ences from epistemology in the strict sense, the epistemology that applies philosophy
to problems of knowledge without making any special effort to short-circuit the politi-
cal question. Steven Shapin (1994) would be an example of the first, and Pierre Duhem
(1904) of the second. I have the greatest respect for my epistemologist colleagues, who
use different tools from mine to try to grasp the secrets of scientific practices. I have
equal respect for the political epistemologists who agree to treat the theory of the sci-
ences and political science asa single philosophical problem.However, I have no respect
at all for those who claim that the “problem of knowledge” has to bedistinguishedfrom
the political question, in order to hold the frenzy of the social world at bay. It is crucial
to combat epistemologists of this last sort. It is in order to distinguish these from the
others that I add parentheses to the expression “(political) epistemology.” Either we
are talking about the organization of public life and must not mix in questions about
the nature of scientific activity, or else we are talking about scientific production and
there is no reason to combine it with considerations about bringing politics into line.
One political epistemology against another political epistemology, fine; epistemology
against epistemology, certainly; epistemology against politics, out of the question.
- The notion of Constitution*, essential to the comprehension of this argument, is
developed at length in Latour 1993: it is a matter of replacing the opposition between
knowledge and power, between nature and society, with a prior operation of distribu-
tion of the rights and responsibilities of humans and nonhumans. It is this notion that
makes symmetrical anthropology possible and makes modernity comparable to other
forms of public organization. - There is a direct and unbroken chain of arguments going from the first sophists
(Cassin 1995) to the “science wars” episodes. - I can be permitted at this point some chauvinism about my own field. And yet I
have never understood how readers of science studies could avoid from the outset see-
ing in this research a questioning of the very notion of the “social” world, “social” ex-
planation, “social” history. On this essential point, see two articles written some time
ago: Callon and Latour 1981, and Callon 1986. I have been involved in two enlighten-
ing quarrels on this point with adherents of “social construction”: see Collins and
Yearley 1992, with our reply in Callon and Latour 1992 and later in Bloor 1999, and my
response, Latour 1999a. Science studies have been accused of politicizing Science,
whereas they have done precisely the opposite: they have depoliticized the sciences by
putting an end to the kidnapping of epistemology by the epistemology police. - The word “politicize” is used from here on in two distinct ways. The first
amounts to reserving power plays for the prison of the Cave alone, and treating the
world of Science as apolitical. The famous “neutrality” of Science arises from this pre-
liminary distribution of functions between Science on the one hand and politics on the
other. “To politicize,” if we accept this division of work at least provisionally, will al-
ways amount to sullying pure and perfect Science by showing the powers at play be-
hind it. Against this pollution of scientific neutrality, it will always suffice to return to
the initial purity, to recall the “absolute difference” that exists between the concerns of
the human world and the cold reality of things. But to politicize also refers tothe very
invention of this absolute difference,to this division of roles between an apolitical reserve
on the one hand and the shrinking of public life to the realm of passions and interests
NOTES TO PAGES 13–17
253