Abstract
The apparent contextual variability exhibited by ‘knows’ and its cognates—brought to attention in examples like Keith DeRose’s Bank Case—poses familiar problems for conservative forms of invariantism about ‘knows’. The paper examines and criticises a popular response to those problems, one that involves appeal to so-called ‘pragmatic’ features of language. It is first argued, contrary to what seems to have been generally assumed, that any pragmatic defence faces serious problems with regard to our judgments about retraction. Second, the familiar objection that the pragmatic effects at issue do not seem to be cancellable is considered. Advocates of the pragmatic defence have suggested that cancellability concerns can be dealt with fairly readily. It is shown both that their recent attempts to respond to those concerns, and some other possible attempts, are unsuccessful. Finally, it is argued that the popular relevance-based accounts, found in the work of Jessica Brown, Alan Hazlett, and Patrick Rysiew, fail to provide a satisfactory explanation of our judgments.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
To avoid potential dispute, let us also just stipulate that commitment to ‘anti-scepticism’ involves commitment to the claim that, in examples like the ‘Bank Cases’ presented below, the knowledge ascription in the ‘low’ context is true.
See DeRose (1992, p. 913) for his original presentation of the Bank Cases. See also Stewart Cohen’s (1999, p. 58) so-called ‘Airport Case’. Similar judgments seem to arise in lottery cases and in examples involving scepticism about the external world. As our judgments concerning examples like the Bank Cases have been the main explanatory focus of our opponents, we will focus on pragmatic explanations of those judgments here. However, we believe our criticisms pass over in fairly straightforward ways to pragmatic explanations of scepticism and lottery cases. For pragmatic treatments of judgments associated with external world scepticism, see Black (2008) and Pritchard (2010).
In what follows, we are going to be fairly liberal about the attribution of truth and falsity to propositions, sentences, and utterances. As far as we can tell, nothing that we say turns on this.
The example has been modified slightly to make it more explicit.
One salient feature of the Bank Cases is that they are centred on knowledge ascriptions and denials made in the first person—claims of the form ‘I know that \(p\)’ and ‘I don’t know that \(p\)’—rather than in the third person. It might be objected that such ‘first person cases’ do not allow us to distinguish between standard contextualist accounts, which differ from TCP in their rejection of invariantism, and standard forms of anti-intellectualism, which differ from TCP in their rejection of intellectualism. However, given our present purpose is not to decide between the various possible alternatives to TCP, but instead to reveal the implausibility of a particular defence of TCP, this is no objection to our focus on first person cases. Whenever the difference between first person and third person cases becomes important to our critique, we will make this explicit.
Our talk of what is ‘pragmatically conveyed’ or ‘pragmatically implicated’ by an utterance should be interpreted in as theory-neutral a fashion as possible. To make our criticisms broadly applicable, we wish to remain neutral on such issues as whether the advocate of TCP must claim that a conversational implicature, in Grice’s (1975) original sense, is present. However, note that at least Hazlett (2009, p. 609) claims that the pragmatic effect associated with knowledge ascriptions is a conversational implicature.
It might be suggested that it is possible to provide a pragmatic explanation of our Bank Case judgments that does not involve appeal to ‘pragmatic implication’ in any natural sense. This might be appropriate, for example, for the kind of view discussed in DeRose (2002, pp. 177–179; 2009, pp. 89–92) and defended in Pritchard (2010). Our critical arguments could just as readily be used against such a position.
The example is adapted from Grice (1975, p. 33).
Cases featuring utterances of (4) are putative instances of what Bach (1994) terms ‘impliciture’. Brown (2006, pp. 415–416) highlights another (controversial) case, featuring referential uses of definite descriptions, in which we also find similar judgments to those found in the Bank Cases. There are plausibly other (still controversial) cases besides—pragmatic treatments of contextually restricted quantification, for example. The example involving (4) is, however, representative of these cases in all the respects that will concern us here.
See e.g MacFarlane (2005, §2.3).
