shown to be wrong. In other words, to have discomfirmable
knowledge.For this panelist, the ability to replicate results is crucial. “Other-
wise, what one is doing is a personal expression...whatisinterest-
ing [about the work of scholars he admires] is not about their own
view of the world, but about the world itself.” Not all social scientists
share this conception of subjectivity (“of the personal”) as a corrupt-
ing influence on the production of knowledge, however. Much in
line with standard practices in high energy physics and other scien-
tific fields, there is a strong tendency in some quarters of the social
sciences to acknowledge the role of interpretation and induction in
research and to point out the researcher’s back and forth move-
ment between theory and empirical analysis (sometimes pejoratively
termed “data massaging”).^18
Much like the gulf between the humanities and the social sciences,
this split within the social sciences, between those fields where em-
piricism is more exclusively favored and those where interpretation
is considered an essential ingredient, also influences panel delibera-
tions. An anthropologist describes sociologists, political scientists,
and economists as tending to emphasize “theoretical models” and
“statistical framework[s],” whereas anthropologists and historians
“put more emphasis on language proficiency, knowledge of the cul-
ture, spending time in the place you’re researching.” A sociologist
elaborates on these distinctions:
Sociology, political science, economics people share one set of cri-
teria, and the people from anthropology and history share differ-
ent criteria. Anthropology and history are much less positivistic,
much more comfortable with single cases. Whereas the political
science, sociology, [and] economics representatives want multiple
cases and stronger research design. They are less comfortable with62 / On Disciplinary Cultures