The thesis that knowledge is a partly evaluative concept is now a wide-spread view in epistemology, informing some prominent debates in the field. Typically, the view is embraced on the grounds that justification is a necessary condition for knowledge and a normative concept � a reasona-ble motivation. However, the view also has counterintuitive implications, which have been neglected. In particular, it implies that J.L. Mackie�s er-ror-theory of value entails global epistemic scepticism and that any true knowledge claim suffices to prove the error-theory is false. In this paper, I elaborate these difficulties and address objections at length.