Hilary Putnam on Meaning:
An important question for the theory of meaning is whether meaning is a private mental entity, or an abstract public entity. For example, when we utter the word, ‘horse’ is the meaning of this word a mental concept? That is, do we understand the meaning of the term ‘horse’ simply by that picture we get in our head when we hear the word horse? Or is meaning some sort of abstract public entity where a term like ‘horse’ is simply understood as that thing to which everyone in a linguistic community refers to when they say the word ‘horse’? In “Meaning and Reference” Putnam argues that meaning is not a mental entity. Or as he says, ““meanings” just ain’t in the head!” (Putnam, 2000). Putnam shows this by constructing a Twin Earth thought experiment where the term ‘water’ on Twin Earth refers to a liquid with a chemical structure XYZ. Putnam argues that the term ‘water’ has the same intension on Earth and Twin Earth, yet has two different extensions. This is meant to show that meaning is not “in the head.” Putnam goes on to argue that the term ‘water’ is a rigid designator, and that what they call ‘water’ on Twin Earth is not really water, because it differs in chemical structure from H20. In this essay, I will argue that if we are consistent in our Twin Earth thought experiment, we will have reason to doubt that the term ‘water’ is a rigid designator, and that it might not only refer to H2O, but could also refer to XYZ. I will begin by defining several key terms that are central to Putnam’s argument. Second, I will outline Putnam’s argument that two terms can have the same intension, yet have different extensions. Finally, I will argue that there is reason to doubt that the term ‘water’ is a rigid designator, and that it might not only refer to H2O, but could also refer to XYZ.
Before I give an account of Putnam’s argument, it is first necessary to define a few important terms, mainly, ‘extension’and‘intension’. The ‘extension’ of a term is the complete set, or class of things to which a term refers. For example, the term ‘creature with a heart’ refers to ‘humans’, ‘dogs’, ‘rabbits’, and all other individuals in the class of creatures that have hearts. The ‘intension’ of a term refers to the mental associations one has with a certain term, or another way to say this: the concept or idea one gets in one’s head when one hears a certain term. So, for example, when I hear the word ‘dog’ I get a picture in my head of a creature with certain defining characteristics, and I can distinguish this mental image of ‘dog’ from other creatures, say, a horse, a cat, and so on. Now the difficulty with the terms ‘intension’ and ‘extension’ is that each is meant to convey meaning, yet neither term fully captures meaning. So, for example, we can say that the terms ‘creature with a heart’ and ‘creature with a kidney’ have the same extension (i.e., apply to the same class of things), and so, are equivalent (i.e., the same as one another) in this respect, however their intensions differ. That is, the term ‘creature with a heart’ means something different than ‘creature with a kidney’. So even though each of the terms are equivalent in that they refer to all of the same types of things (i.e., because all creatures that have hearts also have kidneys), we still think that the two terms mean something different than one another, that they are not entirely equivalent. This ambiguity in the traditional theory of meaning is, in part, what Putnam is responding to in his essay. Now that we have an understanding of these terms, we can move on to see how they factor into Putnam’s argument.
Putnam begins by noting two false assumptions of the traditional theory of meaning. The first is that “meaning” was thought to be a mental entity. That is, the meaning of a term was thought to be equivalent to the concept, or idea, one gets “in the head” when one hears or thinks of a certain term (this is how I defined the term ‘intension’ above). Hence, on the traditional view, it was said that one was in a certain psychological state when one understood the meaning of a term. The second false assumption of the traditional view was that while terms could have the same extension and yet differ in intension, the reverse of this was impossible. So as we saw above, two terms having the same extension (e.g., ‘creature with a heart’ and ‘creature with a kidney’) can differ in intension (i.e., mean different things) – we have no difficulty in understanding this. However, as Putnam notes, the reverse of this was assumed to be impossible: “two terms cannot differ in extension and have the same intension” (Putnam, p. 418). So, put more succinctly, the two false assumptions of the traditional theory of meaning are as follows: “(1) that knowing the meaning of a term is just a matter of being in a certain psychological state…” and “(2) That the meaning of a term determines its extension (in the sense that sameness of intension entails sameness of extension)” (Putnam, p. 419). In his essay Putnam rejects these two false assumptions of the traditional theory of meaning. Putnam makes his argument by use of the following example.
