Contextualism

views updated

CONTEXTUALISM

The term "contextualism" has been used to denote many different philosophical theories. Within epistemology alone, there are two broad categories of theories that have been called "contextualist": subject contextualism and attributor contextualism.

Subject Contextualism

A few basic concepts are needed to explain subject contextualism. Let S be an epistemic subject, a being whose cognitive attitudes are proper targets of epistemic evaluation. Let C be a cognitive attitude that S has. C may be a belief, a judgment, a high degree of confidence, an affirmation or endorsement of some kind—any attitude that is a proper target of epistemic evaluation. C has a propositional content p. Finally, let x be the situation in which S C s that p. We will hereafter specify the target of epistemic evaluation as "S 's C ing that p in x."

According to subject contextualism, whether S C s that p in x constitutively depends on features of x that are metaphysically independent of S 's cognitive attitudes and of the truth-values of the propositional contents of S 's cognitive attitudes. As these features of x vary, so too does the epistemic status of (the degree of truth attached to) S 's C ing that p, even if S 's cognitive attitudes and the truth-values of their propositional contents all remain fixed. One or another version of such a view has been suggested in various passages in the writings of C. S. Peirce, Ludwig Wittgenstein, John Dewey, Karl Popper, W. V. Quine, and J. L. Austin. But only since the mid-1970s has subject contextualism been developed with any precision and generality.

The different versions of subject contextualism differ from each other in at least two ways. First, these different versions specify different features of x as relevant to constitutively determining whether S C s that p in x. Second, these different versions of subject contextualism specify different ways in which the relevant feature of x can determine whether S C s that p. By differences of the first kind, we can distinguish the various theories of subject contextualism that have been propounded into three broad groups.

According to one group of subject contextualist theories (Stine 1976, Goldman 1976, Dretske 1981), the epistemic status of S C ing that p in x constitutively depends on the objective probability of p 's being true in x. Other things being equal, the higher the objective probability of p 's being true in x, the higher the epistemic status of S C ing that p in x.

The most prominent argument in favor of this first variety of subject contextualism proceeds from consideration of case pairs such as the following (from Goldman 1976): Suppose that, in normal daylight, Henry, who has normal visual powers, has an unobstructed view of a barn right in front of him. Henry sees the barn, has a normal visual experience as of a barn, and believes that there is a barn in front of him. If there is nothing unusual about the case, then Henry knows that there is a barn in front of him. But now suppose that Henry's environment is full of barn facades that look exactly like barns from the angle and distance at which Henry is currently viewing the real barn in front of him. In this second case, just as in the first case, Henry has a true belief that there is a barn in front of him. And in both cases, this belief is based on Henry's seeing the barn, on his normal visual experience as of a barn. But in the second case, unlike the first, Henry does not know that there is a barn in front of him. What difference in the two cases could account for this difference in whether or not Henry knows? Subject contextualists of the first variety say that the difference in Henry's extrapsychological environment—in particular, the frequency of barn facades in his environment—is responsible for the difference in whether he knows.

Opponents of this first variety of subject contextualism typically respond to the preceding argument by offering an alternative explanation of why Henry knows, in the first case but not in the second case, that there is a barn in front of him. In each case, they contend, Henry believes that there is a barn in front of him partly because he believes that, in his environment, things that look like barns from his vantage point typically are barns. So his belief that there is a barn in front of him is partly based on the latter epistemic belief. And the epistemic belief is true in the first case and false in the second case. It is the difference in the truth-value of the epistemic belief that explains why Henry knows in the first case but does not know in the second case—or so say the opponents of this first variety of subject contextualism.