DeRose may of course attempt to retract but excuse his previous assertion, perhaps by uttering something along the lines of ‘What I said is false, but I was obviously speaking a little loosely’. Such a response seems natural, but that fact actually seems to make the situation worse for the advocate of the pragmatic account. See footnote 17 for relevant discussion.
Similar judgments are reported in Hawthorne (2004, p. 163), Williamson (2005a, §II), MacFarlane (2005, §2.3), Stanley (2005, p. 115), Davis (2007, pp. 399–400), and elsewhere. Notice that this list includes those who defend views apparently impugned by such judgments—for example, advocates of subject sensitive invariantism (SSI), such as Hawthorne and Stanley. Cohen (2001, p. 89; 2004, pp. 191–192), a prominent defender of contextualism, also assumes that our retraction judgments have the kind of character presented. DeRose (2006, pp. 322–323; 2009, p. 161) is an exception. He defends a form of contextualism and reports, in regard to similar cases, that he has the opposite judgments about cases of retraction.
An anonymous referee suggested to us that if DeRose’s wife waits to challenge DeRose until after his friend has left, DeRose might appropriately respond by saying that he knows that the bank will be open, but that he did not feel comfortable saying that to his friend, given how much his friend has at stake. As long as it is granted that there are ways of fleshing out the case such that it seems appropriate for DeRose to retract his earlier assertion and admit that what he said is false, observations like these do not undermine the objection we are about to press. As we pointed out in footnote 14, the claim that there are ways of fleshing out the case such that it seems appropriate for DeRose to retract his earlier assertion seems to be well-supported in the literature.
Note that the advocate of TCP has no special reason to think that knowledge of knowledge is especially hard to come by. We can therefore also assume that in his current context (resembling high) DeRose knows that he knew (when in low) that the bank will be open.
Of course, it may be natural, especially in this kind of case, for DeRose to not only retract his earlier assertion but also attempt to excuse it. In this vein, he might respond to the judge along the following lines: ‘What I said is false, but c’mon! I was obviously speaking a little loosely—I wasn’t in a courtroom back then’ (cf. footnote 13). However, this actually seems to make the situation worse for the defender of the pragmatic account. First, DeRose was not only speaking truly but presumably also strictly in uttering (1) in low, since his earlier utterance of (1) presumably conveyed just what it semantically expressed. Second, such a judgment is nothing like the judgments we find in parallel examples featuring other (putatively) pragmatic phenomena, such as the cases featuring utterances of (3) and (4). In those cases, it seems both unnatural and false for the speaker to utter, in reference to her earlier claim, ‘What I said is false, but obviously I was speaking a little loosely’, and perhaps especially so if the later context is set up to resemble a courtroom environment.
Cohen (1999, 2001, 2004) suggests that we find retraction judgments similar to those found in the case of ‘knows’ in the case of the plausibly context sensitive expression ‘flat’. He therefore argues that those judgments represent little problem for contextualist accounts of ‘knows’. Notice that a parallel response does not seem to be open to the advocate of the pragmatic account. As we have seen, we do not seem to find retraction judgments similar to those found in the case of ‘knows’ in examples of other (putatively) pragmatic phenomena.
Judgments about disagreement have also been thought to raise problems for non-traditional accounts like contextualism. Montminy (2009) argues that these judgments might also represent a problem for pragmatic defences of invariantism. However, such a disagreement-based objection rests on some potentially controversial claims about the presence of disagreement in the relevant cases, and also on a potentially controversial assumption to the effect that it is insufficient for disagreement that there is a contradiction at the level of what is semantically expressed by the relevant utterances. Our retraction based objection fares better in both respects. The objection does not rest on potentially controversial assumptions about disagreement, and the retraction judgments are widely reported in the literature. See footnote 14 for relevant references.
