(notes omitted)
From A Pessin and S. Goldberg, eds., The Twin Earth Chronicles: Twenty Years of Reflection on Hilary Putnam's "The Meaning of 'Meaning'". (Armonk, New York: M.E. Sharpe, 1996)
The papers in this collection represent a range of responses to my essay "The Meaning of 'Meaning.'" The editors have made an excellent selection, and a reader who works through this collection of essays will certainly be abreast of the debate about my "semantic externalism." To comment on each of the individual papers would require a whole series of essays rather than an "introduction," and I shall not attempt it. Instead, I shall offer a few general remarks, and comment on just two of the reactions to my "Twin Earth" thought experiment and to semantic externalism in general.
"The Meaning of 'Meaning'" was itself a reaction to an assumption about concepts that is literally millennia-old, and I have never ceased to be surprised and gratified by the speed with which and the extent to which the view I proposed in that essay became widespread. When I speak of a millennia-old assumption about concepts, I am not overlooking the fact that there has been disagreement about concepts and about conceptual thought ever since Aristotle disagreed with Plato about the nature of "ideas" in the Greek sense of that word (often translated as "forms"). For example, according to Plato, ideas are extra-mental entities, but knowledge of them is supposed to be innate in the mind, and capable of being brought to consciousness by an act of "recollection." For Aristotle, they are both mental and extra-mental. The very same "idea" or "form" is supposed to be capable of existing in things, but also, minus its matter, in the mind. In the middle ages, Conceptualist and Nominalist views were added to the Platonic and Aristotelian alternatives. For the Conceptualists ideas are explicitly mental entities; for the Nominalists they are, of course, "names," but the understanding of "names" is supposed to lie in each individual mind. In the modem period, for an empiricist like Hume, ideas are hardly distinguished from mental images, and are certainly in the mind. Thus, in spite of the variety of metaphysical theories about the nature of concepts, this much was not doubted: concepts were uniformly thought of as capable of being completely contained in or recollected by "the mind" (which was itself conceived of as a private theater, isolated from other individuals and from the "external world").
It was also taken for granted by almost all the philosophers in the tradition that the idea in the mind, or the possession or recollection of the idea by the mind, determines the extension of the "name" associated with the idea or concept: a name, say, "dog," is true of a particular thing inasmuch as that particular thing falls under the concept in the mind, or the concept recollected by the mind. In short, it is a feature of all these views that one individual in isolation can, in principle, grasp any concept whatsoever, and that the individual's grasp of his or her concepts totally determines) the extension of all the individual's terms. Knowledge of meanings is private mental property.
It was against this conception of meanings and of knowledge of meanings as private mental property that I wrote "The Meaning of 'Meaning,'" and the Twin Earth thought experiment was only one of a number of arguments that I used to argue that, contrary to these traditional views, knowledge of meanings is not something that is possible for a thinker in isolation, and that it presupposes both interactions with the world and interactions with other language users. The interaction with other language users is something that I illustrated with the aid of words which we are all competent to use, but the members of whose extension most of us are not able to recognize. For me at that time the word "elm" was such a word. I possessed what normally counts as "knowledge of the meaning" of the word "elm." For example, I knew that elms are deciduous trees, that they are common in North America and in Europe, I knew the approximate height of a typical elm (I knew that it was not a miniature tree or a bush). True, I was quite unable to distinguish an elm from a beech; yet, to say that I did not know the meaning of the word "elm" could, pace John Searle, have been to confuse lack of botanical knowledge with lack of linguistic competence. Or (another example I used), take the word "gold." I suppose that no one who is not a chemist or a metallurgist or a miner or a jeweller can distinguish true gold from fake gold with genuine reliability; the rest of us have to rely on these experts when we purchase gold jewelry in order to avoid being fooled. Yet this does not mean that we lack knowledge of the meaning of the word "gold." My suggestion was that knowing the meaning of the word "gold" or of the word "elm" is not a matter of knowing that at all, but a matter of knowing how; and what you have to know how is to play your part in an intricate system of social cooperation. Language, I said, is not a tool like a hammer, that anyone can use by him- or herself. It is a tool like a large ship, which it takes many people working together to operate. I can refer to gold, talk about gold, purchase gold, etc., perfectly well without being able reliably to distinguish gold from non-gold because there are others in the communityexperts-upon whom I can rely. In short, there is a linguistic division of labor.
