Abstract
I defend a neo-Lewisean form of contextualism about knowledge attributions. Understanding the context-sensitivity of knowledge attributions in terms of the context-sensitivity of universal quantifiers provides an appealing approach to knowledge. Among the virtues of this approach are solutions to the skeptical paradox and the Gettier problem. I respond to influential objections to Lewis’s account.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
What of Father’s utterance? Father’s intention was similar to Aunt’s—he didn’t mean to be talking about the driver. So perhaps we should consider his sentence true. But Daughter, whom we’ve already agreed spoke truly, also took herself to be disagreeing with Father, as indicated by her use of the word ‘no’. How to treat such apparent disagreement is something of a vexed issue—its parallel in the case of knowledge is much celebrated, but this case shows that the phenomenon also occurs in less controversial cases of context-sensitive discourse.
This is not to say that contextualism about quantifier domains has no detractors; of particular note is the ‘Semantic Minimalism’ of Cappelen and Lepore (2005). Defending contextualism about quantifiers is beyond my present scope (but see Stanley and Szabó (2003)). My present conclusion may be thought of as the conditional claim that if contextualism about quantifiers is well-motivated, then so is contextualism about knowledge.
This, of course, is a significant mischaracterization of Lewis own view; to force a context in which ‘knows’ does not apply is no more to destroy knowledge than is to force a context in which ‘everybody’ does not apply to destroy everybody.
Williamson (2002, Chap. 9).
Other options than foundationalism are also available; one may respond to the classical regress challenge in any of the usual ways. One might, for instance, follow Klein (2007) and demand an infinite chain of known reasons, or admit circular chains of evidence on the model of coherentism.
At least it does so if the statement of E = K is meant to hold true in all contexts; other options are also available. For example, one could, if one liked, identify evidence with what is ‘known’ relative to some particular context; the resulting view is also consistent with my approach to knowledge.
Note also that this treatment explains some dynamic features of knowledge attributions; to claim (1) or (1′) is, in many contexts, felicitous, and (2) and (2′) may well follow in the conversation. But statements of (1) or (1′) are much more problematic when given after (2) and (2′). This apparent dynamic feature is further evidence that there is a contextualist element at work in the case of knowledge.
The term is due to DeRose (1995, pp. 27–28).
Cohen occasionally states the objection in the object language, as here: “Intuitively, because of his [Gettier] situation, S1 fails to know he sees a sheep, regardless of what is salient to the speaker of the context.” (p. 301) Presumably, this is to be interpreted in the metasemantic way suggested; the literal fact given here is not in tension with contextualism of the relevant sort. Compare Lewis’s own remark that his talk of knowledge vanishing with context shifts is an “expository shortcut, to be taken with a grain of salt.” (p. 566).
In introducing this complication, Lewis sets out what he recognizes to be a problem for his own view: since there is always some skeptical possibility that is saliently similar to actuality, how is it that any knowledge attributions come out true? Lewis confesses that he does not know how to avoid this consequence without adhockery. Cohen seems (p. 296) to consider it obvious that the best solution for Lewis is to understand the salient similarity of x and y as requiring that it be salient that x and y are similar. However, as Cohen goes onto show, taking this line brings about further problems for Lewis with respect to Gettier cases. Cohen’s moral is that Lewis has serious problems with Gettier cases; mine is that Cohen’s interpretation of Lewis is not the best one. As I’ll go onto argue, Lewis seems to intend a different notion of salient similarity—one that does not provide an obvious solution to the challenge introduced in the quoted paragraph.
In effect, then, the Lewisean contextualist faces a choice: he can adopt Cohen’s suggestion and require a stronger notion of salient similarity, thus avoiding the skeptical challenge here described without adhockery; or he can use a weaker notion of salient similarity that gets the results Lewis wants about Gettier cases. I side with Lewis in taking the latter option, and therefore inherit his problem of leaving unanswered the question about why ‘similarity with respect to evidence’ is not usually enough to include skeptical possibilities in the domain. But see footnote 12.
We can also find a problem parallel to the one Lewis admits for himself, discussed in my footnote 11. Every clown in the world resembles the clowns Aunt is focusing on in at least some salient sense—the sense of being a clown. Here, as in the knowledge case, the Ruse of Resemblance must be treated with care, and attempts to precicify it will look rather ad hoc.
Cohen (1998, p. 297, footnote 19) objects to this interpretation of Lewis’s ‘salient similarity’ by arguing that it does not explain why the Rule of Resemblance doesn’t make all skeptical scenarios relevant, resulting in widespread skepticism. But this of course is exactly the situation as Lewis sees it; see previous footnote.
Lewis’s own discussion of Bill is in the first person; he gives the special case where the subject is the ascriber: “we know he will never be rich.”
See their footnote 20.
References
Cappelen, H., & Lepore, E. (2005). Insensitive semantics. Malden, MA: Blackwell Press.
Cohen, S. (1998). Contextualist solutions to epistemological problems: Scepticism, Gettier, and the lottery. Australasian Journal of Philosophy, 76(2), 289–306.
DeRose, K. (1995). Solving the skeptical problem. The Philosophical Review, 104(1), 1–52.
Gendler, T., & Hawthorne, J. (2005). The real guide to fake barns: A catalogue of gifts for your epistemic enemies. Philosophical Studies, 124(3), 331–352.
Hawthorne, J. (2004). Knowledge and lotteries. New York: Oxford University Press.
Klein, P. (2007). Human knowledge and the infinite progress of reasoning. Philosophical Studies, 134(1), 1–17.
Lewis, D. K. (1996). Elusive knowledge. Australasian Journal of Philosophy, 74(4), 549–567. (Reprinted from Papers in metaphysics and epistemology, pp. 418–445, by D. Lewis, Ed., 1999, Cambridge: Cambridge University Press.
Oakley, I. T. (2001). A skeptic’s reply to Lewisian contextualism. Canadian Journal of Philosophy, 31(3), 309–332.
Stanley, J. (2005). Knowledge and practical interests. New York: Oxford University Press.
Stanley, J., & Szabó, Z. (2003). On quantifier domain restriction. Mind & Language, 15(2), 219–261.
Williamson, T. (2002). Knowledge and its limits. New York: Oxford University Press.
Acknowledgments
For valuable conversations, and for and comments on previous drafts of this project, I am grateful to Jessica Brown, Yuri Cath, Stewart Cohen, Paul Dimmock, Benjamin Jarvis, Carrie Jenkins, Ernest Sosa, Jason Stanley, Brian Weatherson, and Crispin Wright.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Ichikawa, J. Quantifiers and epistemic contextualism. Philos Stud 155, 383–398 (2011). https://doi.org/10.1007/s11098-010-9576-2
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-010-9576-2