It could be proposed that knowledge ascriptions and denials pragmatically convey something about appropriate action or reasoning. For example, perhaps an utterance of (1) by DeRose in high conveys that he is able to rely on the proposition that the bank will be open in his practical reasoning, while an utterance of (2) conveys that he is not able to rely on that proposition. Whether DeRose can rely on the proposition that the bank will be open in his practical reasoning is obviously broadly connected to his epistemic position, but let us just make clear that we intend the discussion in the present section to apply generally to pragmatic defences of TCP, and, in particular, to apply to pragmatic proposals that appeal to practical reasoning and action just as much as those proposals that involve a more overt connection between what is pragmatically conveyed and the subject’s epistemic position.
The example is drawn from Grice (1975, p. 32).
It should be noted that we are interested in whether a putative pragmatic implication is explicitly cancellable and not just whether it is contextually cancellable. The latter requires only that there is a context in which an utterance of the relevant sentence does not carry the relevant implication.
The point that implicatures are supposed to be cancellable does not apply to so-called ‘conventional implicatures’. We do not seek to criticise a view based on such implicatures here. But note that it is unclear how to develop such an account without encountering the retraction objection pressed in the previous section. For some further objections to a conventional implicature-based story of the sort required to explain what is going on in the Bank Cases, see the appendix of Fantl and McGrath (2007).
It is also worth pointing out that even though Mary’s utterance of (11) does not cancel the relevant pragmatic implication, it is not infelicitous. In that respect, Mary’s utterance of (11) differs from utterances of (8) or (9a) by DeRose in high, which seem infelicitous. This might suggest that we are dealing with different phenomena.
There are some further examples of pragmatic implications that do not seem to be cancellable, but that do not involve irony. As Davis (2007, p. 411) observes, in examples where what is semantically expressed entails what is pragmatically conveyed, the implications cannot be cancelled. Once again, however, there is a clear explanation for why the implication in such examples cannot be cancelled. In attempting to cancel the implication, the speaker would be contradicting herself.
This does not rule out the possibility that there is another and nearly universal pragmatic implication associated with knowledge ascriptions. An anonymous referee suggested to us that an utterance of a sentence of the form ‘S knows that \(p\)’ could universally pragmatically convey that S is able to rely on \(p\). This pragmatic implication might seem to be universal in the same sense as the implication associated with utterances of (13). However, claiming that knowledge ascriptions generally pragmatically convey that S is able to rely on \(p\) does not seem sufficient to explain all of the cases at issue. In particular, that pragmatic effect seems unsuited to explain our judgments about cases that feature an attributor in a situation like high and a subject in a situation like low. For an example of such a case, see DeRose’s (2009, pp. 4–5) case involving Thelma and Louise. In such cases, it can seem inappropriate for the speaker who is in a context that is similar to high to attribute knowledge of the proposition at issue \(q\) to the subject who is in a context that is similar to low even if the case is set up such that it seems appropriate for the speaker who is in the context that is similar to high to assert \(q\). But this cannot be explained by appeal to the idea that the knowledge attribution pragmatically conveys that the subject can rely on \(q\), since, given the subject occupies a context that is similar to low, this implication is presumably true. To amend the proposal to handle cases like these, it seems natural to suggest that utterances of the form ‘S knows that \(p\)’ generally pragmatically convey that S can rely on \(p\) in the kinds of situations relevant in the context of utterance (or something along these lines). But the inclusion of ‘in the kinds of situations relevant in the context of utterance’ raises the concern that this is no more a universal implication than is Rysiew’s proposed implication that S can rule out the salient alternatives to \(p\).
Thanks to an anonymous referee for drawing our attention to Lutz’s proposal.
The fact that John is able to help is necessary for his actually helping. Does this fact somehow render the example featuring John and Mary importantly disanalogous from the example featuring DeRose and his wife in high? It seems not, since, assuming TCP, DeRose knowing (or at least being in a position to know) that the bank will be open is presumably necessary for his occupying an epistemic position such that he and his wife do not need to inquire further in high.