According to a second group of subject-contextualist theories (Annis 1978, Williams 1992, Henderson 1994, Klein 1999), whether S C s that p in x constitutively depends on the inquiry that takes place in x. The inquiry fixes which considerations confer positive epistemic status upon S 's C ing that p. The main argument for this second variety of subject contextualism, propounded in different ways by Annis, Williams, and Klein, has to do with the regress of reasons. According to this argument, neither foundationalism nor coherentism offers a correct account of structure of epistemic reasons, or justifications. Foundationalism cannot offer a correct account, because it is committed to the unsustainable claim that some cognitive attitudes—the foundational beliefs—are intrinsically justified. And coherentism cannot offer a correct account, because it is committed to claiming either that circular reasoning provides justification or that each belief in a coherent set is a foundational belief. But if there are no foundational beliefs and if circular reasoning does not provide justification, then how can positive epistemic status accrue to S 's C ing that p ? According to this second variety of subject contextualism, this question is best answered as follows: S, in a particular context of inquiry, makes certain presuppositions. These presuppositions can provide epistemic reasons, or justifications, for S 's other cognitive attitudes in that same inquiry. But when S moves into a different context of inquiry, some presuppositions in the earlier inquiry may be put into question in the new inquiry, and other propositions that were in question in the earlier inquiry may simply be presupposed in the new inquiry.

To the preceding argument, Henderson adds that because our cognitive competence is limited, in ways that can be empirically ascertained, we are incapable of forming beliefs about our environment without taking a great deal for granted. What we need to take for granted to form needed beliefs will vary from task to task. Since we are incapable of forming the beliefs that we need to form without taking a great deal for granted, we cannot be epistemically obligated to do otherwise. Our epistemic obligations cannot exceed our cognitive potential. Since our belief-forming processes require us to take more or less for granted, depending on the cognitive task at hand, our epistemic obligations must allow us to take more or less for granted, depending on the task.

Opponents of this second variety of subject contextualism typically respond to the preceding arguments by defending foundationalism or coherentism. While foundationalist and coherentist theories of justification might hold us to normative standards that we do not commonly meet, this is not a problem for those theories if, in the epistemological realm, "ought" does not imply "can."

According to a third group of subject-contextualist theories (Fantl and McGrath 2002, Hawthorne 2003, Stanley 2004), whether S C s that p in x constitutively depends on how the truth-value of p affects S 's interests, or perceived interests, in x. Other things being equal, the higher the cost, or perceived cost, of S 's being wrong that p, the less likely that S C s that p. The most prominent argument in favor of this third variety of subject contextualism considers pairs of cases such as the following (adapted from Fantl and McGrath 2002): Suppose that you are at the train station waiting for the train to New York. You would like to get on the express train so that you can be in New York by dinnertime, but it does not matter all that much to you whether you get there by dinnertime or not. You ask someone else on train platform, "Is the next train an express train, or a local?" and your honest and knowledgeable interlocutor sincerely tells you that it is an express. You believe her, and you have no reason to distrust her. In this situation, it seems that you know, and are justified and warranted in believing, that the next train is an express. But now suppose that the situation is exactly the same except that your life depends on your being in New York by dinnertime. In this case, the testimony of your honest and knowledgeable interlocutor does not justify or warrant—let alone give you knowledge—that the next train is an express. When so much depends on your being right, knowledge, justification, warrant, etc., all require more than they otherwise would require.

Opponents of this third variety of subject contextualism typically respond to the preceding argument by claiming that what depends on a subject's actual or perceived interests is not the epistemic status of the subject's C ing that p, but rather the rationality of the subject's acting as if p were true. For these opponents, whether S C s that p is fixed independently of S 's actual or perceived interests.

These are the main varieties of subject contextualism, and the arguments concerning them. Some subject contextualists are also attributor contextualists, and almost all attributor contextualists are also subject contextualists.

Attributor Contextualism

A few more basic concepts are needed to explain what attributor contextualism is. Let A be an epistemic attributor, someone who epistemically evaluates S 's C ing that p in x. A 's evaluation of S 's C ing that p in x will also be a cognitive attitude of some sort, either expressed (as in the case of an assertion) or not (as in the case of silent thought). A 's evaluation may concern whether S knows that p in x, or it may concern whether S is justified, reasonable, rational, or warranted in C ing that p in x, or it may concern whether S has adequate grounds, evidence, or reasons to C that p in x. More generally, an epistemic evaluation or appraisal of S 's C ing that p in x is a determination of whether S C s that p in x. Let y be the situation that A is in when A evaluates whether S C s that p in x. We will hereafter specify the act of epistemic evaluation as "A 's evaluation, in y, of S 's C ing that p in x."

