Bruno Latour (1999) Science in Action: How to Follow Scientists and Engineers Through Society, (1987) Cambridge, Mass.: Harvard University Press.

The word 'black box' is used in cybernetics when a piece of machinery or a set of commands are too complex. Latour sees this as the Janus-face of science, on the one side lively and the other side severe: 'science in the making' in contrast to 'ready made science' (1999: 4). In the way, the seemingly impossible task of opening Pandora' black box is made possible by experiencing science and technology at work, not ready-made - in action, 'before the box closes and goes black' (1999: 21). The black box represents hard facts or indisputable evidence. But science has two faces, Latour explains, one that knows and one that does not yet know (1999: 7). He calls for a 'stereophonic rendering of fact-making instead of its monophonic predecessors' (1999: 100). Universal facts are held in contrast to speculations. The method of 'science in action' follows certain rules, for instance: 'we will enter facts and machines while they are in the making; we will carry with us no preconceptions of what constitutes knowledge; we will watch the closure of the black boxes and be careful to distinguish between two contradictory explanations of this closure, one uttered when it is finished, the other while it is being attempted' (1999: 13-5). 

This method is what Latour calls penetrating the inner workings of science and technology from the outside, following controversies on how it works and is made. Thus science should be understood as a practice that is produced both in a social context and as a result of the technical and institutional apparatus. It is a 'disorderly mixture' revealed 'in action' rather than an 'orderly pattern of scientific method and rationality' (1999: 15). A cultural reading of scientific method is thus an interdisciplinary endeavour broadly characterised under the domain of 'science, technology and society' (at the time of writing, Latour worked as a professor in the sociology of innovation). The book charts a series of rules of method and working principles to explore these ideas around 'techno-science'. He says: 'To do such a study it is absolutely necessary never to grant to any fact, to any machine, the magical ability of leaving the narrow networks in which they are produced and along which they circulate.' (1999: 257) 

First of all, the issue of rhetoric is used to demonstrate how certain statements are made more or less believable, whether the claims are considered subjective or objective. In the anatomy of scientific literature, he discovers that if something is controversial, the literature becomes technical: 'when we go from "daily life" to scientific activity, from the man in the street to the men [sic] in the laboratory, from politics to expert opinion, we do not go from noise to quiet, from passion to reason, from heat to cold. We go from controversies to fiercer controversies.[...] More noise, indeed, not less.' (1999: 30) As in legal disputes, he points out that often controversies end in 'technicalities', more dissent leads to more technical literature and scientific papers. Another strategy for veracity is to seek the opinions of experts in the field, those who hold authority neither in person nor by citing former texts or being cited - despite the orthodox view that science stands against rhetoric (1999: 32-3). Latour says 'The power of rhetoric lies in making the dissenter feel lonely' (1999: 44). Scientists employ style not strictly for aesthetic reasons but to demonstrate a logical argument - to make it seem rational often simply but the use the right kind of adjectives. This is fact-writing not fiction-writing (although admittedly the techniques have some similarities). Facts and truths, like the efficiency of machines, are constructed in such a way, not for their intrinsic qualities but for their transformative qualities in later usage (1999: 59). Technoscience is a weak rhetoric that becomes stronger through this fact-making process - this is how beliefs are turned into knowledge perhaps. In a manner of speaking, he is extending literary criticism to technical writing here in the tradition of structuralism and semiotics.  

These texts and the rhetoric employed though are representation of what really took place in the laboratory. It is here that scientists do their work. They use instruments or what Latour calls 'inscription devices' (1999: 68) to provide visual data perhaps - and the implication is that this is instrumental in terms of providing pre-determined results. Scepticism is a useful critical response, as in the case of Cantor when he declared: 'I see it, but I don't believe it' (in Latour, 1999: 90). For the most part, the instrument that produces the visualisations is kept invisible. When the scientist attempts to prove something by saying 'see for yourself', the critical response should include an analysis of how the evidence has been obtained in terms of the mediating process. Latour asks who is speaking: 
'When we are confronted with the instrument, we are attending an "audio-visual" spectacle. There is a visual set of inscriptions produced by the instrument and a verbal commentary uttered by the scientist. We get both together. The effect on conviction is striking, but its cause is mixed because we cannot differentiate what is coming from the thing inscribed, and what is coming from the author.' (1999: 71) 
The response to these conditions should be to build better laboratories, what Latour calls 'counter-laboratories' to take account of this array of constraints (1999: 79). 

