The Indexicality of ‘Knowledge’∗

Epistemic contextualism (EC)—the view that the content of the predicate ‘know’ can change with the context of utterance—has fallen into considerable disrepute recently. Many theorists have raised doubts as to whether ‘know’ is context-sensitive, typically basing their arguments on data suggesting that ‘know’ behaves semantically and syntactically in a way quite different from recognised indexicals such as ‘I’ and ‘here’ or ‘flat’ and ‘empty’. This paper takes a closer look at three pertinent objections of this kind, viz. at what I shall call the Error-Theory Objection, the Gradability Objection and the Clarification-Technique Objection. Let us briefly consider each of these objections before going into detail. According to the Error-Theory Objection to EC, EC is to be rejected because it is committed to a controversial error-theory, viz. to the view that competent speakers are sometimes systematically mistaken about the content of ‘know’. Objections discussing EC’s error-theory can be found in many different guises in the literature, and their impact has been considerable: opponents of EC usually take them to provide decisive evidence against EC, and even staunch contextualists such as Steward Cohen feel that they ought to address the problem of the non-obviousness of the indexicality of ‘know’. I shall argue in this paper that EC’s error-theory isn’t as implausible as is