settle particular areas of the Pacific Rim. The more familiar case of Darwin's voyage on the Beagle might also serve. A series of familiar social networks got Darwin on board, including the social connections between his father and various Government agencies, and it was also essential that he cope socially as a gentleman companion for the Captain.  The voyage itself was launched for several interlocking reasons, including the Admiralty's desire to chart the Pacific, commercial interests wanting to establish trade with South America, and the Church of England wanting to establish missionary outposts. To achieve the latter, the Beagle carried a missionary and three natives of Tierra Del Fuego who had been 'brought to England' [actually abducted, originally as hostages] several years before and Anglicised: they were to be returned to their homeland to set up a missionary station. Darwin says that returning the Fuegians was the main motive for the journey as far as the captain was concerned. Darwin himself took several risks in his adventures and might well have perished once or twice. The implications for science would have been very interesting!).

The notion of a network is actually rather like that of the rhizome ( in Deleuze, as Latour himself says), extending in different areas and dimensions in a kind of practical problem-solving way -- a  'sociologic'. Theoretical science, social science, and religious beliefs are identical in this sense -- they're all cobbled together as they go, in order to extend themselves and defend themselves against rivals. We're not far away from the notion of bricolage, surely, with a possible role for particularly gifted bricoleurs?

 Of course science is organized and systematized, and this is the secret of its huge explanatory power. Experiences are turned into data, and this data is 'mobile', that is it can be imported back to the centres of networks, abstracted and managed, and slotted into systems of classification of various kinds. Cartography and taxonomy are the bases of the growth of science in Europe. There is a power relation involved, in that, usually, data moves into centres and not the other way around.

 This is an area where Latour is occasionally evasive.  For example, he says that he will not use sociological terms such as class or power because that is not the ways that scientists themselves understand the world, and he is simply trying to map their understandings -- Science in Action. Yet there other reasons as well -- sociological theory is a form of science itself, and he would not wish to privilege one form in order to understand another. The implication is that his concepts, such as network, are more powerful than either, or at least more general. The usual lack of reflexive critique is applicable here -- has Latour simply been cobbling together bits of history, scientific practice, observation and sociological terminology in order to develop ANT?

 De Certeau’s critique of Foucault’s use of history might apply here: Latour is displaying such a grasp of historical and other detail that it is impossible to refute him. His decision to explore the esoteric realm of science also leaves many sociologists without any way to disagree.

 More mundanely, many of the examples are historical, no doubt to help us understand before things got really technical. But is modern science like that? Does it still do simple induction? Isn’t there another stage where mathematicised ‘facts’ are themselves manipulated and never actually ‘applied’? To use Lyotard’s terms, Latour is describing modernist science where it all still ran according to the performance criteria: now it is a mere game is it easier to extend networks? If it is now a research programmes, doesn’t that make the politics of universities more prominent?

 At times he seems apologetic towards the use of power, seeing it as necessary to consolidate networks, inevitable, or so well established that it is irreversible. The links between knowledge and power need be clarified though, unless he really is doing a simple description of the world-view and activities of natural scientists, who could not be expected to want to clarify them.

 index page