Latour, Bruno. Science in Action. Cambridge, Massachusetts: Harvard University Press, 1987.

"[S]urprisingly few people have penetrated from the outside the inner workings of science and technology, and then got out of it to explain to the outsider how it all works." (15)  Through engaging examples and carefully built arguments Latour advances the theory of science as a group of networks tied together by those facts that are accepted as true by the network-- the "black boxes" which are, in effect, nodes of the network. Scientists in action compete for truth, money and prestige, won through alliances made by scientists, literature, labs, machines, communities and cities.
Latour begins with the time pressures and rival competitors faced by Jim Watson and Francis Crick as they researched the DNA structure, and the Eagle/Eclipse team as they tried to build a computer using new the new PAL chip.  As he traces the history of these events, Latour  introduces a character, the double-headed Roman god Janus, who speaks throughout the work to the iterative process that characterizes Latour's understanding of science:  "Once the model works, people will be convinced" / "The model works when all the relevant people are convinced."  (10)   Latour also puts the point this way:  "Facts and machines in the making are always under-determined.  Some little thing is always missing to close the black box once and for all."  (13)  Latour chooses moments when facts and processes, such as DNA and computers, were not so firmly "packed," and shows through various methods, how they become closed,   irrefutable.  Governing and directing his work is the explication of rules of methods and principles.  Rules of method are the "a priori decisions [which] should be made in order to consider all of the empirical facts provided by the specialized disciplines as being part of the domain of science, technology and society.'"  Principles are his own understanding of the facts at hand. (17) How, then, does a fact pass into a crystalline state?
Ideas and their proponents need a web of scientists to use them, cite them, build second and third generation articles on them.   The more technical the literature is, the more social -- the more embedded in the reputations of reputable scientists -- and the harder it is for dissenters to make headway against ideas.  Not only must ideas be cited in articles of reputable journals, and cited again, they must also acquire laboratories to support them.  Nature is not so much tested in the laboratory as decided upon in the laboratory. So a dense web of literature and laboratory resources grows up as a hedge around an idea.  "Everything else being equal, the winner is the one with the bigger laboratory or the better article." (103)

In chapters three and four Latour offers two central concepts in his work:  translation, and the black box.  An idea is established through socialization.  How, then, do inventors and innovators control the participation of others?  They translate the needs of others into resources that advance their own interests.  For instance, an investor who wants the best possible factory flooring might be asked to invest in wood laminate in order to create new possibilities for materials that might answer her wishes.  Latour lists a host of translation strategies and tactics to form alliances. A black box, then, is a series of alliances and chain of associations that has become an automated. (130-131) When allies act as a unified whole, they create black boxes, irrefutable facts and objects, which in their turn, become one more piece of equipment in a lab.  Insiders and outsiders to science and technology are always shifting.  Science is not so much done or diffused as  translated into the terms that touch on the interest of various social groups. Latour sets the discussion of science, technology and science on its ear by declaring no such thing can exist. They are indivisible elements, each the media in which the others live, move and have their being. At this point he has built his model, and chapters five and six simply show the consequences.

The concluding chapters offer a look at specific case studies where reason itself is called into question, and how the centers of meaning are established through accumulation of data. Moving outside of Western culture, Latour cites chains of reasoning that seem, initially, to be nonsensical or insane.  Simply by adding more information about the context in which the interactions take place, Latour appears to reframe reason.  (In fact, he simply gives enough depth for Western reasoning to swim again).  Again, moving from one culture to another, Latour demonstrates how some places become powerful centers of calculation; apparently, cities or nations that do the best job of compiling information about far-away places are likewise best able to impose their own civilization on others.

What are the consequences of Latour's thinking?  First, he allows scientists to lower their stentorian tones and speak with uncertainty. That offer may not be taken up, but in highlighting the alliances that establish truth, he may teach historians to look for the reasons that seemed reasonable to the network where "new facts" were established.  Pinpointing the moment the black box stabilized may likewise be useful. Second, he offers a vision of science that is not so much blocked by walls as it is a wallflower, waiting for the next right partner.  His theory may help address puzzling counterfactual questions, for example, "Why did the Romans invent water organs but not steam engines?" Finally, Latour's two assays into cross-cultural anthropology offer ways to look at translation/diffusion that may be less invested in ethnocentric explanations.