There are many other cases of pragmatic phenomena, besides the example featuring (20) and (21), that exhibit the relevant kind of structure. The example featuring (6), discussed at the start of Sect. 5, arguably represents one such case. In (6), it might seem plausible that the most relevant issue is whether there is an open garage nearby, and not whether there is some garage nearby. Indeed, it seems that the former is what A is really inquiring about in asking her question ‘Is there a garage nearby?’. Nevertheless, as noted at the start of Sect. 5, it seems perfectly felicitous for B to cancel the pragmatic effect supposedly associated with his response in (6), as (7) illustrates.
In reply to this particular case, Lutz (2013) contends that B’s utterance of (7) is likely to be uncomfortable if it is stipulated that B is intelligent and trying to be helpful. However, we do not agree with this contention, even with that stipulation in place. Indeed, (7) seems like a natural and appropriate thing to say in the circumstances (even if the information is not very helpful to A). This is in stark contrast to the situation arising in regard to an utterance of (8) by DeRose in high.
Rysiew (2007, p. 639) offers a somewhat similar proposal to Lutz’s. In the case of (8), Rysiew’s proposal is that the reason an utterance of (8) seems infelicitous is that the speaker is thereby uttering something —viz. ‘We know’—that conveys that she occupies a strong enough epistemic position for present purposes (or something along those lines). But Rysiew then asks why a speaker would choose to utter something that conveys that she occupies a strong enough epistemic position for present purposes, only to then deny that she occupies such a position, as the speaker does by adding ‘but we need to investigate further’. Why would the speaker not simply deny that she occupies a strong enough epistemic position for present purposes? For instance, she could just say ‘We need to investigate further’.
However, in light of the discussion of Lutz’s proposal, there is a fairly straightforward answer to these questions. Assuming TCP, the motivation for DeRose to utter (8) in high is presumably to impart the information that although he knows that the bank will be open, he and his wife nevertheless need to investigate further. As we noted in our discussion of Lutz’s proposal, this information seems like something relevant and useful given the direction of the conversation in high, since it indicates not merely that further inquiry is required in order for DeRose and his wife to safely wait until Saturday to go to the bank, but it also indicates how much further inquiry is called for. In fact, it seems that DeRose might be motivated to convey this information simply because he wishes to first answer the question that is semantically expressed by his wife’s utterance of ‘Do you know that the bank will be open?’, before then answering the question that is ‘really at issue’ in their conversation. In examples of other pragmatic phenomena that exhibit a similar kind of structure, this appears to be something that speakers can comfortably achieve. We conclude that Rysiew’s comments fail to supply a reason to expect an utterance of (8) to be uncomfortable. Similar remarks apply to (9a).
Given TCP, an utterance of (25a) by DeRose in high is very plausibly true. Thus, even if the defender of TCP were to express doubts about the inappropriateness of an utterance of (24) by Mary, it still seems that some explanation is required for why an utterance of (25a) seems infelicitous. None seems to be forthcoming.
Dougherty and Rysiew (2009, 2011) offer an explanation for the infelicitousness of utterances of the form ‘I know that \(p\), but it might be that not-\(p\)’. They suggest that such utterances typically seem infelicitous, despite often being true, because it is natural to assume that a speaker would only mention the possibility that not-\(p\) if the epistemic chance (for the speaker) that not-\(p\) were sufficiently large to be relevant to the conversation. However, Rysiew and Dougherty contend, if that chance were sufficiently large to be relevant to the conversation, it would be large enough to preclude knowledge. Thus, utterances of the form ‘I know that \(p\), but it might be that not-\(p\)’ tend to seem infelicitous because the utterance of the second conjunct can be relevant only if the utterance of the first conjunct is false.
