ABSTRACT
According to the ideal of value-free science, non-epistemic values have no role to play in science proper. Even among those who deny that science is value free, many hold that epistemic values must take priority over non-epistemic values when it comes to the inferential processes in science. The author defines epistemic standards as non-algorithmic factors that characterize theories as epistemically good qua scientific theory, for example, empirical adequacy, logical consistency, predictive accuracy and quantitative precision. In ideal cases, students do not need epistemic priority, because they can successfully satisfy both epistemic standards and social values. In the non-ideal cases, there are various reasons that scientists may make non-ideal decisions in which epistemic standards and non-epistemic values pull in opposite directions. Epistemic priority implies, or is strongly associated with, a sort of noncognitivism or anti-realism about value judgments. EPT treats epistemic standards as criteria for successful scientific inquiry, rather than as values that are good if we can have them.
