One of the earliest papers arguing for the importance of the sociology of science for the epistemology of testimony was John Hardwig’s “The Role of Trust in Knowledge.” There, Hardwig argues, among other things, that (1) merely prudential reasons for trusting others—i.e., reasons stemming from one’s knowledge of the fact that scientific fraud would be imprudent on the part of the person attempting to defraud the scientific community—are insufficient, given the demonstrative prevalence of fraud in science. (702-5) Furthermore, Hardwig argues (2) that if prudential reasons are insufficient to guarantee good science, then ethical researchers are required. (705)
Let’s leave aside the plausibility of whether, despite the much-hyped Darsee, Imanishi-Kari, and Schoen cases, among others, serious fraud really is so widespread within the scientific community. The issue I wish to consider briefly here is whether prudential reasons are exhausted by the fear of exposure of fraudulent research practices and of the attendant punishment for those transgressions.
The point I want to raise here is that the prevalence of the use of familial terminology to describe the relationship between mentors and their graduate students and research assistants might point to an extremely powerful mechanism for promoting honest communications among collaborators—if the collaborators are linked by such “familial” bonds.
First, note that many of the most successful scientists are linked by “familial” bonds. There are numerous references in the literature of the history and sociology of science (most famously, perhaps, Merton’s seminal “Matthew Effect”) to the power of one’s academic genealogy as a predictor of success in a given scientific field.
Second, note that many invoke those “familial” relationships to explain their communications with other, often otherwise unknown, researchers. Just to cite one example, consider H.M. Collins’s quote from a Scottish scientist, explaining why some of their American colleagues singled their group out as one with whom to share the then cutting edge information as to how to construct a Transversely Excited Atmospheric Pressure C02 (or, simply, TEA) laser, in Collins’s “The TEA Set”:
these people of course had got a Ph.D. working in the same group under the same supervisor, albeit eight or ten years ago. But they’re still one of our family.
This quote suggests the power of familial relationships in research science to forge collaborative links. However, it doesn’t explain that power.
My suggestion is that “familial” links of the sort under consideration here provide a further prudential motivation for relying on other researchers. In the same way that one might expect the successes of members of one’s biological family to redound to one’s own advantage, one may also expect the successes of researchers with whom one shares a common academic lineage to increase one’s own prestige. If this is the case, however, it would not be strategic for one to act in such a way to jeopardize one’s status within one’s academic “family”—and it would be strategic for one to act in such a way to make the “family” proud. Fraudulent research would put one, thus, in jeopardy—even in cases in which the academic community failed sufficiently to exact formal penalties for wrongdoing. In this way, then, bonds of academic lineage provide shorthand evidence of trustworthiness, and it is little wonder that they therefore prove of such importance to working researchers.
Of course, this discussion leaves still open the question of whether such “familial” relationships provide strong evidence of the importance of trust in science, as Hardwig, for one, might claim, or whether the predominance of such relationships points to the importance of obviating the need for truly trusting relationships in science—i.e., trusting relationships in which the trustor is forced to rely on the trustee despite uncertainty as to the trustee’s motives, competence, etc. The resolution of this question, however, awaits a more detailed discussion of the nature of trust itself.