According to attributor contextualism, the semantic value (the truth) of A 's evaluation (of S 's C ing that p in x ) constitutively depends on features of y. As these features of y vary, so too does the semantic value of A 's evaluation, even with everything else held fixed. The earliest prominent statements of such a view appear in Lewis (1979) and Dretske (1981). The view gained widespread notice following the publication of Cohen (1988), DeRose (1995), and Lewis (1996).

Different versions of attributor contextualism specify different features of y as relevant to constitutively determining the semantic value of A 's evaluation of S 's C ing that p in x. And even if two attributor contextualists agree about which features of y are relevant to constitutively determining the semantic value of A 's evaluation, they might still disagree about precisely how those features of y are relevant. By differences of the first kind, we can distinguish the various attributor-contextualist theories on offer into several groups. Although there is thus some diversity among attributor-contextualist theories, a single line of argument has generally been used to support attributor contextualism.

The argument in question proceeds from consideration of cases similar to those commonly adduced to support the third variety of subject contextualism. Suppose that Jones and Smith are at the train station trying to catch a train to New York. They want to know whether the next train is an express. They ask a bystander if he knows whether the next train is an express. The bystander looks at a schedule and replies, "Yes, I know. It is an express." It turns out that Jones and Smith have to be in New York as soon as possible, and cannot afford the mistake of getting on a local train. Jones says, "That schedule could easily have been outdated. That guy does not really know that the next train is an express." So the bystander claims to know that the next train is an express, but Jones claims that the bystander does not know that the next train is an express. Who is right?

If the bystander is right that he knows, then is Jones making a false assertion when he says that the bystander does not know. Suppose that the bystander's warrant for thinking that the next train is an express is precisely what Jones takes it to be. If such warrant is strong enough to give the bystander knowledge, then, it seems, it is also strong enough to give Jones and Smith knowledge. But if Jones and Smith have warrant strong enough to give them knowledge that the next train is an express, they have no reason to check further whether the next train is an express. Since they clearly do have a reason to check further whether or not the next train is an express, their warrant cannot be strong enough to give them knowledge of whether or not it is. But if they do not have enough warrant for knowledge, then it seems that the bystander does not know, since he does not have any more warrant than they do.

Suppose that the bystander is wrong, that he does not know that the next train is an express. In this case, it seems that most of the knowledge attributions that we make in ordinary life are wrong as well, since our warrant for most of what we claim to know is no greater than is the bystander's warrant for the claim that next train is an express. So if the bystander does not know that the next train is an express, then most of us know very little of what we ordinarily claim to know.

How can we avoid simply granting that Jones and Smith are right to deny that the bystander knows, or that the bystander is right to claim to know? The attributor contextualist avoids granting this by claiming that the truth-values of knowledge attributions are relative to the context in which the attribution is made. Relative to the context in which the bystander claims to know, her claim is true. But relative to the context in which Jones claims that the bystander does not know, his claim is true. So both claims are true, and they do not contradict each other. These assertions only appear to contradict each other because we fail to notice that "knows" requires a higher standard (or signifies a more stringent epistemic relation) in one context of attribution than in the other context of attribution than in the other context of attribution. Analogous arguments may support the conclusion that ascriptions of other epistemic properties, not just ascriptions of knowledge, are semantically sensitive to context.

Opponents of attributor contextualism will typically reply to an argument like the preceding in one or both of the following two ways. First, like Bach (2005), they may claim that, although the bystander knows that the next train is an express, Jones and Smith do not know, and that is because knowledge requires sufficient confidence. Although the bystander is sufficiently confident that the next train is an express, Jones and Smith are not, and so they do not satisfy one of the necessary conditions of knowledge. If, without any further investigation, Jones and Smith claimed to be sufficiently confident of the bystander's claim, then, given how much is at stake in their being right, their degree of confidence would be irrational.

