Testimony, science and the social: Obtaining knowledge from authorities

The normal way of forming a belief upon the testimony of a speaker, namely, by taking the speaker's word for it, involves granting the speaker epistemic authority on the asserted proposition. While contemporary epistemologists have recognized our extensive reliance on other people's words, they have failed to adequately deal with the normal way of forming testimonially-based beliefs, because they have either ignored or misunderstood our reliance on the epistemic authority of others in forming such beliefs. I present an account of epistemic authority and explain how we can be entitled in believing the testimony of speakers upon their authority.; At the center of the dissertation is a study of the concept of 'epistemic authority'. To have epistemic authority is to have the ability to transfer entitled doxastic attitudes to trusting and deferring thinkers. According to the standard account of epistemic authority, one has epistemic authority on p if and only if one knows that p. Against this I argue that to have epistemic authority on p one must have non-shared knowledge, which entitles one to one's doxastic attitude towards p, and one must know that one has good grounds for one's doxastic attitude towards p. Thus knowledge that p is not necessary for epistemic authority on p, and, moreover, may not be sufficient for authority on p, unless our concept of knowledge is more demanding than is usually thought.; At the core of the Enlightenment approach to testimony is the claim that deference to epistemic authorities cannot be justified. I attempt to explain the legitimacy of such deference without endorsing a radical rejection of the empiricist approach to testimony. I argue that our justification in believing the testimony of others is always normatively dependent on empirical evidence available to us. Nonetheless, we can be entitled in disregarding relevant empirical evidence available to us in deferring to an authoritative speaker, because such deference allows us to become sensitive to better evidence available to the speaker. I defend the claim that epistemic deference can be legitimate against a challenge that emerges from scientific norms disallowing epistemic deference, by suggesting that these norms are reasonable because a scientific community governed by them allows members to better serve as epistemic authorities.