Could this explanation have application in the current dispute over cancellability? The answer seems to be ‘no’. The issue in high seems to be whether DeRose and his wife can safely wait until Saturday to go to the bank—indeed, it seems that DeRose’s wife is issuing the question ‘Do you know that the bank will be open?’ precisely in order to determine whether they can safely wait. But it seems that, given the stakes, the chance (for DeRose) that the bank will not be open is sufficiently large such that he and his wife need to inquire further before they can safely wait until Saturday. The chance that the bank will not be open therefore seems sufficiently large to be relevant to the conversation in high. Nevertheless, it is not large enough to preclude knowledge, since, given TCP, DeRose knows that the bank will be open. It therefore seems that the epistemic chance that the bank will not be open is relevant in high, but it is relevant without being large enough to preclude knowledge (see also Fantl and McGrath’s (2009, pp. 20–23) objection to Rysiew and Dougherty). Rysiew and Dougherty’s account therefore supplies no reason to expect that a cancellation attempt along the lines of ‘I know that the bank will be open, but it might not be’ by DeRose in high should seem infelicitous. For similar reasons, it does not seem possible to extend the explanation to account for the infelicitousness of the other cancellation attempts considered at the start of Sect. 5, such as (8) or (9a). Thanks to an anonymous referee for encouraging us to consider this issue.
Rysiew 2001, p. 495) actually makes the more general claim that utterances of the form ‘I know that \(p\), but I cannot rule out the bizarre alternatives to \(p\)’ seem felicitous.
Brown (2006, p. 428) actually makes the more general claim that utterances of the form ‘S knows that \(p\), but S is not in a really strong epistemic position with respect to \(p\)’ and ‘S knows that \(p\), but S’s belief that \(p\) does not match the facts of the matter in a really distant possible world’ are felicitous.
It might be proposed—as we considered in Sect. 5.1—that the infelicity of (8) is somehow tied to the constraints on appropriate assertion. However, as we pointed out, this kind of explanation cannot be extended to various third-person cases.
One might think that there is no special burden on the defender of TCP to explain the difference between (8) and (29a) and (29b). However, it is worth noting that contextualists about knowledge ascriptions seem able to explain the difference between (8) and (29a) and (29b) as uttered by DeRose in high. Assuming standard forms of contextualism about ‘knows’, and given the circumstances in high, it seems plausible that there would be strong pressure towards interpreting a use of ‘knows’ by DeRose in response to his wife’s question ‘Do you know that the bank will be open tomorrow?’ as one such that if an utterance of ‘I know the bank’ll be open’ by DeRose is true then he does not need to inquire further into the bank’s opening hours. Perhaps the contrast between (8) and 29(a–b) is explained by appeal to the fact that extra effort is required to overcome the pressure to interpret the use of ‘knows’ in this way. There is no obvious way for a non-contextualist to provide an explanation along these lines.
Drawing on a suggestion in Brown (2008), an anonymous referee proposed to us that an utterance of ‘I know, but I need to make sure’ might cancel the pragmatic effect associated with DeRose’s utterance of (1) in high. In response, let us make two points. First, insofar as an utterance of (1) by DeRose in high conveys something about his need to ‘make sure’, it seems to convey that DeRose does not need to do more to make sure that the bank will be open. However, notice that when the proposed cancellation attempt is clarified to make clear that this is what is being denied, it no longer seems felicitous. For example, an utterance by DeRose in high of ‘I know the bank’ll be open, but I need to make sure it will be’ seems infelicitous. This might seem to suggest that an utterance of ‘I know, but I need to make sure’, although felicitous, is not cancelling the pragmatic effect associated with an utterance of (1) by DeRose in high. Second, the same kind of problem that affects Lutz’s proposal will also affect this proposal. In particular, it seems that the pragmatic defender of TCP will have to come up with some explanation for why ‘I know but I need to make sure’ can cancel the pragmatic effect associated with knowledge ascriptions, while utterances of (8) cannot. But no such explanation seems to be forthcoming.
This is one of three constraints that DeRose (1999, § 10–11) proposes on pragmatic explanation. In Sect. 3, we encountered one of the other two constraints, viz. that an appearance of truth and appropriateness cannot be explained by appeal to the claim that a false utterance pragmatically conveys a truth. As we saw, cases like those involving utterances of (4) raise potential problems for that kind of constraint. For critical discussion of DeRose’s remaining constraint, see Brown (2006, pp. 411–413).