This suggests a second line of response to the attributor-contextualist argument above: Even if Jones and Smith, prior to doing any further investigation, are sufficiently confident that the next train is an express, their confidence is unreasonable, given how much is at stake. Knowing that p requires not simply that one be sufficiently confident that p, but moreover that one's level of confidence be reasonable. Since, without doing any further investigation, Jones and Smith cannot be reasonably confident that the next train is an express, they also cannot know that the next train is an express, even if they are sufficiently confident of it (by the bystander's standards), even if they share all of the bystander's evidence for it, and even if the bystander himself knows. On this second line of response, we resolve the problem set out by the attributor contextualist's argument by appealing to subject contextualism of the third variety distinguished above.

To bolster the argument for attributor contextualism in the face of these objections, attributor contextualists must now attempt to run their thought experiments while controlling for variation in S 's level of confidence and also in S 's level of reasonable confidence. To do this, they must focus on a particular epistemic subject S in a particular context x, and then find variation in the truth conditions for asserting that S C s that p in x. To make the case for attributor contextualism, they must make sure that the variation they discover is variation in the truth conditions of an ascription of knowledge, and not simply in the conditions under which we are inclined to make, or are warranted in making, the ascription. DeRose (2004) undertook to do all this.

Three other sorts of argument that have commonly been used to support attributor contextualism. The first is an argument to the effect that attributor contextualism provides the best response to a skeptical argument like the following:

Premise 1: I cannot possibly know that I am not a brain in a vat being electrochemically stimulated to have realistic experiences.
Premise 2: If I knew that I have hands, then I could deduce, and thereby come to know, that I am not a brain in a vat being electrochemically stimulated to have realistic experiences.
Conclusion: I do not know that I have hands.

While some philosophers would simply to deny one of the premises, attributor contextualists typically take such denial to be implausible. So how can attributor contextualists avoid accepting the skeptical conclusion of such an argument? They can do so by claiming that the skeptical conclusion is true only relative to contexts of attribution that we enter into by thinking (in some way or other) about premise 1. Relative to other, more commonplace, contexts of attribution, premise 1 is false, as is the skeptical conclusion of the argument. Attributor contextualists typically take this response to the skeptical argument above to be more plausible than any alternative response, and they take this to count as a point in favor of attributor contextualism concerning knowledge attributions. Analogous arguments have been adduced in favor of attributor contextualism concerning attributions of other epistemic properties.

A second style of argument in favor of attributor contextualism proceeds from premises concerning the epistemic properties in question. For instance, Dretske (1981) and Lewis (1996) both claim that knowledge that p involves having infallible grounds for one's belief that p. For them, one's grounds are infallible just in case all alternatives to p are ruled out. But "all," like other quantifiers, involves a contextually restricted domain of quantification. Ruling out "all" alternatives means ruling out all those that fall within the contextually restricted domain of quantification. So on this account, some form of attributor contextualism is true of knowledge attributions. Analogous arguments show that attributor contextualism is true of attributions of other epistemic properties as well.

Finally, a third style of argument in favor of attributor contextualism proceeds from premises concerning the conversational function of epistemic-property attributions and epistemic appraisal. Such arguments (e.g., Neta 2002, Schaffer 2004) claim that for attributions of epistemic properties to serve the function that they are supposed to serve, they must be semantically sensitive to contexts. For instance, if a knowledge attribution of the form "S knows why p " functions to signal to one's interlocutors that they can trust S on the topic of why it is that p, then whether it is appropriate to make such an attribution depends on whether one should signal to one's interlocutors that they can trust S, and this in turn depends on features of the conversational context. If this appropriateness depends on conversational context because the truth conditions depend on conversational context, then attributor contextualism is true of knowledge-why attributions. Analogous arguments may lead to attributor contextualism for attributions of other epistemic properties as well.

