Many responses have been offered to the Harman-Vogel paradox in recent years, most notably a series of innovative rival proposals about either the semantics of knowledge-ascribing sentences, or the metaphysical determinants of knowledge itself. Semantic proposals such as contextualism (Cohen 1999; DeRose 2009; Lewis 1996) and relativism (MacFarlane 2005; Richard 2004) hold that the verb ‘to know’ has some surprising features. Contextualists aim to explain the shifting intuitions of the paradox by claiming that ‘knows’ picks out different relations in different conversational contexts. Once worries about remote possibilities have been raised, ‘knows’ comes to denote a more demanding relation, so that a subject needs to be in a stronger epistemic position to count as knowing. Relativists take the more radical step of saying that the truth of a knowledge-ascribing sentence for an assessor depends in part on a hidden variable, the assessor’s current epistemic standards parameter. Meanwhile, metaphysical proposals such as interest-relative invariantism argue that the paradox is better handled by taking knowledge to involve some factors overlooked in traditional epistemology, factors such as the subject’s practical concerns (Fantl & McGrath 2007; Hawthorne 2004; Stanley 2005). On such views, whether a subject knows a given proposition may depend in part on what he is worried about, or whether that proposition is a high-stakes matter for him. It is safe to say that each of these novel semantic and metaphysical theories is thought to have a number of counter-intuitive consequences.
« Prev Next »
Top of page Top of page