The view is defended in Brown (2005, 2006), Hazlett (2009) and Rysiew (2001, 2007). Rysiew (2001, 2007) actually proposes an alternative explanation of those examples, like the Bank Cases, which feature first person knowledge ascriptions, reserving a relevance-based explanation for examples featuring third person knowledge ascriptions. In essence, his alternative explanation is that when one asserts \(p\) one already conveys that one knows \(p\), and so when one asserts that one knows \(p\), one ends up conveying something more—viz. that one can rule out the possibilities salient in the conversation (or something along those lines). However, this explanation cannot, as Rysiew acknowledges, be extended to third person cases. It is also thrown into doubt by DeRose’s (2002) observation that assertions of \(p\) already seem to convey that the speaker can rule out the possibilities salient in the conversation (or something along those lines). We therefore set the explanation aside.
It does not matter whether the proposal is presented in terms of Grice’s (1975) original conversational maxims or in terms of a different set of conversational rules. Thus, while we are going to present the issues using Grice’s familiar machinery, it would not make a difference if we were to couch the discussion in terms of, for example, a so-called ‘neo-Gricean’ framework. For discussion of such frameworks, see e.g. Horn (1984, 1989) and Levinson (2000).
We are using the term ‘knowledge alternatives’ rather than the more familiar ‘relevant alternatives’ associated with the relevant alternatives theory of knowledge. This is to avoid confusion with the notion of conversational relevance.
We are not alone in worrying that the relevance-based explanation of DeRose’s utterance of (1) in high is inadequate. McGrath (2013) also expresses some concerns with the explanation. See e.g. also DeRose (2009, pp. 121–3) for an objection to the relevance-based explanation of the apparent truth of (2) as uttered by DeRose in high. For considerations of space, we will not discuss that objection here. In this section we are going to focus exclusively on DeRose’s utterance of (1) in high.
It also seems that DeRose’s utterance of (1) in high would convey useful information both if it conveyed that he knows that the bank will be open and can rule out that the bank has changed its hours, and if it conveyed that he knows that the bank will be open but cannot rule out that possibility.
There are some similarities between the objection that we are pressing here and the so-called ‘symmetry problem’, familiar from the literature on scalar implicature. See e.g. von Fintel and Heim’s unpublished MIT lecture notes ‘Pragmatics in Linguistic Theory’.
For the purpose of the present discussion we are going to assume that the knowledge alternatives are a subset of the salient alternatives. However, this is only a simplifying assumption and we could just as easily talk about the union of the knowledge alternatives and the salient alternatives.
There is also the issue of accounting for what happens when DeRose’s wife asks her question in high. The invariantist will presumably have to come up with some pragmatic account of how she is able to convey what she does by asking her question. This is not something that can be taken for granted.
The most obvious way to develop a semantic treatment of (6) would be to make use of the resources that have been developed in order to provide a semantic treatment of quantifier domain restriction. The idea would be that the relevant quantifier is restricted in such a way that B’s response is true just in case there is an open garage around the corner. See e.g. Stanley and Szabó (2000) for further details and relevant discussion.
Gauker (2001) proposes an account of examples like (6) which puts a lot of emphasis on real world knowledge. It needs to be acknowledged that Gauker is engaged in the more ambitious project of attempting to provide an alternative to a more traditional implicature-based account of (6). However, while we do not wish to endorse such a view here, we are open to the possibility that real world knowledge could have an important role to play when it comes to making sense of examples like (6). A view that puts a lot of weight on real world knowledge would seem to predict that B’s response would not convey that the garage is open if, for instance, the conversation takes place late at night in a region in which it’s common knowledge that garages are unlikely to be open at that time (though it may well convey that, for all B knows, it is open). For what it is worth, we do not take this prediction to be obviously incorrect, but we do not wish to take a stand on the issue here.
Thanks to an anonymous referee for making a suggestion along these lines.
