Philosophers of science disagree on the extent to which epistemology transcends the social sphere in mature branches of science. In this paper I suggest a way of vindicating a key aspect of the transcendence thesis without questioning the social nature of science. Such vindication requires pistemological autonomy to prevail along channels having to do with (1) selection of research goals, (2) use of human subjects and public resources in research, (3) social interventions aimed at helping science fulfill its epistemic goals, and (4) social interventions aimed at helping people and the community protect themselves from harmful scientific activity. This paper focuses on type (3), specifically on social pressure to diversify the points of view represented in scientific research. My exploration roceeds by contrasting two case studies involving pluralist enrichment of scientific research. Both encompass epistemological reform. In one (Feminist Biology) reform is pushed largely from outside the scientific sphere; in the other (Einstein’s development of Special Relativity) reform originates largely from within. Examination of these cases hows why general pluralist arguments fail and also why social intervention in epistemological matters is a misguided activity – or so I argue.