« Linguistic Just-So Stories | Main | Now that's funny! »


TrackBack URL for this entry:

Listed below are links to weblogs that reference The Sapir-Whorf Hypothesis:

» for mitch from the hanged man at askee dot net
Close Range is a new-ish blog by a philosopher at the University of Wyoming. He has an interesting post up on Burge-style externalism and the Sapir-Whorf Hypothesis, which seems very contrary to Mark Rowlands' argument about the Sapir-Whorf Hypothesis ... [Read More]


Kerim Friedman

It is important to note that the version of the Sapir-Whorf Hypothesis you propose here is much closer to the one actually proposed by Whorf than the kind of arguments Mark Abley seems to be making.

Whorf drew from his experience working for an insurance company reviewing insurance claims in developing his theory of how linguistic categories influence behavior. For instance, he argued that someone smoking a cigarette next empty oil canisters was misled by the category "empty" into believing that they did not pose a danger (even though empty canisters are actually more dangerous than full ones, which are sealed).

Unfortunately, a lot of people, including Steven Pinker, have found it useful to portray Whorf as making much stronger claims than he was actually making.

Danny Korman

Marc –

I think you’re exactly right about the possibility of categorially descriptive “water”-words. And I suspect that many linguistic communities *do* have words roughly translatable into English as ‘water’ that express something like the concept of being a wet, transparent, drinkable ... stuff.

That said, if there really are “reference magnets” – that is, properties (or concepts) that are especially eligible to be the referents (or meanings) of the expressions of a language – then the Burgean defense of linguistic relativity might be in trouble. So suppose that our externalist water-concept really is more eligible than the corresponding internalist concept. In that case, while it’s *possible* to introduce a word that expresses the internalist concept, this requires very deliberate stipulatory work in order to prevent the more eligible concept from being meant. On the other hand, when a “water”-word is introduced less cautiously – for instance, by simply pointing at water and saying ‘water’ (‘agua’, etc.) – the most eligible concept in the vicinity wins out.

So, although there are in all probability plenty of linguistic communities that wouldn’t have the Kripkean intuition when you ask them whether their “water”-word would refer to the stuff on XYZ-world, this probably wouldn’t move a proponent of reference magnets (“reference suckers”?) To these guys, intuitions are a pretty good guide to what concept we have, but intuitions must ultimately be weighed against the naturalness or eligibility of candidate meanings of our words in order to determine what's really meant. (That’s why Brian Weatherson thinks that knowledge is justified true belief – he argues for that in his Counterexamples paper.)

So, while it’s *possible* for there to be twin communities that mean twater rather than water by their “water”-word (and likewise for tharthritis), the existence of reference magnets would make it pretty unlikely that there actually are any such communities.



DZ -

There is more room in here than your proposal suggests. Specifically, Lewis articulates his theory of eligibility against a background descriptivist semantics (and associated descriptivist referential intentions). Because of this, it is possible to put to one side referential intentions as part of the eligibility equation--they are assumed to be descriptivist.

However, when we consider the concept of eligibility without such background commitments, it seems to me that it is best construed as a match between referential intentions and the world. Thus, if x has descriptivist intentions, then the most eligible concept will be a "macro-concept"; but if one has nondescriptivist intentions, then the most eligible concept may be a microstructural or de re concept. And so on. [Maybe Lewis and/or Putnam would maintain that the only relevant referential intentions are our intentions "to refer in such a way as to come out right." I am presupposing that we have more fine-grained referential intentions than that--though the story I want to tell here is complex.]

Two questions: (1) Is only one sort of referential story possible for all languages? (2) If not, do natural languages in fact differ with respect to their referential semantics (and associated referential intentions), perhaps in part because of cultural differences? If the answers to (1) and (2) are "no" and "yes" respectively, then the point in the post would seem to hold. [If the answer to both is "no", then the S-W Hyp stands as an unrealized possibility.]

When you say "when a “water”-word is introduced less cautiously – for instance, by simply pointing at water and saying ‘water’ (‘agua’, etc.) – the most eligible concept in the vicinity wins out," I am not sure what your conception of eligibility amounts to. Are some concepts, as it were, intrinsically more eligible to be referents /simpliciter/ (regardless of one's referential intentions)?

Danny Korman

I guess there are various grades of eligibility theories. On one extreme, we have highly fine-grained referential intentions which completely constrain the referents (meanings) of our expressions, and according to which the eligibility of a property (concept) is not intrinsic, but rather varies with our referential intentions. Eligibility only comes into play at the point that our intentions underdetermine the referent.

On the other extreme, certain properties are intrinsically more eligible than others, and manage to become the referents of our terms *regardless* of any of our referential intentions. This of course is too extreme, since on this picture *every* expression ends up referring to the single most eligibile property (i.e., The Good). Less extreme versions maintain that certain properties are intrinsically more eligible (e.g., because the members of their extensions are objectively more similar to one another), but admit that our referential intentions play *some* role in determining what our expression denote. But the intrinsic ineligibility of what we’re intending to refer to might sometimes result in our ending up referring to something that explicitly violates certain of our referential intentions, but which it more eligible. (Cf, again, Weatherson on knowledge and JTB.)

I guess that whether or not eligibility theories pose a problem for the Burgean defense of S-W wholly depends on (a) how fine-grained our referential intentions really are, and (b) the extent to which eligibility is able to contravene referential intentions. I have my doubts about “intention-contravening eligibility”; but it does seem like that’s how some of these reference suckers are thinking about eligibility. And if there is such a thing, then there’s a chance that – despite the mere possibility of languages with different referential semantics – no natural languages differ with respect to their referential semantics (of, say, their ‘water’-terms).

Incidentally, I had in mind intrinsic eligibility when I gave the ‘water’ (‘agua’) baptism case; but it may be that a certain kind of extrinsic eligibility might work just as well. Perhaps, because of general facts about the way our brains are wired, the relevant externalist concept is more eligible to be possessed by us than is the associated internalist concept. Nevertheless, I might, on some dispositional level, intend to express the internalist concept, but (because of the contravening-strength of more eligibile concepts in the vicinity) I might still end up expressing the externalist concept.


Marc and Danny,

First, Marc, congratulations on the new place. Second, Danny when are you going to open shop?

I wanted to know more about this contrast between water and macrowater and the contrast between functional vs. mirco kinds. Are we supposed to think that this categorial distinction is exhaustive? Are we supposed to think that matching the category with the term is related to conceptual and linguistic competency?

If we think of the distinction as exhaustive, then it seems that kinds at the most basic level should be conceived of as functional--but this doesn't seem to be something that fits our actual practice. In fact, those who are competent enough to be ascribed attitudes with the term don't need to know whether 'atom', for example, is a functional kind or not, what its function is, etc., as evidenced by the fact that prior to Dalton, 'atom' was treated as that which played a certain role but subsequently we were able to say before Dalton scientists wrongly thought that atoms were the smallest bit of indestructible matter from which larger bits are made.

viagra online

It is an interesting point of view, at the end it is uncertain.

The comments to this entry are closed.

August 2011

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad