Bespoke Knowledge

A central finding of social epistemology is that phenomena desirable at the individual level might have hidden undesirable effects that emerge on the social level. These emergent effects have important implications for the management of values in science. Using a historical case study of the safety of hormonal pregnancy tests, I argue that the meta-analysis of studies risks compounding value-influence in undesirable ways. In statistical analysis, researchers must either constrain values such that they do not impact the measure of central tendency or accept that they have created “bespoke knowledge”, knowledge which cannot be reused or exported to other contexts.