A Note on the Nature of 'Water'

Citation preview

A Note on the Nature of "Water" Barbara Abbott Mind, New Series, Vol. 106, No. 422. (Apr., 1997), pp. 311-319. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28199704%292%3A106%3A422%3C311%3AANOTNO%3E2.0.CO%3B2-2 Mind is currently published by Oxford University Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri Jun 22 07:36:36 2007

A Note on the Nature of "Water" BARBARA ABBOTT

1. Introduction In his essay on the study of language Chomsky (1995) notes that substances such as tea and Sprite, although they (are believed to) contain roughly the same proportion of H 2 0 molecules as tap water, are nevertheless not called "water". In so doing he wants to challenge the essentialist semantics of Putnam (1975) and Kripke (1972), according to which reference for words like "water" is determined by internal structure properties, and suggests instead that "special human interests and concerns" (p. 22) play a role in categorization in this case no less than in the case of artifact terms. Coincidentally, in a paper titled "Water is not H20" Malt (1994) has argued along very similar lines. Both Chomsky and Malt echo concerns expressed by Wilson (1982, p. 578). However, while Wilson (1982) generally supported Putnam's psychologically externalist semantics, and Malt remains neutral on this issue, Chomsky's remarks are part of an extended plea for psychological internalism. In this paper I want to reassert the claim that water is H20, and respond in a way that is consistent with that fact to Chomsky's observations. This requires explaining why substances which are largely H 2 0 may be called something else (§3), and explaining why what we call "water" does not have to be pure H 2 0 (94). But first I want to clarify the extent to which the conclusions of Putnam (1975) have been challenged.

2. Putnam Putnam's main and best known claim in "The Meaning of 'Meaning"' (1975) was that typically unknown internal structure properties are necessary in fixing the reference of natural kind terms-atomic or molecular structure in the case of words for substances, and genetic makeup in the case of plant and animal species names. Putnam's famous twin earth thought experiments were designed to show that something with the external, readily observable properties of, say, a tiger or water, would nevMind, Vol. 106.422 .April 1997

O Oxford University Press 1997

312 Barbara Abbott

ertheless not be a tiger or water unless it had the right internal structure. This is true even when, as is usually the case, ordinary speakers of the language do not know what the internal structure properties actually are. To overturn this result one would have to show that internal structure properties are not necessary in the determination of the extension of a natural kind term-for example that there could be a substance that is considered by people to be water even though it doesn't have the right internal structure. I want to make clear that this fundamental aspect of Putnam's view is not at issue here. Chomsky's (and Malt's) main attacks concern instead the sufficiency of internal structure to determine extension. A note on the phrase "natural kind": in an everyday totally compositional sense it may refer to kinds that are found in nature as opposed to ones that are manufactured by people. However this phrase (and correspondingly "natural kind term") has acquired a special use following Putnam (1975) and Kripke (1972). In the latter use it does not apply to all kinds of things found in nature, but rather to species of plants and animals, well-defined substances such as gold and molybdenum, and natural phenomena like heat and lightning. These are things whose names (natural kind terms) were argued by Kripke to be nondescriptional. So although mud and bugs exist in nature, they are not natural kinds in the technical sense, and the words "mud" and "bug" are not natural kind terms in this sense. When Malt is considering whether the meanings of natural kind terms are in fact more similar to those of artifact terms than is commonly assumed, she points to examples like trees and vegetables (pp. 670; but these also are not natural kinds (and "tree" and "vegetable" are not natural kind terms) in the technical sense that is relevant to issues of word meaning. Chomsky casts aspersions on Putnam's twin earth thought experiments (despite the fact that he has not challenged their main result), suggesting that the intuitions are those of philosophers and not normal language users. However Keil(1989) reports results of the same type as Putnam's that were obtained with 10 year olds, whose ears are probably "not previously contaminated by philosophical theory" (to use the words of Stich (1983), cited by Chomsky in this connection (1995, pp. 42-3)). Chomsky also complains that normal speakers do not have intuitions about theoretical semantic relations like reference and extension (1995, p. 42), but to get Putnam's results one does not need to use terms like "reference" or "extension" which, indeed, ordinary speakers may not be familiar with in their technical uses. All one needs to ask is whether, for example, under such-and-such conditions such-and-such a substance would be water or such-and-such an animal would be a tiger. In this respect gathering semantic data is no different from gathering syntactic data, where native speakers who do not (consciously) know anything about Principle B of the