Suppose that there is a planet nearly identical to Earth, called Twin Earth. There are even people on Twin Earth, some of whom speak English. In fact, these two planets are so similar to one another that each person on Earth has an identical copy on Twin Earth. But suppose there is one difference between the two planets: on Twin Earth the substance ‘water’ is not H2O, but consists of some other chemical composition, abbreviated XYZ. Now, H2O looks, smells, and tastes exactly like XYZ, and the lakes, rivers, and oceans on Twin Earth all appear exactly the same as those on Earth, however, on Twin Earth ‘water’ is XYZ and not H2O.
Next, Putnam supposes that were a spaceship from Earth to visit Twin Earth, we would initially think that ‘water’ has the same meaning on Twin Earth, as it does on Earth. That is, we would think that Earthians and Twin Earthians were referring to the exact same thing when each referred to water. We would continue to think this until we actually tested the chemical composition of water on Twin Earth and found that it was not in fact H2O, but XYZ. Once the chemical composition of water was known, the term ‘water’ would have two different meanings: water on Twin Earth would mean XYZ, and water on Earth would mean H2O. Another way to put this is the term ‘water’ would have two different extensions. So, as Putnam says, on Earth, the extension of the term water would refer to “…the set of all wholes consisting of H2O molecules…” but on Twin Earth, the extension of the term water would refer to “…the set of all wholes consisting of XYZ molecules” (Putnam, p.420). We will see the significance of this below, if it is not already apparent.
Putnam continues with his example by supposing that we are back in the year 1750, approximately 50 years before it was discovered, on Earth, that water was composed of H2O, and similarly, 50 years before it was discovered, on Twin Earth, that ‘water’ was XYZ. Now suppose two people, Oscar1, on Earth, and Oscar2, on Twin Earth, both have beliefs about water. For example, both Oscar’s believe that water is that substance with which lakes, rivers, and oceans are composed of, and water is that which we drink, bathe in, and so on. In fact, we might say that Oscar1 and Oscar2 are physiologically and psychologically identical such that each has all the same beliefs about water that the other has. However, in 1750, neither of the Oscar’s knew the chemical composition of water on their respective planets. What does Putnam conclude from this? Putnam concludes that “Oscar1 and Oscar2 understood the term ‘water’ differently in 1750 although they were in the same psychological state…Thus the extension of the term ‘water’ (and, in fact, its “meaning” in the intuitive pre-analytical usage of that term) is not a function of the psychological state of the speaker by itself” (Putnam, p. 420). In other words, Putnam believes he has shown to be false the two above-noted assumptions of the theory of meaning, mainly “that knowing the meaning of a term is just a matter of being in a certain psychological state;” and “that sameness in intension determines sameness of extension.” So is this true? Has Putnam shown both assumptions to be false?
Well, it certainly seems clear that Oscar1 and Oscar2 are in the same psychological state as one another when they think of ‘water’. After all, Oscar1 and Oscar2 share all the same beliefs about water. Each has the same mental picture or concept in his mind when he thinks of water. It also seems clear that each is thinking about two different substances when thinking about water: one is thinking about H20, the other is thinking about XYZ. So, given that they are in the same psychological state as one another while thinking about water, and given that they are really thinking about two different substances, then it mustn’t be true that knowing the meaning of a term is simply a matter of being in a psychological state.
As for assumption (2) it seems clear that we have sameness in intension: the term ‘water’ on Earth has all the same mental associations as the term ‘water’ on Twin Earth. When the Earthian forms a mental concept of water in her head, she thinks about ‘that which, lakes, rivers, and oceans are composed of, etc’, just as the Twin Earthian does. And it also seems clear that the term ‘water’ has two different extensions: water on Earth is H20 and water on Twin Earth is XYZ. Hence, sameness in intension does not entail sameness in extension.
