Full Idea
Williamson says that instead of being viewed as a concept to be analysed, knowledge should be seen as something useful in the analysis of all sorts of other concepts to epistemology - and to philosophy of mind as well.
Gist of Idea
Don't analyse knowledge; use knowledge to analyse other concepts in epistemology
Source
report of Timothy Williamson (Knowledge and its Limits [2000]) by Keith DeRose - The Case for Contextualism 1.8
Book Reference
DeRose,Keith: 'The Case for Contextualism' [OUP 2009], p.20
A Reaction
I just don't believe this, because knowledge is obviously a complex state of mind, which invites breaking it down into ingredients. How could knowledge possibly be prior to truth?