A Note on the Nature of "Water" 313

Binding Theory will nevertheless agree with Chomsky's intuition that somebody who says "John expects to like him" is saying that John expects to like some male person other than himself (cf. Chomsky, 1995, p. 36). In addition to the well known twin earth result, Putnam put forward a more complete theory of word meaning. He suggested a multi-part "normal form" for the description of word meaning which would include not only a specification of the extensiow-in the case of "water", "H20 (give or take impurities)"---but also "semantic markers" ("natural kind; liquid) and a "stereotype" ("colorless; transparent; tasteless; thirst-quenching; etc.") (Putnam, 1975, p. 269). Putnam was clearly aware of the fact that what we call "water" may (be believed to) contain impurities, which is one of the points that is stressed by Chomsky. His (and Malt's) other main observation in this regard is that many substances, such as tea, clam juice, blood, and Windex, which are correctly believed to contain a high percentage of water, are nevertheless not called "water". None of these substances have the stereotypical properties of water and hence to that extent they would not fit the meaning of "water" under Putnam's approach. Some loose ends remain here however: Putnam's stereotypical features are given ad hoc and it is not clear what official role they are to play in meaning. Putnam's theory is less radical than Kripke's view, according to which "water" and other natural kind terms are nondescriptional. That means that they do not have a meaning in the traditional Fregean sense. Kripke did not include any features, stereotypical or otherwise, in his account of the semantics of natural kind terms. (Although internal structure properties determine the reference of natural kind terms, they are not part of meaning in Kripke's view---there was no change in meaning felt when these properties were discovered by scientists.) In fact Kripke's approach will be supported over Putnam's in what follows.

3. Why tea is not called "water" Why don't we use the term "water" for substances like tea, Windex, and Sprite, which people believe to consist partly or almost entirely of water? Both Chomsky and Malt suggest that it is because of the existence of other features which are involved in the concept of "water"-features such as source, location, and function, which reflect "special human interests and concerns" (Chomsky 1995, p. 22). One immediate problem with claiming that such features actually belong to the concept of water is that different types of water differ on these features. Water can come out of faucets, fall out of the sky, or be bought at the store. It can be located in bottles, or pipes, or just be lying around on the surface of the earth in various quan-

314 Barbara Abbott

tities. As for function, or the role of water in the lives of humans, it is difficult to think of any substance which has more different kinds of function. And the other side of this coin is not surprising either. Consider Evian or Perrier. These kinds of water, like beer or Sprite, come from a bottling plant, are kept in the refrigerator, and are used only for drinking. (Ordinarily, that is--of course you could wash your hair in them if you wanted to, but you could also wash your hair in beer or lemon juice.) Thus they have the same kind of source, location, and role in our lives, as substances which are not called "water" and hence these features cannot explain why beer or Sprite are not called "water" while Evian and Perrier are. In contrast to the functional features, Putnam's stereotypical features of colorlessness, transparency, and tastelessness do a good job in these cases. Thus unlike beer or Sprite or lemon juice, or tea, or Windex, Perrier and Evian look and taste like water. It is interesting to note why these features work better than the type of considerations proposed by Malt and Chomsky: it is apparently because they are a better clue to internal structure properties. Small as the proportion of additives or other substances in beer or radiator coolant are believed to be, it is their presence that seems in these cases to be crucially responsible for what these things are called and they are what give the substance its distinctive taste and appearance. We will return to this fact below. Other intuitions support the conclusion that functional and locational properties do not seem in general to do a good job of predicting what is called "water" and what is called by another name. Were a magician to change the water in the Mississippi to Sprite or wine we would no longer call it "water" despite its location and its use for floating barges. Blood is the same stuff whether it is in someone's veins, in a plastic bag at the blood bank, or lying in a puddle in the street. We could even have an ocean of calamine lotion. These intuitions go contrary to those of Chomsky, who holds that if a giant tea bag were installed at the water processing plant, so that what came out of people's faucets was indistinguishable from tea, it would nevertheless be called "water" (Chomsky, 1995, pp. 22-3). Here I believe Chomsky is in the minority--possibly he has missed all those grade B horror movies in which blood pours out of a shower head. We have considered substances such as Sprite, celery juice, and Windex which consist to a greater or lesser extent of water but which contain distinguishing additives. Of course these are not natural kinds, and the terms for them are descriptional.' For these kinds of terms, then, we might I It has been convincingly argued that artifact terms are to be distinguished from natural kind terms in this respect. See e.g. Goosens (1977) and Schwartz (1978), as well as Abbott (1989) and the references cited there. These arguments respond to the contrary view expressed in Putnam (1975) and Wilson (1982) on this issue.