See e.g. Gazdar (1979) and Levinson (1983) for relevant discussion of so-called ‘clausal implicatures’. An example of a clausal implicature is the implicature generated by an utterance of (30a) to the effect that the speaker does not know that the bank will be open. Gazdar and Levinson give more detailed criteria for the derivation of this kind of implicature. Among the factors they mention as being relevant to the derivation of the implicature is whether the verb in question is factive, in the sense that the truth of its complement is entailed or presupposed. This might be a reason for distinguishing utterances involving ‘knows’, like (1), from utterances involving ‘believes’, like (30a). It is complications like this that make us wary about taking on commitments regarding the pragmatic mechanisms at work.
References
Bach, K. (1994). Conversational impliciture. Mind & Language, 9, 124–162.
Black, T. (2008). A warranted-assertibility defense of a Moorean response to acepticism. Acta Analytica, 23, 187–205.
Brown, J. (2005). Adapt or die: The death of invariantism? Philosophical Quarterly, 55, 263–285.
Brown, J. (2006). Contextualism and warranted assertibility manoeuvres. Philosophical Studies, 130, 407–435.
Brown, J. (2008). Subject-sensitive invariantism and the knowledge norm for practical reasoning. Noûs, 42, 167–189.
Cohen, S. (1999). Contextualism, skepticism, and the structure of reasons. Philosophical Perspectives, 13, 57–89.
Cohen, S. (2001). Contextualism defended: Comments on Richard Feldman’s ‘sceptical problems contextualist solutions’. Philosophical Studies, 103, 87–98.
Cohen, S. (2004). Contextualism and unhappy-face solutions: Reply to Schiffer. Philosophical Studies, 119, 185–197.
Davis, S. (Ed.). (1991). Pragmatics: A reader. Oxford: Oxford University Press.
Davis, W. (2007). Knowledge claims and context: Loose use. Philosophical Studies, 132, 395–438.
DeRose, K. (1992). Contextualism and knowledge attributions. Philosophy and Phenomenological Research, 52, 913–929.
DeRose, K. (1999). Contextualism: An explanation and defense. In J. Greco & E. Sosa (Eds.), The Blackwell guide to epistemology (pp. 187–205). Oxford: Blackwell Publishing.
DeRose, K. (2002). Assertion, knowledge, and context. The Philosophical Review, 111, 167–203.
DeRose, K. (2006). Bamboozled by our own words: Semantic blindness and some arguments against contextualism. Philosophy and Phenomenological Research, 73, 316–338.
DeRose, K. (2009). The case for contextualism. Oxford: Clarendon Press.
Dougherty, T., & Rysiew, P. (2009). Fallibilism, epistemic possibility, and concessive knowledge attributions. Philosophy and Phenomenological Research, 78, 123–132.
Dougherty, T., & Rysiew, P. (2011). Clarity about concessive knowledge attributions: a reply to Dodd. Synthese, 181, 395–403.
Fantl, J., & McGrath, M. (2007). On pragmatic enchroachment in epistemology. Philosophy and Phenomenological Research, 75, 558–589.
Fantl, J., & McGrath, M. (2009). Knowledge in an uncertain world. Oxford: Oxford University Press.
Gauker, C. (2001). Situated inference versus conversational implicature. Noûs, 35, 163–189.
Gazdar, G. (1979). Pragmatics: Implicature, pressupposition, and logical form. New York: Academic Press.
Grice, P. (1975). Logic and conversation. In P. Cole & J. Morgan (Eds.), Syntax and Semantics 3 (pp. 41–58). New York: Academic Press. Page references are to the reprint in Grice 1989, p. 22–40.
Grice, P. (1978). Further notes on logic and conversation. In P. Cole (Ed.), Syntax and semantics 9 (pp. 113–128). New York: Academic Press. Page references are to the reprint in Grice 1989, p. 41–57.
Grice, P. (1989). Studies in the way of words. Cambridge, MA: Harvard University Press.
