Abstract

This paper will investigate whether constraints on possible forms of limitation of the autonomy of science can be derived from epistemological considerations. Proponents of the autonomy of science often link autonomy with virtues such as epistemic fecundity, capacity to generate technological innovations and capacity to produce neutral expertise. I will critically discuss several important epistemological assumptions underlying these links, in particular the “unpredictability argument”. This will allow me to spell out conditions to be met by any form of limitation of the autonomy of science to be epistemologically acceptable. These conditions can then be used as a framework to evaluate possible or existing forms of limitations of the autonomy of science. And it will turn out that the option of direct public participation might not be the best way to go to democratize the setting of research agenda.