There is Some Truth in That

5 Responses to “Review of DeRose’s Case for Contextualism”

1. on 08 Dec 2009 at 2:44 pm1Keith

   Dear Jonathan,
   A Google search of some of my favorite topics brought me to your blog & then to your Review of DeRose. That caught my eye! When you wrote “Comments welcome,” I doubt you had me in mind. But, anyway, just a few quick (quite pro-DeRose, I imagine) comments, maybe helpful, maybe not. You of course have to consider the source…
   First, thanks for your interest in this work, and for taking the time to so carefully engage with it — and also for the kind words you have about it.
   On my use of the ‘best’ — most contextualist-friendly — cases (pp. 3-4 of your review): While there are places toward the beginning of Ch. 5 where this may be questionable, I would have thought the two specific examples you give are places where it’s fairly unproblematic for me to use such cases. P. 56 is in Chapter 2, where I’m giving the basic pro-contextualist case from pairs of examples, and describing what kind of examples to use to make the case the strongest. What’s illegitimate about making the case as strong as it can be? There will be case pairs where it isn’t clear that the epistemic standards vary the way the contextualist claims they can — and case pairs where on the best version of contextualism they actually *don’t* vary — the contextualist of course needn’t hold there will always be a noticeable variance in standards between any pair of cases we might come up with. But if there are cases where it is clear that there is a variance of the type the contextualist allows but invariantists deny can ever happen, those are the case for the contextualist to (quite legitimately, it would seem) focus on in making his positive case. So, maybe we’re just seeing this differently, but my reason for attending to such best cases, given on pp. 53-54 of the book, just seem like good sense to me (but consider the source…) Because of the nature of their positions, invariantists have to boldly deny there is contextualist variation (& here I mean variation of the type invariantists can’t accept–so we’re talking mostly about 3rd person cases) in all of the case pairs: “The best case pairs will differ with respect to as many of the features that plausibly affect the epistemic standards, and especially those features which most clearly appear to affect epistemic standards, as possible. It’s about such pairs of cases that the intuitions supporting contextualism will be strongest. And, given the content of their position, invariantists must resist the intuitions supporting contextualism even with respect to those strongest cases. By contrast, it’s of course open to contextualists to hold that there is no difference in truth-conditions between the cases in some of the case pairs.”
   P. 164 concerns Hawthorne use of DSK, a very general principle. When someone declares and then starts swinging as a weapon some general principle — suppose someone had declared “All Fs are G” –, especially when the only basis he gives is a couple of barely described examples in which the principle seems to hold, does it not behoove us to consider some other Fs, and particularly some of them that look the least G-like? That’s just a matter of checking general principles against the cases that are most challenging to them.
   On a related matter, I was surprised by your closing paragraph about comparative subtlety. It’s not that I didn’t think I was open to criticism on the matter of level of subtlety. But I thought the problem might be that I was comparatively *too* subtle. The above dispute over DSK and its use is a case in point. John makes some very quick remarks, considers a couple of barely described cases, and then BAM!, out comes this extremely powerful general principle that he’s out to use as a sledgehammer. I’m then all like, “Wait. What about these cases where it doesn’t seem so right to disquote? Maybe you don’t want to use such a blunt weapon, but instead make the case that, even if there are cases where disquotation involves seems problematic, there are cases where it seems clearly right, but it’s being so right may seem surprising on contextualism. Now, to evaluate that better objection, let’s consider what happens in parallel situations involving clearly context-sensitive terms…” That’s from Chapter 5, but similar situations are constantly arising in Ch. 7 — John reaching very quickly (from my pov) for simple & powerful (sledgehammer) principles, and me saying, “Not so fast! There are different kinds of cases to consider…”. It’s possible to get too subtle — and even *way* too subtle — when one takes the approach I do to some of these issues, and I was expecting those who tend to side with John H to think that *that’s* what happened to me.
   On my use of intuitions. Of course, I use them, but I worry that people might get the wrong idea. That’s my fault for not heading off the wrong idea more forcefully. But I myself actually take a dim view of the use of epistemic intuitions (intuitions about whether subjects do or don’t know things, or do or don’t have justified beliefs, in various examples) in much recent epistemology. The intuitions I use tend to be semantic ones — about whether certain claims about what is “known” are true or false, whether pairs of such claims are compatible with one another, etc. (Some explicit discussion of this in note 2 of Chapter 2.) And I get fairly explicit about the conditions under which such claims are, and are not, to be trusted. Many of my appeals to intuitions that can sound most problematic out of context are defensive uses in Chapter 5: There I’m responding to objections based on putative intuitions being wielded against contextualism; maybe I should take an x-phi survey, but that charge should perhaps be made in the first instance against the original objection. On the topic of x-phi, most of the studies I’ve seen, given the questions they use — about whether the subjects know, rather than about whether claims made by speakers are true or false (compare with that 2nd note of Chapter 2) –, seem to be testing, if anything, anti-intellectualism, rather than contextualism. As an intellectualist myself, it’s a bit disconcerting to hear the results reported as problems for contextualism. I think I’ve seen one study that asks the right kind of question to test contextualism, but doesn’t set up the situation the right way, at least for my particular contextualist views. So it did seem to me premature to comment on the x-phi situation. Josh Knobe tells me he & Jonathan have found a way to raise the salience of error possibilities so as to get somewhat contextualist-friendly results (though they don’t ask the right kind of question, either, from my point of view, if I’m remembering right) — though they think, at least in the draft I saw, that these results most strongly confirm the contrastivist version of contextualism. About that I disagree.
   I did find this part of the review a bit odd: “On this [DeRose's] view, there will be many true sentences of the
   form ‘S knows p but isn’t in a good enough epistemic position to assert it’ and ‘S is in a good enough epistemic position to assert that p, but doesn’t know it.’ These will arise in cases in which judges and speakers speak in relevantly divergent conversational contexts. I worry that any principle predicting the truth, and indeed propriety, of these sentences cannot plausibly be identified with the knowledge norm of assertion. (DeRose does offer some discussion of this worry much later in the book; again, one could wish for a more unified treatment.)” I work hard to argue that my view does not have the consequences you cite — I guess that’s what you’re acknowledging in your parenthetical sentence. Is the worry that my argument at pp. 244-257 is not successful, or just that it’s in the wrong place? If the problem is just that that argument is in the wrong place, then the way you write what precedes the parenthetical sentence is a bit strange.
   Well, I’ll stop there. Like I said, maybe helpful, maybe not. Thanks again for the very well thought-out review. When it happens that very smart philosophers think hard and carefully about one’s work is one of the most gratifying aspects of this job. –Keith