There are, however, words whose extension we are all competent at determining (at least in normal circumstances). All of us, for example, are extremely reliable at identifying water, at least under normal circumstances (although under doubtful circumstances even here we might have to consult an expert). But even the understanding of a word like "water" is not, I argued, simply a matter of having an item tucked away in a private mental theater.
In arguing that meaning depends not only on other speakers, but also on the environment, I used several examples (in "The Meaning of Meaning'" and its predecessor, "Is Semantics Possible?").2 For instance, I imagined at one point that the words "aluminum" and "molybdenum" had their meanings unintentionally switched by English-speaking colonists on another planet. (Something like this happened when English colonists came to the United States and ignorantly applied words like "sparrow" to species of birds which are not called "sparrows" in England.) What I associate with the word "aluminum" is not very different from what I associate with the word "molybdenum" (if we prescind from the knowledge that pots and pans are often made of aluminum and not of molybdenum-but we can also imagine that on the colony, pots and pans are normally made of molybdenum). Thus, we can imagine that an English-speaker on Terra and an English-speaker on one of Terra's colonies might be in the same brain state, at least in all relevant respects, on a particular occasion of the use of the sentence, "This pot is made of aluminum," and yet one of them might be saying that the pot is made of aluminum and the other might be saying that it is made of molybdenum. To mean aluminum when one uses the word "aluminum," I argued, it is not enough to have certain associations, certain mental images, etc.; it is necessary that the metal referred to in the linguistic community by that name actually be aluminum, and whether that is the case depends on whether one has certain direct or indirect interactions with aluminum itself, including interactions that go through experts in one's community.
The example that became the most famous, however, and the one that gives this anthology its title, was the one involving Twin Earth. I wrote,
One of the peculiarities of Twin Earth is that the liquid called "water" is not H2O but a different liquid whose chemical formula is very long and complicated. I shall abbreviate this chemical formula simply as XYZ. I shall suppose that XYZ is indistinguishable from water at normal temperatures and pressures. In particular, it tastes like water and it quenches thirst like water. Also, I shall suppose that the oceans and lakes and seas of Twin Earth contain XYZ and not water, that it rains XYZ on Twin Earth and not water, etc.
If a spaceship from Earth ever visits Twin Earth, then the supposition at first will be that "water" has the same meaning on Earth and on Twin Earth. This supposition will be corrected when it is discovered that "water" on Twin Earth is XYZ, and the Earthian spaceship will report somewhat as follows:
"On Twin Earth, the word 'water' means XYZ."
I went on to claim that the difference in the meaning of the word "water" existed even before chemistry was developed on either Earth or on Twin Earth (say, in 1750); it is just that in 1750 neither community knew the chemical nature of the substance each called "water." The meaning was different because the stuff was different. And I pointed out that an Earth speaker and his Twin Earth Doppelgänger might be in the same brain state neuron for neuron, and it would still be the case that what the Earthian meant by his word "water" was not what his Twin Earthian Doppelgänger meant by the word. It was at this point that I wrote, "Cut the pie any way you like, 'meanings' just ain't in the head."
Of course, denying that meanings are in the head must have consequences for the philosophy of mind, but at the time I wrote those words I was unsure as to just what those consequences were. After all, such accomplishments as knowing the meaning of words and using words meaningfully are paradigmatic "mental abilities"; yet, I was not sure, when I wrote "The Meaning of 'Meaning,'" whether the moral of that essay should be that we shouldn't think of the meanings of words as lying in the mind at all, or whether (like John Dewey and William James) we should stop thinking of the mind as something "in the head" and think of it rather as a system of environment-involving capacities and interactions. In the end, I equivocated between these views. I said, on the one hand, that "meanings just ain't in the head," and, on the other hand, that the notion of the mind is ambiguous, and that, in one sense of "mental state" (I called mental states, in this supposed sense, `narrow mental states'), our mental states are entirely in our heads, and in another sense (I called mental states in this supposed second sense "broad mental states"), a sense which includes such states as knowing the meaning of a word, our mental states are individuated by our relations to our environment and to other speakers and not simply by what goes on in our brains. Subsequently, under the influence of Tyler Burge and more recently of John McDowell as well, I have come to think that this conceded too much to the idea that the mind can be thought of as a private theater (situated inside the head).