Latour uses the example of the development of the diesel engine to make the point that it is collective action that transforms a good idea into a working product (the black box under the bonnet of the vehicle). Diesel, the engineer, was simply the inspiration - 'the motor behind the engine' so to speak (1999: 106) - the catalyst behind a complex process of interactions, itself inspired by earlier discoveries in thermodynamics. Latour explains this in terms of 'translation' to describe: 'the interpretation given by the fact-builders of their interests and that of the people they enrol' (1999: 108). Claims are transformed into facts through this process of translation, partly reinforced by collective action and shared interests (perhaps financial or ideological). The example of motor vehicles is a good one, and subsequent transport policies that serves the interests of the economy despite the environmental damage. Yet, there is nothing inherently meaningful about calling a claim 'absurd' or knowledge 'accurate' (Latour, 199: 205). It is the form of the claim or knowledge that require inspection.

For Latour, the term translation means simultaneously offering new interpretations of these interests and moving things from one place to another. This allows particular issues are tied to larger ones in such a way that 'threatening the former is tantamount to threatening the latter. Subtly woven and carefully thrown, this very fine net can be very useful at keeping groups in its meshes.' (1999: 117; for instance, more efficient transport threatens the economy). Enmeshed in this are issues of attribution of intellectual property (and social responsibility). This partly explains the difficulty of attributing particular discoveries to individual scientists, evident in many of the books I have come across on popular science, keen to elevate particular names into the romantic myth of genius. Deciding who was first to do something is simply explained by the analysis of these processes whereas the question itself is simply a naive question. These 'machiavellian' strategies are employed by scientists and technologists, but clearly by those in the cultural field too (in fact there is long established critical tradition in making these machinations transparent). Translations are used to keep the allies in line. These are interests are further tied together through a 'machine', not a tool, as an assembly of forces greater than that of the tool and independent of the human hand (Latour, 1999: 129). The machine is seen as a system of 'subtle and versatile alliances' acting in unison. It is almost as if a society is created for these ideas and machines: 'In this model, society is simply a medium for different resistances through which ideas and machines travel.' (Latour, 1999: 136) The machine only holds things and people together if the other strategies are in place, and exists in a social context of associations and linkages. 

To guarantee success, the scientist is required to gather resources, talk with authority, convince others and equip laboratories. Success is measured by attention, money and confidence. All this explains how the domains of science, technology and society are kept apart. This is what Latour refers to as a model of 'diffusion' wherein the associations and translations are kept apart artificially. It is not simply that science is determined by technical or social factors but Latour recognises the significant influence of class, the economic infrastructure, capitalism, business, gender, culture (1999: 142). Instead, he proposes that science and society are tied together 'symmetrically', face to face presumably in a janus-like relation of human and non-human resources. This leads Latour to ask who is really doing science, as this is not confined to bearded men in white coats in laboratories. Clearly there is a complex institutional mesh around science, linked to funding structures of research and enterprise, with relations of production in place to protect some of the interests described. Statistics of money spend on research and development programmes as well as education provision are revealing in this connection. Latour explains this stratification of scientists across laboratories in the same country or between countries: 'This asymmetry modifies what is called the visibility of a scientist or a claim' (1999: 166). Success or failure rests on these factors of visibility. 'All laboratories are not equal' before God, says Latour (1999: 166). It is no surprise to learn that budgets remain proportional to amounts of interest generated. Latour concludes that success rests on the 'presence or absence of already aligned interest groups' (1999: 173).