8
simplifying the relationships between them. The main difference between the original meaning
of the concept and Galison’s use of it, one the one hand, and the situation in an interview is that
despite all of the interviewer’s preparations, the language must be created almost
instantaneously, namely in the course of one interview.
As a skeleton of such a language, some ‘common ground’ of natural science and social
science research can be used. In any empirical research process, researchers start with a question
that is somehow rooted in a theoretical background, investigate a research object by applying
methods that must be developed or adapted, and interpret the empirical results. Although only
few research projects follow exactly that sequence of steps, the steps themselves will occur in
one form or another in all research processes, and scientists’ perceptions of research processes
correspond to this model.
We can use this very abstract level of common experiences to formulate questions about the
interviewee’s research. In the investigation of scientists' collaborations interviewees were asked
about the elements of their research processes, e.g. by using the following questions:
What research problem do you deal with?
Could you explain to an outsider what it is you try to find out?
What methods do you apply? What equipment do you use?
What substances do you use? Where do these substances come from?
Wherever possible, these questions were specified by detailed knowledge that had been acquired
in the preparation of the interview by reading research proposals, research reports, etc. (see 3.2).
The questions about elements of research processes led to hints about other researchers who
contributed to the interviewees’ research in different phases. Thus, the cognitive links that were
created via the exchange of substances, joint use of equipment, etc. hinted to other researchers
who were identified as collaborators.
Q: And have you applied additional methods in your project?
A: Well, I would say yes, I did try something different… We tried to characterise these layers by
ellipsometric methods, for example, because we were never sure how they look like …
Q: Where did you do this?
A: I gave the layers to these people, that means in the clinical research centre. Y and the current
undergraduate student are working at it. And they tried it, because I do not know the equipment very well. I
looked at the equipment and watched them when they were working.
In this part of the interview, the interviewer and the interviewee jointly deconstructed the
interviewee’s research project, which led to the identification of collaborations. This strategy
worked well in all interviews on collaboration. By ‘disassembling’ the research process into its
elements it was possible to find opportunities for collaboration, as well as real (successful and
unsuccessful) collaborations. The variation of links between the research processes that were
reported in the interviews supported the construction of types of collaborations.
The interviewers’ success depended primarily on the common language that was constructed
at an appropriate level of simplification. This was especially problematic in the investigation of
Convention