These are some of the main lines of argument that have been adduced in favor of one or another variety of attributor contextualism. Here is a review of those varieties:

According to one group of attributor-contextualist theories (Cohen 1986, 1988, 1999), what varies with the context of attribution is the threshold of evidential support for p that must be exceeded for S to C that p (for S to know, to be justified, to be warranted, etc.). In a particular context x, S 's evidence confers a certain level of epistemic support on the proposition that p. Does that level of support suffice to warrant asserting that S C s that p ? This first kind of attributor contextualism takes the answer to this question to depend on features of the context of attribution y.

According to a second group of attributor-contextualist theories (Dretske 1981, Lewis 1996, Schaffer 2004), what varies with the context of attribution is the range of relevant alternatives to p that S 's evidence must rule out for S to C that p. To say that S 's evidence must rule out these alternatives to p is not to say anything about what S does or does not know, or about what S does or does not believe. Rather, it is to say that S has adequate evidence only if these alternatives to p do not obtain. Epistemologists standardly assume that S 's evidence cannot rule out all alternatives to p —in particular, it cannot rule out the alternative that p is false but S is being deceived by a deceiving spirit into believing that p. But S 's evidence does not need to rule out all alternatives to p for S to C that p. Rather, for S to C that p, S 's evidence must rule out only the relevant alternatives to p. Which alternatives are relevant? That depends upon the context of attribution y.

According to a third group of attributor-contextualist theories (DeRose 1992, 1995; Heller 1995, 1999), what varies with the context of attribution is the range of possible situations throughout which, in order for S to count as knowing that p, p must be true if and only if S C s that p. In a particular context x in which S C s that p, S has the disposition to C that p just in case certain conditions obtain. S C s that p if and only if there is an adequate range of conditions under which S is disposed to C that p. But what range of conditions is "adequate"? That is relative to a context of attribution y.

According to a third group of attributor-contextualist theories (DeRose 1992, 1995; Heller 1995, 1999), what varies with the context of attribution is the range of possible situations in which, for S to count as knowing that p, it is required that p is true if and only if S C s that p. In a particular x, S C s that p just in case certain conditions obtain (such as, for instance, S 's being an authority). What is the range of conditions throughout which this biconditional must hold for S to know that p ? That is determined by the context of attribution y.

A view that combines features of the last two varieties of attributor contextualism is defended by Rieber (1998). According to Rieber, S knows that p if and only if the fact that p explains S 's belief that p. An explanation answers the question "Why?" "Why" questions are contrastive: To ask "Why is it that p ?" is always, at least implicitly, to ask "Why is it that p rather than that q ?" For Rieber, ascriptions of knowledge inherit the contrasts of explanation statements. Thus, for Rieber, for S to know that p is for there to be some contrast proposition q such that S knows that p rather than that q. And for the latter to hold true, on Rieber's account, the fact that p, rather than the fact that q, must explain S 's belief that p. On Rieber's view, then, S knows that p if and only if the fact that p (rather than the contrast proposition q ) explains the fact that S believes that p. On Rieber's view, the context of attribution y determines the contrast proposition.

Finally, according to the most recently espoused version of attributor contextualism (Neta 2002, 2003a, 2004), what varies with the context of attribution is the range of propositions, or of psychological states, that count as part of S 's evidence set. Relative to some contexts of attribution, S 's evidence set may include nothing more than S 's own current states of consciousness. But relative to more ordinary contexts of attribution, S 's evidence set may include various propositions about, say, widely known results of experiments that took place completely independently of S. More generally, according to attributor contextualism concerning evidence, whether S C s that p depends on S 's evidence for p for other epistemic properties C as well (e.g., knowing, being justified, having warrant).

The Recent Controversy over Attributor Contextualism

Since the late 1990s, attributor contextualism has been subject to two sorts of objections. According to the first sort of objection, the problem with attributor contextualism is that it implausibly attributes to native speakers a significant level of semantic self-ignorance. We can see this either by thinking about attributor-contextualist responses to skeptical arguments, as Schiffer (1996) and Rysiew (2001) do, or by thinking about the consequences of attributor contextualism for our practices concerning disquotation of knowledge attributions, as Hawthorne (2003) and LePore and Cappellen (2004) do.

