Remarks on Hansson’s model of value-dependent scientific corpus

Abstract

This article discusses Hansson's corpus model for the influence of values (in particular, non-epistemic ones) in the hypothesis acceptance/rejection phase of scientific inquiry. This corpus model is based on Hansson's very convincing concepts of scientific corpus and science ‘in the large sense’. I present Hansson's corpus model of value influence, comment on its origins, analyse its advantages and disadvantages, and conclude that it is a very good candidate for managing value influence, because contrary to other models in the literature it is simple, it provides a universal procedure for dealing with values, and it systematically preserves the epistemic integrity of science. I comment on potential difficulties associated with this model: how it manages controversial non-epistemic values; the difficulties associated with systematically taking the maximum corpus entry requirement (difficulty to identify this maximum requirement, and non-optimality with respect to other less demanding requirements); and the issue whether a claim belonging to the corpus could still be deemed not reliable enough for practical applications. I note that these issues do not put into question the applicability of the corpus model, but rather underline the need of precisely defining the frontier between the corpus and its context of application. Finally, I call for empirical work (in particular, interviews of practising scientists) in order to better assess these issues.