Blog Coordinator

Knobe's X-Phi Page

X-Phi Grad Programs

« How the Experience Machine Works | Main | CfP: Formal Epistemology Meets Experimental Philosophy (Tilburg, 29 - 30 September 2011) »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Jonathan Ichikawa

Thanks for linking this -- an interesting read.

I often have a hard time being confident about just what the target of this critique is supposed to be. The new paper adds some more data with respect to this question, but I'm still a bit confused. I was surprised to read that the "goal has never been to challenge Kripke’s argument against descriptivism." I had always interpreted that as a goal of the MMNS and MMNS projects.

You say now that the goal is broader: it is "to challenge the way philosophers of language go about determining what the right theory of reference is."

What puzzles me is that it's hard to see how the arguments might successfully challenge that latter, without also successfully challenging Kripke's argument. I'm assuming that Kripke's methodology -- whatever its details -- is an example of the way philosophers of language go about deciding between theories of reference. If the experimental data don't impugn Kripke's methodology, then it's hard to see how they could impugn the more general methodology -- especially since Kripke's is the case study discussed.

Is the moral that Kripke's methodology isn't, after all, an instance of the intuition-based method of cases described in the text? If so, is there reason to think that anybody is practising the bad methodology?

Julian Kleinknecht

I think your reply to what Ichikawa and colleagues present as Objection 4 rests on a misinterpretation of Kripke. You claim that Kripke used intuitions as evidence for the claim that 'Gödel' refers to Gödel rather than Schmidt. To support this claim you cite the well-known remark Kripke makes in N&N: 'Some philosophers think that something’s having intuitive content is very inconclusive evidence in favor of it. I think it is very heavy evidence in favor of anything, myself. I really don't know, in a way, what more conclusive evidence one can have about anything, ultimately speaking.' (p.42)

To evaluate this remark one should look at its background. Interestingly, Kripke writes about something having *intuitive content*; he doesn't write about a judgment (like '"Gödel" refers to Gödel') being intuitive. This is because his well-known remark is a reply to the accusation that the notion of necessity is 'just a doctrine made up by some bad philosopher.' (p. 41) Since the notion of necessity plays an important role in the modal argument, one could argue that intuitions play a role in the modal argument against descriptivism.

By contrast, Kripke doesn't use the word 'intuition' when he's concerned with the semantic argument, i.e. when he writes about judgments like '"Gödel" refers to Gödel' or '"Peano" refers to Peano.' He only writes things like:
-'So, since the man who discovered the incompleteness of arithmetic is in fact Schmidt, we, when we talk about "Godel", are in fact always referring to Schmidt. But it seems to me that we are not. We simply are not.' (p. 84)
-'So on the theory in question the term "Peano", as we use it, really refers to -- now that you've heard it you see that you were really all the time talking about -- Dedekind. But you were not.' (p. 85)
-'But we do in fact refer to Godel.' (p. 89)

One could claim that Kripke has to appeal to intuitions to provide evidence for these claims, but Kripke *himself* doesn't do it.

R Mallon

Jonathan: The claim is that "determining the right theory of reference" requires more than showing that some particular version of descriptivism is wrong (even if Kripke accomplishes the latter). As we say in the response, we're skeptical that one can accomplish this broader goal with the sort of intuitions that Devitt and you and your coauthors claim accomplishes the narrower one.

Julian Kleinknecht: I'll think more about the issue you raised with regard to Kripke's remark. But I think it would be incorrect to read our response to objection 4 as resting entirely on a claim about what Kripke took himself to be doing. It looks like Kripke's judgment in the Godel case rests on a spontaneous judgment about the referent of a name in a counterfactual case. And it's that sort of judgment we're worried about the variation of (even if Kripke doesn't call it an "intuition" or "intuitive," etc.).

jonathan weinberg

I would be interested in hearing more about how you would reply to JI's question about how you understand Kripke's methodology. I don't mean this as a challenge, but I'm genuinely curious as to what y'all's take on it is. What parts of his argument do you take to be generally caught in the net of MMNS-type results, and what parts not? E.g., I might hazard a guess that the modal arguments would count as targeted, but predictions about people's dispositions to affirm or retract claims in light of novel evidence (as in Ichikawa et al. mss. p. 4, towards the end of section 1) maybe would not. But what's y'all's thinking about that?

Edouard Machery

Sorry for the slow response!

1. Julian,

I agree with Ron that our response does not depend on Kripke's own view about intuitions. The relevant question is: How are you justified in believing, e.g., that in the situation described by Kripke, "Gödel" refers to the man who stole the theorem.

That said, I wonder what Kripke could mean by "intuitive content" if not a content (e.g., a proposition) that elicits an intuitive judgment when considered. If this is what Kripke means (and, again, what else could he mean?), then it would seem that whether or not some content elicits an intuitive judgment is relevant after all.

2. Jonathan I. If we are right, to the extent that Kripke relies on the intuition-based methodology in this argument against descriptivism, then his argument fails. Now, it does not matter very much whether his argument relies on intuitions, since, as Ron rightly notes, refuting the particular forms of descriptivism considered by Kripke is not equivalent to establishing a theory of reference - it is not even equivalent to refuting all plausible forms of descriptivism.