Searle's Reaction
John Searle's view is that meanings are literally in the head. He replies to the arguments in "The Meaning of 'Meaning"' by claiming that English speakers who are not able to tell elms from beeches "don't know the meaning or know it imperfectly." At the same time, he holds that even such speakers have a different concept of an elm and of a beech; for, he argues, they have "enough conceptual knowledge to know the two are distinct species." Here, I can reply that if, indeed, they "don't know the meaning" at all, then it follows at once that this "conceptual knowledge," whatever it is, is not meaning. So let us assume that Searle's assertion that they "know it imperfectly" or "don't know it at all" was a slip.
The only "conceptual knowledge" about elms and not about beeches that such a speaker typically possesses is that elms are trees that experts upon whom he or she relies refer to by the name "elm." Does the existence of such conceptual knowledge show that meanings are in the head as Searle believes? (Searle regards all mental states as "features of the brain.")
Well, if it does, then "the tree that experts upon whom I rely call an 'elm'" must be part of what such a speaker means by "elm," for otherwise nothing in the the part of the meaning known to the speaker (in the "Intentional content" of "elm," to employ Searle's notion) distinguishes elms from beeches. (I mean, of course, that this is a consequence of Searle's view; in my view, knowing the meaning of "elm" isn't, for the most part, a matter of knowing that at all, but of participating successfully in the linguistic division of labor. And, in my view, the sound and spelling of the word, although of course known to speakers, are not part of its meaning.)
The whole notion of "Intentional content" on which Searle relies seems to me intensely problematic for reasons which I shall explain shortly. But, in any case, even if there were Searlian "Intentional contents," they could not be identical with the meanings of words in a public language. I don't mean to say that some other objects are the meanings of words in a public language; in my view, meanings aren't objects at all. I agree with Wittgenstein and with Quine that the idea of its meaning as some kind of "entity" with which a word is "associated" and which determines how the word is to be used in every context, is as untenable today as belief in Homeric gods. The real questions are (as many philosophers have pointed out) not what meanings are, but (1) what it is to know the meaning of a word; and (2) when should we say that two words have the same meaning, i.e., how to understand the relation of synonymy. When I say that "Intentional contents" in Searle's sense (if there are such things) are not the same as meanings in a public language at all, I mean that sameness of "Intentional contents" in Searle's sense does not have the properties that synonymy has, or even that similarity of meaning has, in a public language. Wittgenstein famously remarked that for many purposes (though not for all) we may say that the meaning of a word is its use in the language. Thus, to say of words in a public language, or in two different public languages, that they have "the same meaning" is to say that in certain respects they have the same use. But to know what that comes to (and it comes to different things in different cases), we have to know in each case which features to discount and which features are relevant. For example, when we say that in English the word "elm" has the same meaning as the word "Ulme" in German, we discount the differences in the sound and spelling of the words themselves. But on Searle's theory (as I just argued) what an English speaker means by the word "elm" would be the kind of tree that experts upon whom I rely refer to by the name "elm" and what the German speaker means by the word "Ulme" is the kind of tree that experts upon whom I rely refer to by the name "Ulme." These are two entirely different "Intentional contents." Thus, if we suppose that Searle's theory of "Intentional contents" is a theory of meaning in a public language, we would have to say that a word and its translation do not have the same meaning at all. Indeed, they don't even have similar meanings-for let us suppose that, by some accident, some group of English speakers have switched the words "elm" and "beech." Let us suppose that in Nova Scotia, as it might be, beeches are called "elms" and elms are called "beeches." On Searle's theory, the word "elm" would, in this case, have the same meaning in Massachusetts and in Nova Scotia, notwithstanding the fact that (speaking as we ordinarily speak) in Nova Scotia "elm" means "beech," while "elm" in English and "Ulme" in German would have totally different meanings. Our actual practices of individuating the meanings of words would, in fact, be totally ignored by such a theory.