According to attributor-contextualist responses to skeptical arguments, such arguments gain their plausibility because, when going through these arguments, we confuse the propositions that our epistemic-property attributions express with the propositions that these attributions would express in certain other contexts. But, according to the proponents of this objection, it is implausible to claim that native speakers do indeed suffer from this confusion. Again, according to attributor contextualists, attributor A can, in some contexts, truthfully assert something of the following form: "S does not know that p, even though S speaks truthfully when S says, 'I know that p,' " or, more generally, "S is not justified or warranted in C ing that p, even though S speaks truthfully when S says, 'I am justified or warranted in C ing that p.' " But such assertions appear self-contradictory to native speakers. Thus, the attributor contextualist is committed to claiming that native speakers are wrong to think that such assertions are self-contradictory. Once again, the attributor contextualist is committed to attributing a significant level of semantic self-ignorance to native speakers.

In response to this first line of objection, Neta (2003b) raises the question of whether the level of self-ignorance that attributor contextualism posits is any greater than the level of self-ignorance about their own language that native speakers routinely display at other levels of linguistic analysis (e.g., pragmatics, syntax, phonology). If native speakers generally do not realize that there is a difference between the "t" sound in "butter" and the "t" sound in "putter," then why should they realize that there is a difference in the meaning of terms of epistemic appraisal, "C," as they occur in different contexts? Of course, naive speakers can be brought to notice the difference between the "t" sound in "butter" and the "t" sound in "putter." Can the attributor contextualist bring naive speakers (even if not theoretically invested philosophers) to discern a difference in the meaning of terms of epistemic appraisal as they occur in different contexts? This remains an open empirical question.

Stanley (2000, 2004), pursuing the second line of objection, has argued that attributor contextualism is empirically implausible because there is no well-established precedent for the particular kind of semantic context-sensitivity that attributor contextualists posit in our epistemic-property attributions. These empirical arguments have been rebutted most recently by Ludlow (2005) and DeRose (2005), but this issue, like many other empirical issues in semantics, remains unsettled. Indeed, Unger (1984, 1986) has argued that there is no empirically ascertainable fact of the matter as to whether attributor contextualism is true.

Bibliography

Annis, David. "A Contextualist Theory of Epistemic Justification." American Philosophical Quarterly 15 (1978): 213–219.

Austin, J. L. "Other Minds." Proceedings of the Aristotelian Society, supp. vol. 20 (1946): 149–187.

Bach, Kent. "The Emperor's New 'Knows.' " In Contextualism in Philosophy, edited by Gerhard Preyer and Georg Peter. Oxford, U.K.: Oxford University Press, 2005.

Cappelen, Herman, and Ernest LePore. Insensitive Semantics: A Defense of Semantic Minimalism and Speech Act Pluralism. Oxford, U.K.: Blackwell, 2004.

Cohen, Stewart. "Contextualism, Skepticism, and the Structure of Reasons." In Epistemology, 1999, edited by James E. Tomberlin. Atascadero, CA: Ridgeview, 1999.

Cohen, Stewart. "Knowledge and Context." Journal of Philosophy 83 (1986): 574–583.

Cohen, Stewart. "Knowledge, Context, and Social Standards." Synthese 73 (1987): 3–26.

Cohen, Stewart. "How to Be a Fallibilist." In Epistemology, 1988, edited by James E. Tomberlin. Atascadero, CA: Ridgeview, 1988.

DeRose, Keith. "Contextualism and Knowledge Attributions." Philosophy and Phenomenological Research 52 (1992): 913–929.

DeRose, Keith. "The Ordinary Language Basis for Contextualism and the New Invariantism." Philosophical Quarterly 55 (2005): 172–198.

DeRose, Keith. "The Problem with Subject Sensitive Invariantism." Philosophy and Phenomenological Research 68 (2004): 346–350.

DeRose, Keith. "Solving the Skeptical Problem." Philosophical Review 104 (1995): 1–51.

Dretske, Fred. "The Pragmatic Dimension of Knowledge." Philosophical Studies 40 (1981): 363–378.

