2
three main practical tasks that must be solved in interviewing professionals (3). As a conclusion,
we will discuss risks and limitations inherent to the strategy we propose (4).
2. Three ‘pros’ and one ‘contra’
2.1 Informed interviewing is necessary because of the role of epistemic conditions of action
Scholars in Science and Technology Studies today widely agree that in order to understand
scientists’ actions, the way in which scientific objects and methods affect these practices must be
accounted for in sociological explanations. This necessity has been reflected in concepts like the
‘non-human actants’ of Actor-Network Theory (e.g. Callon 1986; Latour 1988; Law and Callon
1988); the “Mangle of Practice” (Pickering 1995) or “thin description” (Knorr-Cetina and Merz
1997). What all these concepts have in common is that they regard the epistemic content of
scientific work as important for sociological explanations. Similarly, an investigation of lawyers’
work could not ignore the content of the law and the opportunities and constraints for the
lawyers’ behaviour that are produced by the law.
In order to demonstrate how scientific content interferes with sociologists’ explanations, let us
turn to an empirical example from an investigation of institutional conditions for
interdisciplinary collaboration (Laudel 1999, 2001). In order to find causal relationships between
the institutional conditions of action and results of collaborative work, all factors that promoted,
hindered, enabled or prevented a collaborative project’s success needed to be analysed. When a
scientist answered: "the collaboration didn't work", it had to be clarified what "it didn't work"
actually meant, i.e. to what kinds of causes the scientist referred. In one case, the further probing
solicited the following explanation:
The .. protein .. he [the biochemist] gave us, .. was always too contaminated .. it has never worked. .. If you
want to crystallize it, it must be perfectly pure, otherwise it doesn’t work. Some proteins are very difficult to
purify...
The scientist referred to a ‘material resistance’ (the protein’s insufficient purity) as the main
cause for the collaboration’s failure. This was confirmed by other interviews and documents. It
became clear that neither lack of resources, nor difficult personal relations (the partners
collaborated successfully in other projects and got along well) nor other social reasons could
explain the collaboration’s failure. Epistemic conditions of action (the difficulties of protein
purification and the high purity that is required by crystallization methods) had to be included in
this explanation. More generally, epistemic conditions of action had to be included in the
investigation in order to provide accounts for the success or failure of collaborations. In order to
do that, we had to include the research content and research actions into the interview strategy.
On a more general level, differences between collaborative behaviour in different fields
needed to be explained. For example, one could ask "why do sociologists collaborate relatively
seldom whereas experimental physicists do it all the time?" An answer to that question would
have to include not only social properties and practices, but epistemic conditions of actions as
well. Knorr-Cetina applied this approach by describing epistemic cultures that clearly differ in
their collaborative practices (Knorr-Cetina 1999).
Convention