3. Jonathan W. We are not concerned with the modal arguments, which do not bear how how names refer (metasemantics), but on how their contribution to the propositions expressed by sentences (semantics).

IMO, all the arguments involve in Kripke's criticism of descriptivism relies on intuitions.

Some intuitions ("Cicero" refers to Cicero) seem far more secure than others ("Gödel" refers to the man who stole the theorem in the Gödel case) - I agree about this with Jonathan I and colleagues and with Michael too.

But where we disagree with them is about the importance of the secure intuitions. To establish any theory of reference (in contrast to merely refuting some particular versions of descriptivism), intuitions similar to the Gödel intuition are likely to be used. Furthermore, to undermine some plausible versions of descriptivism, one needs to appeal to those intuitions that are similar to the Gödel intuition.


jonathan weinberg

Hi Edouard,

Hmh, I'm not sure y'all can really go this way here: "We are not concerned with the modal arguments, which do not bear how how names refer (metasemantics), but on how their contribution to the propositions expressed by sentences (semantics)." First, I thought that, in an important sense, you're not concerned with how _any_ of Kripke's arguments are going -- your interest, I thought, is in the "method of cases", wherever it may be. Are the cases used in the modal argument insufficiently case-y? Putting it differently: whether they are or are not what you are _concerned_ with, can they reasonably be taken to fall within the scope of y'all's challenge? If not, then you might need to do a bit more to get around JI's concern that it sounds like maybe only a tiny slice of what's out there in philosophy of language is in peril.

Second, and relatedly, even if the overall argumentative strategy at that point of NN is metasemantic, the _intuitions about cases_ there are not. They are, rather, about things like whether Aristotle was _necessarily_ the teacher of Alexander the great. So I don't think anything about the semantics/metasemantics distinction is going to be relevant here. It might be that the difference between semantic judgments and modal ones would be something to invoke, but (i) the semantic ones are, after all, a sub-species of modal ones (since they are counterfactual judgments), and (ii) given the sorts of linkages between modality and semantics that Kripke is arguing for (as indeed is embodied in the conjunction of the book's title!) it would be a very surprising result if EA intuitions about semantics varied, but about modality they did not. And so variation in the semantic intuitions should give us some prima facie reason to expect some corresponding variation in the relevant sorts of modal intuitions.

Jeff Maynes

I have one small comment about a passage I found unclear:

Ichikawa's "Objection #4" is the argument that Kripke is interested in semantic facts about the reference of proper names, and that he does not make any assumptions about the universality of our intuitions. In response, you make two claims: first, the intuitions are still evidence for his claims about the semantic facts, and second, Ichikawa, et al., are inconsistent (or "not obviously consistent").

Well, according to the Method of Cases, and some remarks in the original MMNS paper, the method under attack is the position that getting the intuitions right is what determines the correct theory of reference. If, as Ichikawa, et al. argue, it is the semantic facts which matter, then while intuitions might be one way to get at them, we do not need the universality assumption (perhaps only a select few judge the facts rightly). This objection would fit quite plausibly with an expertise account, namely that the argument is best understood as Kripke providing an expert's judgment on the relevant facts. If so, then (a) it does matter whose intuitions we are paying attention to, and (b) the intuitions can still be evidence, but their evidential force is based on the expertise, rather than their universality; and these are not only consistent with each other, but they fit Ichikawa, et al.'s arguments. If so, then the response to Objection #4 really is the argument against expertise a bit earlier in the paper, and the inconsistency charge does not stick.

(I take it that the key phrase here is "or otherwise," taken to suggest that for Ichikawa et al., the intuitions do not matter *at all.* But their qualifications of point 4 indicate that they do think expertise matters in intuitions, e.g., asking people whether GB-Theory was an improvement over Aspects).

Edouard Machery


1. The target of Machery et al. 2004 and Mallon et al. 2009 was the method used to determine which theory of reference is correct. We never intended to bring havoc to the whole philosophy of language.

2. Our argument is appropriately circumscribed: so the target is not the "method of cases, *wherever it may be*. First, that intuitions about reference vary does not cast doubt about the use of intuitions in general (e.g. in ethics) because our concerns might not carry over to the use of intuitions in other fields. Furthermore, it is not clear whey the variation of intuitions in the Godel case would suggest that the intuitions used in the modal argument (e.g., it is not necessary that Aristotle was the teacher of Alexander) also vary. So the target just is the use of intuitions about reference (what a proper name refers to, what someone is talking about in some situation, etc.).

You suggest this in your second point, but you note that:

- 'the semantic ones are a sub-species of modal ones". This is unconvincing. Not all intuitions about reference are counterfactual ("Frege" refers to Frege). Second, the counterfactual judgments about reference are not about what is possible or necessary.

- " it would be a very surprising result if EA intuitions about semantics varied, but about modality they did not." I don't see why. The connection between metasemantics and modality is intricate, and I really don't think that one's intuitions about reference have much bearing on the kind of modal intuitions that are relevant for semantic questions.

This seems to be a possible way to understand what Ichikawa and colleagues are asserting. I wish they had put it more clearly.


The comments to this entry are closed.

3QD Prize 2012: Wesley Buckwalter