Fantl, Jeremy, and Matthew McGrath. "Evidence, Pragmatics, and Justification." Philosophical Review 111 (2002): 67–94.

Gendler, Tamar, and John Hawthorne, eds. Oxford Studies in Epistemology. Vol. 1. Oxford, U.K.: Oxford University Press, 2005.

Goldman, Alvin. "Discrimination and Perceptual Knowledge." Journal of Philosophy 73 (1976): 771–791.

Hawthorne, John. Knowledge and Lotteries. Oxford, U.K.: Oxford University Press, 2003.

Heller, Mark. "Contextualism and Anti-luck Epistemology." In Epistemology, 1999, edited by James E. Tomberlin. Atascadero, CA: Ridgeview, 1999.

Heller, Mark. "The Simple Solution to the Problem of Generality." Noûs 29 (1995): 501–515.

Henderson, David. "Epistemic Competence and Contextualist Epistemology: Why Contextualism Is Not Just the Poor Man's Coherentism." Journal of Philosophy 91 (1994): 627–649.

Klein, Peter. "Human Knowledge and the Infinite Regress of Reasons." In Epistemology, 1999, edited by James E. Tomberlin. Atascadero, CA: Ridgeview, 1999.

Lewis, David. "Elusive Knowledge." Australasian Journal of Philosophy 74 (1996): 549–567.

Lewis, David. "Scorekeeping in a Language Game." Journal of Philosophical Logic 8 (1979): 339–359.

Ludlow, Peter. "Contextualism and the New Linguistic Turn in Epistemology." In Contextualism in Philosophy, edited by Gerhard Preyer and Georg Peter. Oxford, U.K.: Oxford University Press, 2005.

Neta, Ram. "Contextualism and the Problem of the External World." Philosophy and Phenomenological Research 66 (2003a): 1–31.

Neta, Ram. "Perceptual Evidence and the New Dogmatism." Philosophical Studies 119 (2004): 199–214.

Neta, Ram. "Skepticism, Contextualism, and Semantic Self-Knowledge." Philosophy and Phenomenological Research 67 (2003b): 396–411.

Neta, Ram. "S Knows That p." Noûs 36 (2002): 663–681.

Preyer, Gerhard, and Georg Peter, eds. Contextualism in Philosophy. Oxford, U.K.: Oxford University Press, 2005.

Rieber, Steven. "Skepticism and Contrastive Explanation." Noûs 32 (1998): 189–204.

Rysiew, Patrick. "Contextualist Solutions to Scepticism." Noûs 35 (2001): 477–514.

Schaffer, Jonathan. "Contrastive Knowledge." In Oxford Studies in Epistemology. Vol. 1, edited by Tamar Gendler and John Hawthorne. Oxford, U.K.: Oxford University Press, 2005.

Schaffer, Jonathan. "From Contextualism to Contrastivism." Philosophical Studies 119 (2004): 73–103.

Schiffer, Stephen. "Contextualist Solutions to Scepticism." Proceedings of the Aristotelian Society 96 (1996): 317–333.

Stanley, Jason. "Context and Logical Form." Linguistics and Philosophy 23 (2000): 391–434.

Stanley, Jason. "On the Linguistic Basis for Contextualism." Philosophical Studies 119 (2004): 119–146.

Stine, Gail C. "Skepticism, Relevant Alternatives, and Deductive Closure." Philosophical Studies 29 (1976): 249–261.

Tomberlin, James E., ed. Epistemology, 1988. Atascadero, CA: Ridgeview, 1988.

Tomberlin, James E., ed. Epistemology, 1999. Atascadero, CA: Ridgeview, 1999.

Unger, Peter. "The Cone Model of Knowledge." Philosophical Topics 14 (1986): 125–178.

Unger, Peter. Philosophical Relativity. Minneapolis: University of Minnesota Press, 1984.

Williams, Michael. Unnatural Doubts: Epistemological Realism and the Basis of Scepticism. Oxford, U.K.: Basil Blackwell, 1992.

Ram Neta (2005)

Encyclopedia of Philosophy