Trust to Testimony: Reductionism and Non-Reductionism

Introduction. The article is devoted to the epistemology of communicative knowledge. It is argued that the central problem in the analysis of such knowledge is the question of the status of testimony. The author discusses reductionism and non-reductionism as two traditional approaches to the problem of trust to testimony. The aim of the article is to describe the arguments of both approaches and to carry out their critique.Methodology and sources. The author uses the method of conceptual analysis to address the task at hand. The primary sources of the paper include the works of the classics of the theory of knowledge – D. Hume and T. Reid. The secondary sources include works of modern authors, belonging to the Anglo-American philosophical tradition.Results and discussion. Reductionism argues that although testimonial beliefs can be accepted on the basis of the reliability of the informant, the testimony itself does not provide justification. Ultimately, reductionism reduces testimony to another source of justification, such as perception or memory. The article explains that reductionism's weakness is to account for normal social interactions when the reliability of the speaker is unknown. Non-reductionism argues that testimony does not provide justification by reduction to other sources. Testimony itself is a valid form of social proof. The informant does not need additional positive reasons to accept the speaker's claims as valid. Testimony is justified by default unless it is proven false or unreliable. However, the weakness of non-reductionism is the relativistic implication in cases of cognitive asymmetry between the informant and the informant.Conclusion. The article concludes that reductionism and non-reductionism are equally unsatisfactory theories. The author believes that the formation of communicative knowledge does not depend on trust in the informant's testimony. It is proposed to consider the issue of trust in the informant himself as an epistemic agent. It is suggested that such an analysis should start from an virtue epistemology.