2. on 08 Dec 2009 at 6:07 pm2Jonathan

   Hi Keith — I can’t imagine many people whose comments I’d be more interested in; thanks for stopping by! I’m sorry that you and I haven’t yet crossed paths; you’re definitely somebody I’d like to spend some time talking to. (I’ve been developing, over the past couple of years, a sort of new-Lewisean infallibilist contextualism about ‘knows’.)

   Here are a couple of quick notes on your notes. (I don’t have your book handy at the moment, so I can’t go back to the citations you mention just now.)

   On the ‘best cases’ treatment: what you say here is helpful. So the idea is, if the relevant shiftiness in truth-conditions occurs anywhere, that’s good for the contextualist and bad for the invariantist, but no particular instance of failing to find it is bad for the contextualist. I’m inclined to think that’s close to right. As you say, the invariantist is committed to a universal generalization; the contextualist can refute it by looking in the best places. That all sounds good.

   What I was worried about was an additional methodological choice that I thought might lie behind the focus on best cases: in addition to the way the situations are selected, there is a question of how they are described. Take a particular case; how we talk about it can affect our intuitions about it. (Suppose we describe some ‘low-standards’ case thus: “and even though S has no way of ruling out the possibility that the bank has changed its hours in the last week, she says ‘I know the bank will be open tomorrow’.” You want us to intuit that her sentence is true, but I’ve just stacked the deck in an invariantist-friendly way, that makes that intuition less likely. Maybe you weren’t claiming the rights to make these kinds of decisions when you talked about best cases; maybe it was just about what cases we discuss, not about how we talk about them. (Certainly the particular features of best cases I quote you as discussing are about selection, not description.) But I still think there’s an important methodological issue here.

   This bears on some of the comments I made at the end of the review about the kind of methodological picture you’re relying on. You put a lot of weight on ‘ordinary intuitions’. But who gets to decide what circumstances are the ones conducive to ordinariness? You’ll probably say of the description above that it’s illegitimately invariantist-friendly; but by what grounds do we adjudicate these claims?

   On a related matter, I find this remark a bit puzzling:

   I myself actually take a dim view of the use of epistemic intuitions (intuitions about whether subjects do or don’t know things, or do or don’t have justified beliefs, in various examples) in much recent epistemology. The intuitions I use tend to be semantic ones — about whether certain claims about what is “known” are true or false, whether pairs of such claims are compatible with one another, etc.