Hazlett, A. (2009). Knowledge and conversation. Philosophy and Phenomenological Research, 78, 591–620.
Hawthorne, J. (2004). Knowledge and lotteries. Oxford: Clarendon Press.
Horn, L. (1984). Toward a new taxonomy for pragmatic inference. In D. Schiffrin (Ed.), Meaning, form and use in context (pp. 11–42). Washington D.C.: Georgetown University Press.
Horn, L. (1989). A natural history of negation. Chicago: University of Chicago Press.
Levinson, S. (1983). Pragmatics. Cambridge: Cambridge University Press.
Levinson, S. (2000). Presumptive meanings. Cambridge, MA: MIT Press.
Lutz, M. (2013). The pragmatics of pragmatic encroachment. Synthese. doi: 10.1007/s11229-013-0361-6.
MacFarlane, J. (2005). The assessment sensitivity of knowledge attributions. In T. S. Gendler & J. Hawthorne (Eds.), Oxford studies in epistemology (Vol. 1, pp. 197–233). Oxford: Oxford University Press.
MacFarlane, J. (2011). Relativism and knowledge attributions. In S. Bernecker & D. Pritchard (Eds.), Routledge companion to epistemology (pp. 536–544). London: Routledge.
McGrath, M. (2013). Two purposes of knowledge attribution and the contextualism debate. In J. Greco & D. Henderson (Eds.), Epistemic evaluation. Oxford: Oxford University Press.
Montminy, M. (2009). Contextualism, invariantism and semantic blindness. Australasian Journal of Philosophy, 87, 639–657.
Nagel, J. (2010a). Knowledge ascriptions and the psychological consequences of thinking about error. The Philosophical Quarterly, 60, 286–306.
Nagel, J. (2010b). Epistemic anxiety and adaptive invariantism. Philosophical Perspectives, 24, 407–435.
Pritchard, D. (2010). Contextualism, scepticism and warranted assertibility maneuvers. In J. K. Campbell, M. O’Rourke, & H. Silverstein (Eds.), Knowledge and skepticism (pp. 85–104). Cambridge, MA: MIT Press.
Richard, M. (2004). Contextualism and relativism. Philosophical Studies, 119, 215–242.
Rysiew, P. (2001). The context-sensitivity of knowledge attributions. Noûs, 35, 477–514.
Rysiew, P. (2007). Speaking of knowing. Noûs, 41, 627–662.
Sadock, J. (1978). On testing for conversational implicature. In P. Cole (Ed.), Syntax and semantics 9 (pp. 281–297). New York: Academic Press. Page references are to the reprint in Davis 1991, pp. 365–376.
Stanley, J. (2005). Knowledge and practical interests. Oxford: Clarendon Press.
Stanley, J., & Szabó, Z. (2000). On quantifier domain restriction. Mind and Language, 15, 219–261.
Weiner, M. (2006). Are all conversational implicatures cancellable? Analysis, 66, 127–130.
Williamson, T. (2005a). Contextualism, subject-sensitive invariantism, and knowledge of knowledge. Philosophical Quarterly, 55, 213–235.
Williamson, T. (2005b). Knowledge, context, and the agent’s point of view. In G. Preyer & G. Peter (Eds.), Contextualism in philosophy: Knowledge, meaning, and truth (pp. 91–114). Oxford: Clarendon Press.
Acknowledgments
Earlier versions of this paper were presented at the Friday Graduate Seminar (University of St Andrews, December 2009) and the CSMN Colloquium (University of Oslo, February 2010). We are grateful to the audiences on those occasions for useful comments and criticisms. We would also like to thank Jessica Brown, Herman Cappelen, Patrick Greenough, Daniele Sgaravatti, and two anonymous referees for helpful discussion and comments.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Dimmock, P., Huvenes, T.T. Knowledge, conservatism, and pragmatics. Synthese 191, 3239–3269 (2014). https://doi.org/10.1007/s11229-014-0442-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-014-0442-1