A Note on the Nature of "Water" 315

expect features involving human interests and concerns to figure in their semantics. It is interesting that in these cases apparently it is the nature of the substance instead that is crucial. Of course that is possible since the internal structure of e.g. Windex did not have to be scientifically discovered but was known from the start. And perhaps it should not be surprising that the internal structure of substance artifacts is crucial to their function, and hence possibly part of their nominal essence, since as substances they do not maintain distinct shapes. As the perceived proportion of water in a substance increases, we go from substances which we would judge to consist partly or mostly of water to substances whose internal structure is completely that of water, but which have other types of properties by which they are classified. Tea and coffee come very close to being in this situation. Paul Bloom has suggested to me that it may be the case that tears are in fact, or are believed to be, of exactly the same chemical constitution as the Pacific Ocean. Another example, which would work better in English if there were a special word for it, is holy water. Does this mean that I must take back my claim of nondescriptionality for "water"? If the same substance is called "tears" when it comes out of someone's eyes, but "water" when it is in a massive quantity lapping up against the coast of California, doesn't that mean that the property of source is forming part of the concept of "water" and the meaning of the word "water"? No. And here's why. The difference is not that one substance (e.g. the stuff coming out of your eyes) is not water and the other stuff is water. They are both water.2 The difference is only that one is called "water" and the other is not. And that only goes to show that we need to distinguish what something is from what it may be called on particular occasions. But it is clear that we needed to do this anyway. After all, all and only physical objects are physical objects, but rarely do we actually call any of them that. Some of these physical objects are animals, and some of those animals are people, and we often call them "people". Some of these people may on occasion appear to hotly deny that they are animals or physical objects, but of course they would be meaning that they are not merely that. Some of these people are students, and some of those students are vegetarians, and so forth. What you call something depends not only on what it is, what categories it falls into, but also on why you want to refer to it, what your purposes of the moment are. So we usually ask for a cup of tea rather than asking for a cup of hot water containing an infusion of tea or other herbal matter, because the former turn of phrase is easier for most Chomsky remarks, concerning a cup of tea, that "we do not have water at all (except in the sense that milk is mostly water, or a person for that matter)" (1995, p. 22). Apparently he is invoking an ambiguity in the word "water", and my reply to that is in $4 below.

316 Barbara Abbott

contexts and purposes. (For a number of examples of unusual language use of the latter type, see the movie The Coneheads.) Let me summarize our conclusions so far. First, it is clear from water's variety of sources, locations, and functions in our lives that features concerning these aspects do not form part of the meaning of the word "water", nor do they play a role in determining its denotation. Second, substances such as celery juice, Sprite, and beer are partly or mostly water in virtue of being partly or mostly HzO. Incidentally, it must be stressed that to deny that water is (coextensive with) H20 would be to deny this basic fact, a punishment which should be severe enough to deter anybody. Third, some other substances such as tears or holy water may consist virtually completely of water (give or take impuritiessee $4 below), and tea and coffee may be on the edges of this category. Hence we must attend to the difference between what something is and what it is called. Fourth, substances which are not called "water" are called by other names because they belong to particular categories in virtue of other properties they may have. These other categories are not natural kinds (in the relevant philosophical sense), and the names for them are not natural kind terms. Hence in principle features of any type may form a part of their meaning. We might expect features related to human interests and concerns to be crucial, so it is interesting that in the majority of these cases internal structure properties appear to play a distinguishing role. However in the case of substances such as tears and holy water, which have or are believed to have the same internal structure as ordinary water, other features are crucial, including possibly features of source or function. Gricean considerations can readily explain why in typical situations "holy water" or "tears" are better expressions to use than "water".

