Epistemology

7. Knowledge first epistemology

An important research programme in contemporary philosophy takes a ‘knowledge first’ approach to questions in epistemology, action theory and philosophy of mind. Whereas traditional epistemology has devoted great attention to defining knowledge in terms of belief, justification, evidence and related concepts, the knowledge first approach takes knowledge as a theoretical primitive, and uses it to explain other important phenomena in epistemology and beyond, including the concepts of belief and evidence. In this respect, the approach reverses the usual direction of analysis, making the concept of knowledge theoretically fundamental.

A major motivation for the knowledge first approach is pessimism with regard to finding a traditional definition of knowledge; that is, an account that gives insight into the nature of knowledge by stating individually necessary, jointly sufficient, and informative conditions. One reasons for pessimism is epistemology’s poor track record with regard to providing such an account. But some knowledge first theorists have provided more principled reasons for pessimism. For example, Williamson (2000) argues that recent developments in the philosophy of mind make it doubtful that knowledge can be ‘factored’ into purely internal components (such as belief) and purely external components (such as truth), as traditional analyses would seem to require.

The knowledge first approach abandons the project of providing a traditional account of knowledge. But this does not mean that it abandons questions about the nature of knowledge altogether. For example, it is still possible to defend substantive theses regarding what is necessary for knowledge, and also what is not necessary. In this regard, Williamson argues that knowledge requires a kind of safety. More exactly, knowledge that p requires that one’s belief that p is safe from error in relevantly close situations. However, a belief can be safe without being safely safe, reliable without being reliably reliable. For these reasons, Williamson argues, knowing does not require knowing that one knows. For similar reasons, knowledge is not ‘luminous’; that is, it is possible to know without knowing that one knows, and also possible to not know without knowing that one does not know.

Regarding the value of knowledge, the knowledge first approach endorses and develops a compelling picture regarding the relations among knowledge, assertion, action and practical reasoning. In slogan form, the purpose of knowledge is to serve action. More specifically, knowledge first theorists have argued that knowledge serves as the epistemic ‘norm’ of action, practical reasoning and assertion. For example: If you know that p, then you can act on p; If you know that p, then you can use p as a premise (a reason) in your practical reasoning; Assert that p only if you know that p. The idea is that norms such as these govern, and thereby help to explain, both individual planning and action, as well as the coordination of planning and action. Such a picture speaks to the practical value of the concept of knowledge, our knowledge language, and also knowledge itself.

The knowledge first approach has implications regarding the scope of knowledge as well. For example, a number of sceptical arguments trade on the assumption that our evidence is the same both in real world cases and in sceptical scenarios involving evil demons and brains in vats. One thought behind this ‘same evidence’ assumption, in turn, is that our evidence is constituted by our mental states, which are the same in both cases. But if our evidence is insufficient to ground knowledge in the sceptical scenario, this reasoning goes, then it becomes problematic as to how it can ground knowledge in the good case. Williamson (2000) argues that the knowledge first approach provides resources for rejecting this sceptical line of reasoning. First, if knowledge can itself be understood as a mental state, then it is false (or at least question-begging) that our mental states are the same in the two cases. Second, if evidence is to be defined in terms of knowledge, rather than the other way around, then it is false (or at least question-begging) that we have the same evidence in the two cases.

Finally, a promising line of research combines a knowledge first approach with virtue epistemology (Millar 2010; Miracchi 2015; Kelp 2019). A guiding idea of virtue epistemology is that knowledge is a kind of achievement in the following sense: knowledge is a kind of success grounded in ability, as opposed to a mere lucky success. For example, perceptual knowledge can be understood as successful perceptual representation, grounded in perceptual ability or competence. Knowledge first virtue epistemology adopts this general framework, but eschews the idea that cognitive abilities can be defined independently of knowledge itself. Accordingly, the approach rejects the idea that knowledge can be defined in terms of success from ability, at least insofar as philosophical definitions are supposed to be non-circular. Such an approach preserves much of the explanatory power of traditional virtue epistemology, however. For example, it can retain the idea the knowledge is an achievement, and that therefore knowledge has the same kind of value as achievements do in general. (See also Williamson, Timothy §3.)