Dan O'brien

Hume is usually taken to have an evidentialist account of testimonial belief: one is justified in believing what someone says if one has empincal evidence that they have been reliable in the past. This account is impartialist: such evidence is required no matter who the person is, or what refotions she may have to you. I, however, argue that Hume has another account of testimony, one grounded in sympathy. This account is partialist, in that empincal evidence is not required in order for one to be justified in believing some of the assertions of one's friends.

_An Introduction to the Theory of Knowledge_ guides the reader through the key issues and debates in contemporary epistemology. Lucid, comprehensive and accessible, it is an ideal textbook for students who are new to the subject and for university undergraduates. The book is divided into five parts. Part I discusses the concept of knowledge and distinguishes between different types of knowledge. Part II surveys the sources of knowledge, considering both _a priori_ and _a posteriori_ knowledge. Parts III and IV provide an in-depth discussion of justification and scepticism. The final part of the book examines our alleged knowledge of the past, other minds, morality and God. O'Brien uses engaging examples throughout the book, taking many from literature and the cinema. He explains complex issues, such as those concerning the private language argument, non-conceptual content, and the new riddle of induction, in a clear and accessible way. This textbook is an invaluable guide to contemporary epistemology

One kind of successful communication involves the transmission of knowledge from speaker to hearer. Such testimonial knowledge transmission is usually seen as conforming to three widely held epistemological approaches: reliabilism, impartialism and evidentialism. First, a speaker must be a reliable testifier in order that she transmits knowledge, and reliability is cashed out in terms of her likelihood of speaking the truth. Second, if a certain speaker's testimony has sufficient epistemic weight to be believed by hearer1, then it should also be believed by hearer2. Third, the normative constraint here is evidentially grounded: whether or not a hearer should believe a speaker depends on the evidence the hearer has that the speaker is telling the truth. I argue that there are cases of testimonial knowledge transmission that are incompatible with these three claims. This is when one accepts the testimony of an intimate friend.