4. Vagueness: why water can contain impurities If water really is H20, why do we not reserve use of the term "water" for only those quantities which contain H20 molecules and nothing else? One possible conclusion is that "water" is ambiguous, with one sense meaning pure water and the other meaning impure water. Chomsky (1995, p. 22) speaks in terms of different senses of "water" (as did Putnarn (1975, pp. 239m). But this is unsatisfactory. Methodologically speaking, postulation of ambiguity is weak. Furthermore our intuitions say "water" is not ambiguous in this way, and grammatical evidence agrees. For one thing, there is apparently no language which regularly distinguishes pure from impure water with two different words. Secondly, VP proforms do not dis-

A Note on the Nature of "Water" 31 7

tinguish two senses for "water", as opposed to the clearly ambiguous "bat". Compare (1) and (2). (1) Mary drank some water (it was pure) and so did Sue (yucky polluted stuff?). (2) ? Mary bought a bat (to practice line drives) and so did Sue (to eat the mosquitoes in her garden). Similarly (3) can easily apply to a situation where what the chemist discarded was pure water, but what was found in the puddle was polluted.

(3) The water that the chemist discarded was later found in a puddle in the street. The burden of proof for those who would still want to claim ambiguity is probably too heavy for anyone to bear. The fact that we apply "water" to substances containing other things is part of a very general and natural type of vagueness in our use of linguistic expressions, and one which therefore need not and should not be incorporated separately into our account of the meaning of each word and phrase. For example, when someone asks to borrow our car we are not required to remove from the glove compartment the maps, box of Kleenex, registration, etc.-i.e. everything that is not part of the car strictly speaking. Nor when we ask for a plate do we insist that the plate be sterilized, or when we ask to see someone do we assume that they will take their clothes off. It is true that Portia was able to get Shylock on this type of technicality, but had the legal advisor been anyone else, Antonio probably would have been a goner. Unger (1975) has argued that nothing is really flat, because everything is less flat than some other real or imaginable thing. Lewis (1979) replies that part of any context is an implicit standard for precision in the use of terms, and for a particular claim to be true in a given situation it is only required that the local standard of precision be met. When we imagine something flatter than the pavement or the desk which we have previously judged to be flat, we are implicitly shifting standards. Similarly I am claiming that the use of the term "water" for substances containing impurities is a case of ordinary vagueness, and that the appearance of ambiguity or different uses of the term "water" depending on the amount of impurities in question is actually a reflection of different implicit standards of precision. By the same token "water" may not be felt to be entirely synonymous with "HZ07'because of the higher standard of precision regularly invoked by the latter.

318 Barbara Abbott

5. Conclusion I have reasserted the not-so-bold claim that water is in fact H20, despite the fact that the word "water"'may be applied to substances containing other than H 2 0 molecules. I have argued that features of source, location, and function do not form a part of the meaning of the word "water". Such features may form a part of the meanings of terms that refer to substances which consist either partly or wholly of water, but since these are not natural kinds that should not be surprising. Thus the conclusion that the meanings of natural kind terms are similar to those of other words is not supported, and neither is the strict internalist view of Chomsky (1995).3

Department of Linguistics and Germanic, Slavic, Asian and African Languages Department of Philosophy Michigan State University A-614 Wells Hall East Lansing, MI 48824-1027 USA [email protected]

