Mol 139-40: "'Network' was [a] term mobilized in the early eighties for understanding how science might be geographically situated. The practices of science are not confined to a single site, and yet the old idea that science is universal doesn't hold either, for it skips over the fact that we are dealing with practices. Practices are not everywhere: they are somewhere."
"Scientific experiments may [...] give the same results in Ghana as they do in London. But this is only the case if the laboratory in Ghana is equipped with the same instruments as that in London and staffed with equally well-trained people. As soon as there is a power cut in either place (and this happens more often in Ghana) the network is no longer capable of maintaining similarity. It fails. The question of whether Newton's laws are true in Ghana, then, does not depend on its distance from London in kilometers, but on whether steady electricity and some other crucial network nodes are persistently present"
(see Latour's Pasteurization of France)
141: "When the network holds, there is similarity. When it fails (when one of the alliances between the notes gets disrupted or one of the nodes falls apart) then there is difference. So [...] the crucial transition point from similarity to difference is not a boundary, but the stability of the network elements and the concomitant functionality. [...] In Africa, in some circumstances, the laboratory network doesn't hold."