So, here we have our story. During the 17th century, strange, new artifacts came to life, able to produce strange, new phenomena under carefully constructed circumstances. To conclude, let us recapitulate by sketching the method by which these fact-factories came to be operated.
First, we need a good machine. ‘Good’, meaning sealed air-tight, fully controllable. The practice of fact-production is characterized by a continuous effort to find better materials and techniques in order to approximate the ideal machine: a world of itself, unhampered by the world outside.
However important materials and craftsmanship may be for the experimental practice, we cannot overlook the social conditions necessary for the production of a matter of fact. In order for a new phenomenon to become a fact, it must be witnessed by a community of as many witnesses as possible, trusting each other’s observational abilities and agreeing on what they observed.
This community of witnesses furthermore protects the borders of the laboratory by excluding considerations that cannot be experimentally tested, such as questions about the nature of animals, or the existence of a ‘subtle matter’.
Unfortunately, in contrast with the hopes of the experimentalists, facts don’t speak for themselves. Moreover, their silence is a consequence of their definition: if they ought to be produced in abstraction from the world outside of the laboratory, then how are they going to tell us something about that world?
And this leaves the matter of fact suspended in the void. If we want it to be meaningful, we will have to interpret it, and bring back into the laboratory what was expelled from it. The fact-factory cannot be sealed off completely from the theoretical speculations that are brought in, wanted or not, with the witnesses. And as our story of the air-pump shows, the fragility of the fact-factory, the cracks in its foundations, already started to appear in one of its first products. However, even though the fact-factory might be built on less solid foundations than was hoped for, that does not change anything about its productivity. Every anomaly leads to new experiments, engendering new surprises, investigated with still more experiments, and so on and so forth, in an endless swirling stream. And science, as it turns out, has always been more than a fact-factory, there has always been something ‘behind’ the facts. Facts don’t speak for themselves, it is our task to speak for them.