BARBARA ABBOTT

REFERENCES

Abbott, Barbara 1989: "Nondescriptionality and Natural Kind Terms". Linguistics and Philosophy, 12, pp. 269-92. Chomsky, Noam 1995: "Language and Nature". Mind, 104, pp. 1-6 1. Goosens, William K. 1977: "Underlying trait terms", in Stephen Schwartz, ed., Naming, Necessity, and Natural Kinds. Ithaca: Cornell University Press, pp. 133-54. Grice, H. Paul 1975: "Logic and Conversation", in Peter Cole and Jerry L. Morgan, eds., Syntax and Semantics, Vol. 3: Speech Acts. New York: Academic Press, pp. 41-58. Keil, Frank C. 1989: Concepts, Kinds, and Cognitive Development. Cambridge, MA: MIT Press. Kripke, Saul A. 1972: "Naming and Necessity", in Donald Davidson and Gilbert Harman, eds., Semantics of Natural Language. Dordrecht: D. Reidel, pp. 253-355. An earlier version of this paper was read at the annual meeting of the Linguistic Society of America, January 1996, in San Diego, California and I would like to thank that audience for their comments. I would also like to express my gratitude to Paul Bloom, Noam Chomsky, Gene Cline, Rich Hall, Larry Hauser, Larry Horn, Barbara Malt, Stan Mortel, Carol Slater, and an anonymous reader for their comments on earlier drafts of this paper, but without necessarily implying their agreement with the finished product.

A Note on the Nature of "Water" 319

Lewis, David 1979: "Scorekeeping in a Language Game". Journal of Philosophical Logic, 8, pp. 339-59. Malt, Barbara C. 1994: "Water is not H2O9'. Cognitive Psychology, 27, pp. 41-70. Putnam, Hilary 1975: "The Meaning of 'Meaning"', in his Mind, Language, and Reality: Philosophical papers, Vol. 2. Cambridge: Cambridge University Press, 1975, pp. 2 15-7 1. Schwartz, Stephen P. 1978: "Putnam on Artifacts". The Philosophical Review, 87, pp. 56tG74. Stich, Stephen P. 1983: From Folk Psychology to Cognitive Science. Cambridge, MA: MIT Press. Unger, Peter 1975: Ignorance: A Case for Scepticism. Oxford: Oxford University Press. Wilson, Mark 1982: "Predicate Meets Property". The Philosophical Review, 91, pp. 549-89.

http://www.jstor.org

LINKED CITATIONS - Page 1 of 2 -

You have printed the following article: A Note on the Nature of "Water" Barbara Abbott Mind, New Series, Vol. 106, No. 422. (Apr., 1997), pp. 311-319. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28199704%292%3A106%3A422%3C311%3AANOTNO%3E2.0.CO%3B2-2

This article references the following linked citations. If you are trying to access articles from an off-campus location, you may be required to first logon via your library web site to access JSTOR. Please visit your library's website or contact a librarian to learn about options for remote access to JSTOR.

[Footnotes] 1

Putnam on Artifacts Stephen P. Schwartz The Philosophical Review, Vol. 87, No. 4. (Oct., 1978), pp. 566-574. Stable URL: http://links.jstor.org/sici?sici=0031-8108%28197810%2987%3A4%3C566%3APOA%3E2.0.CO%3B2-N 1

Predicate Meets Property Mark Wilson The Philosophical Review, Vol. 91, No. 4. (Oct., 1982), pp. 549-589. Stable URL: http://links.jstor.org/sici?sici=0031-8108%28198210%2991%3A4%3C549%3APMP%3E2.0.CO%3B2-U 2

Language and Nature Noam Chomsky Mind, New Series, Vol. 104, No. 413. (Jan., 1995), pp. 1-61. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28199501%292%3A104%3A413%3C1%3ALAN%3E2.0.CO%3B2-W

References

NOTE: The reference numbering from the original has been maintained in this citation list.

http://www.jstor.org

LINKED CITATIONS - Page 2 of 2 -

Language and Nature Noam Chomsky Mind, New Series, Vol. 104, No. 413. (Jan., 1995), pp. 1-61. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28199501%292%3A104%3A413%3C1%3ALAN%3E2.0.CO%3B2-W

Putnam on Artifacts Stephen P. Schwartz The Philosophical Review, Vol. 87, No. 4. (Oct., 1978), pp. 566-574. Stable URL: http://links.jstor.org/sici?sici=0031-8108%28197810%2987%3A4%3C566%3APOA%3E2.0.CO%3B2-N

Predicate Meets Property Mark Wilson The Philosophical Review, Vol. 91, No. 4. (Oct., 1982), pp. 549-589. Stable URL: http://links.jstor.org/sici?sici=0031-8108%28198210%2991%3A4%3C549%3APMP%3E2.0.CO%3B2-U

NOTE: The reference numbering from the original has been maintained in this citation list.