More importantly (since this is an anthology of papers on Twin Earth), Searle denies that the Twin Earth thought experiment shows that meanings aren't just features of the brain. According to Searle, "most people do not go around baptizing natural kinds; they just intend to use words to mean and refer to whatever the community at large, including the experts, use the words to mean and refer to." What this seems to mean is that, for most people, "water" means "stuff that the community at large, including the experts, call `water.' " This account has the same defects as the account of the meaning of "elm" I just considered.
Searle also argues that even if the meaning of water did have an indexical (ostensive) element of the kind I suggested in "The Meaning of 'Meaning,'" that would not mean that it was not a private mental content. What a typical Earthian English speaker might mean by the word "water" is, he suggests, "whatever is identical in structure to this stuff," and a typical Twin Earth English speaker might associate the very same "Intentional content" with the word. Thus, on this alternative proposal, the word would have exactly the same meaning on Earth and Twin Earth; however, it would have a different extension because the Earth speakers are referring to what comes out of Earth faucets and the Twin Earth speakers are referring to what comes out of Twin Earth faucets. Such a view, it seems to me, fails to conform to our actual practice of deciding on sameness/similarity of meaning in a public language by denying that what we are in fact referring to plays any role at all.
In addition, it is uncertain whether Searlian "Intentional contents" really exist at all. How is the Searlian speaker in isolation* supposed to have the notion of an "expert"? What is the "Intentional content" of "expert"? Perhaps Searle would say that the "Intentional content" of "expert" is: speaker whom other speakers call an "expert." But then how is the "intension" of "speaker" supposed to be fixed? Perhaps Searle would say that the "Intentional content" of "speaker" is: person with whom I engage in a certain kind of communication. But then how is the "Intentional content" of "communication" fixed?
It is clear that Searle is making the assumption that at least some very significant concepts (concepts like "speaking," "communicating," and "calling something a so-and-so") are concepts that a speaker could master and possess without relying on the actual nature of any of the objects with which he or she interacts, and without the benefit of a linguistic division of labor. This is a sort of "private language" assumption. I myself do not believe that mysterious "Intentional contents," totally inside the head but with the power of fixing reference to external realities all by themselves, exist at all. But I do not need to argue against Searle's metaphysics here-for, in any case, as I have just shown, even if there were Searlian "Intentional contents," sameness and difference of Searlian "Intentional contents" could not be the same thing as sameness and difference of meaning or concept as meanings and concepts are actually individuated.
Remember, "Intentional contents" are supposed to be features of the brain, and thus presumably do not depend on relations to the environment or to other speakers at all!
Tyler Burge's Criticism and a Concluding Remark
Tyler Burge's criticism comes, so to speak, from the opposite direction. Burge agrees with my semantic externalism, but argues that I did not carry it far enough. In Burge's view, my attempt in "The Meaning of `Meaning' " to hold a place open for a notion of "narrow content" and for "narrow mental states" represented a confusion on my part, and I have come to believe that he right. But, since I agree with his paper, I will only refer the reader to it rather than attempt to summarize his arguments.
In this Introduction, I have focused on the two doctrines in "The Meaning of 'Meaning'" that have aroused the most controversy: semantic externalism and the division of linguistic labor. Unfortunately, more than one author has misread the essay by overlooking the fact that it asserted that there are additional factors involved in meaning. "The Meaning of 'Meaning'" also contains the beginning of a theory of the semantics of natural-kind words, and that theory does not claim that their meaning is entirely determined by the nature of the objects in their extension (whatever that nature is discovered to be by scientific investigation), although I did and do still claim that it is partly so determined. Another factor that partly determines the meaning of a natural-kind term is the stereotype that speakers have of a typical member of the kind. If there is an area in which there is further work to be done (linguistically and historically as well as philosophically), it is the study of the ways in which these two components-the stereotype and the objective nature of the objects in question-interact with one another. I assume that the reader of this anthology will read (or re-read) "The Meaning of `Meaning' " itself. I urge him or her not to neglect the less thematized but nevertheless (I believe) important discussions sprinkled through it of such matters as stereotypes, the various senses that natural-kind words can have, the problems with what I called "California semantics," etc.
Hilary Putnam Cambridge, Massachusetts Spring 1995