   I have a hard time making sense of this suggestion. I don’t think I believe in the distinction you rely on here between ‘epistemic intuitions’ (i.e. intuitions about what is known) and ‘semantic intuitions’ (i.e. intuitions about what ‘knows’-sentences are true). For it to seem to me as if S knows p just is for it to seem as if “S knows p” expresses a truth in my context.

   On x-phi: you’re probably right about those particular arguments. I tend to agree with you that they’re pretty unpersuasive. I just thought it worthy of mention, given that the claims made there bear so heavily on the kinds of evidence you rely on.

   Is the worry that my argument at pp. 244-257 is not successful, or just that it’s in the wrong place?

   I’m going to have to defer answering this question until I’ve had a chance to have another look at that bit of the book.

   By the way, you might be interested in some of the other recent thoughts I’ve written up about your project over the past few weeks; you can actually click on the ‘DeRose’ tag for a quick link to related posts. I had quite a lot of thoughts — many more than could fit into this review. So I’ll be writing some more related things soon. I’ll probably post thoughts and/or links to material here in my blog; maybe, if you’re willing, I’ll email you some thoughts along the way, too.

3. on 09 Dec 2009 at 12:36 pm3Keith

   I don’t think I believe in the distinction you rely on here between ‘epistemic intuitions’ (i.e. intuitions about what is known) and ’semantic intuitions’ (i.e. intuitions about what ‘knows’-sentences are true). For it to seem to me as if S knows p just is for it to seem as if “S knows p” expresses a truth in my context.”

   Well, ok, but it may be easier to discern the truth-conditions and truth-values of ‘knows’ claims in some contexts than in others. (Or maybe it’s sometimes worse than that in the difficult cases: maybe in some cases it isn’t that there are objectively clear standards in place but they’re difficult to discern, but rather that there are no objectively clear standards in place (omniscient God doesn’t even know).) And maybe philosophical discussions about epistemology — say, epistemologists discussing/arguing about the analysis of knowledge and in connection with that talking about what is and is not “known” in various imaginary cases — is one of the types of cases where it’s at least difficult to discern the truth-values & truth-conditions of ‘knows’ claims (for reasons I won’t go into here). (And maybe some freshman in college taking an x-phi survey, being asked on paper whether a certain subject ‘knows’ something or not, is an example of another kind of unclear case.) Then, when in such a context, our judgments about ourselves and our own present claims (judgments at both levels) may be very untrustworthy. But there may be other cases — cases where it’s clear why the the subjects are saying that they or other subjects do or don’t know (what their conversational purposes are, what’s at stake, etc.), and what of relevance has been said so far, etc., and all of these do point together quite clearly to a fairly determinate set of standards, and so we are able to make relatively secure judgments about the truth-values of these (usually imaginary) speakers’ ‘knows’ claims. If someone were inclined to accept something like the above (as I am), that could then make pretty good sense of their being very suspicious of many of the uses of epistemic intuitions made in recent epistemology, while being more bullish about various semantic intuitions about the truth-values of certain ‘knows’ claims used in certain exercises in the philosophy of epistemological language.

4. on 10 Dec 2009 at 6:39 am4Jonathan

   Thanks, Keith — that’s helpful. I agree that one could find reason to find some kinds of intuitions more reliable than others, and that if one found that pattern in the way you suggest, it would make good sense of the kind of methodological preference you express. Can you say a bit about why you think that’s the way the patterns should play out?

5. on 19 Dec 2009 at 10:07 am5Jonathan

   Hi Keith, on the 244-47 passage you mentioned: it seems that you’re developing (in the part where you consider a way for the contextualist to avoid making sentences like the ones I’m considering true) a sort of contextualist ‘parasitic’ strategy according to which the speaker inherits features of the subject’s practical situation. But it seems that this does not have a plausible general application. Consider a case where the speaker is in a nonskeptical context and the subject is in a skeptical situation, but the speaker is unaware of this latter fact. Now consider the sentence “the subject knows but ought not assert.” I don’t see how you keep it from being true in the speaker’s context. It is, I submit, implausible that the subject’s hidden situation affects the conversational context in the way the strategy of 246 requires.

   You go on to suggest that maybe sentences like these should be considered true after all. I don’t agree, but that’s not my sticking point here — if you want to argue that there are truths like that, that’s fine by me, and a reasonable way to resist Hawthorne’s argument against contextualism. My point was just that doing so just doesn’t seem to be a way of accepting the knowledge norm of assertion.

Leave a Reply

Name (required)

Mail (hidden) (required)

Website