rise—in short, everything that makes it possible to articulate proposi-
tions.^3 As a result, the use of the term “fact” without further precau-
tions to designate one of the territories outlined by the frontier be-
tween facts and values completely obscures the immense diversity
of scientific activity and obliges all facts, in every stage of their pro-
duction, to become fixed, as if they had already reached their defini-
tive state. This freeze makes it necessary to use the same words to des-
ignate a multitude of sketches, prototypes, trials, rejects, and waste
products, for want of a term that makes it possible to diversify the
gamut, rather as if we called all the successive stages of an assembly
line “cars,” without noticing that the word designates sometimes iso-
lated doors, sometimes a chassis, sometimes miles of electrical wire,
sometimes headlights. No matter what term we choose later on to re-
place “fact,” it will have to highlight the process of fabrication, a pro-
cess that alone makes it possible to record the successive stages as well
as the variations in quality or finishing touches that depend on it; it
will have to encompass matters of concern* as well as matters of fact.
The notion of fact has another, better-known defect: it does not al-
low us to emphasize the work of theory that is necessary for the estab-
lishment of the coherence of the data. The opposition between facts
and values, in fact, unfortunately intersects with another difference
whose epistemological history is very long, the opposition between
theory and the data that are called, in contrast, “raw.” The philosophy
of science, as we are well aware, has never been able to put forward a
united front on this issue. If the respect for matters of fact appears es-
sential to the deontology of scientists, it is no less true that an isolated
fact always remains meaningless as long as one does not know of what
theory it is the example, the manifestation, the prototype, or the ex-
pression.^4 In the history of the sciences one finds as much mockery
against builders of vain theories that have been overturned by some
tiny bit of evidence as one finds jokes at the expense of avaricious
“stamp collectors” who accumulate heaps of data that a single astute
thought would have sufficed to predict. An effort to shape, form, or-
der, model, and define seems necessary if one wants brute facts, speak-
ing facts, obtuse facts, to be able to stand up forthrightly to those who
chatter on about them. Here again, there are too many hesitations be-
tween positivism and rationalism for us to take the word “fact” as an
adequate description of these multiple tasks. To our set of specifica-
POLITICS OF NATURE
96