There is, however, one objection to Putnam’s argument that we might like to raise at this point. We might argue that due to the fact that Twin Earthians intend the same thing as we do when they say the word ‘water’ (and what I mean by this is that when they say ‘water’ they mean ‘that which lakes, rivers, and oceans are composed of’, and when we say ‘water’, we mean ‘that which lakes, rivers, and oceans are composed of’) that this shows the term ‘water’ on Twin Earth really does mean the same thing as on Earth, the only difference is the chemical composition (and thus, the extension) of ‘water’ on the two planets (Putnam, p. 424). So our thesis might be that it is a linguistic community that defines the meaning of a term, regardless of how that same term might be used outside of that community, and regardless of any difference in extension. In other words, water can differ in chemical composition from one linguistic community to another (and thus, the extension of the term differs), yet the term still retains the same meaning, due to the fact that people on Earth and Twin Earth, are all referring to all the same types of things (e.g., lakes, rivers, rain, liquid in a glass, etc) when they use the word ‘water’. However, as Putnam points out, it is the chemical composition of water that makes water, water. Water is H2O and nothing else. And so, any possible world where the word ‘water’ refers to something other than H2O will not be water at all. As Putnam says,
You will not have described a possible world in which “water is XYZ,” but merely a possible world in which there are lakes of XYZ, people drink XYZ (and not water), or whatever. In fact once we have discovered the nature of water, nothing counts as a possible world in which water doesn’t have that nature. Once we have discovered that water (in the actual world) is H2O, nothing counts as a possible world in which water isn’t H2O (Putnam, p. 426).
So Putnam’s response would seem to indicate that the meaning of water is determined by the actual world, and given that in the actual world, water is H2O, any possible world where some liquid referred to as water is not H2O, is not really referring to water at all.
But Putnam’s response raises a second objection. While it may be true that ‘water’ on Earth and Twin Earth do not share the same chemical composition, why think ‘water’ on Earth is the “real” water, and that ‘water’ on Twin Earth ought not to count as water? In other words, why is water H20 and not XYZ? One obvious reply is to say that our world actually exists, whereas Twin Earth, as far as we know, does not exist. And so, meanings defined in the real world are real meanings (i.e., the ones we ought to accept), and any meanings arising from some “possible world” thought experiment, are just “made-up” meanings – meanings we construct in order to make a philosophical point –and so these meanings do not count as “real” meanings. But doesn’t this reply dismiss our above question a bit too quickly?
It seems unfair to dismiss the existence of a possible world (and thereby dismiss the meanings of words in those possible worlds), for the only reason the Twin Earth example was able to be effective for Putnam was is if we imagined that a Twin Earth actually did exist. And so if we want to suppose the existence of a possible world in order to make an argument (as Putnam did), we must continue to suppose the existence of this world in order to entertain objections to that argument. To do so is to be consistent. In other words, we cannot use the example of a possible world only when it suits our purpose, and then deny the existence of that same world if it looks as though such a though experiment might pose potential trouble for our argument. So I think we will need to continue supposing that Twin Earth does exist, if we are to have a fair chance at criticizing Putnam’s view. So, key to our criticism here, is the supposition that Twin Earth does exist.
So if we are supposing that Twin Earth exists, and that on Twin Earth water is XYZ, then don’t Twin Earthians have just as much reason to think that water is XYZ as we, on Earth, have to believe that water is H2O? Putnam wants to say that the term ‘water’ is a natural-kind term that is indexical. What this means, Putnam says, is
[w]hen I say “this (liquid) is water,” the “this” is, so to speak, a de re “this”–i.e., the force of my explanation is that “water” is whatever bears a certain equivalence relation [the relation “same liquid as”] to the piece of liquid referred to as “this” in the actual world (Putnam, p. 424).
In other words, Putnam claims that natural-kind terms like ‘water’ designate rigidly, which is to say that that “water = H2O” holds true for all possible worlds. But couldn’t we make the same argument from Twin Earth and wouldn’t it carry just as much force as our argument on Earth? So, for example, if Oscar2, on Twin Earth, points to a glass of XYZ and says “this is water” wouldn’t he also claim that the term ‘water’ (meaning XYZ) was indexical, and so the claim “water=XYZ” would hold true for all possible worlds? If we continue to suppose that Twin Earth does exist, which it seems we must if we want to remain consistent in our thought experiment, then Oscar2 has just as much reason to believe water is XYZ as Oscar1 has to believe that water is H2O.
Another way to make this point is to suppose, as part of our Twin Earth thought experiment, that if there is a Putnam2 on Twin Earth writing the exact same essay on meaning and reference, won’t Putnam2 make the same argument as Putnam1, and won’t his claim that the term ‘water’ (XYZ) designates rigidly have to be taken just as seriously as Putnam1’s argument that water (H2O) designates rigidly?
The point of this criticism is not to show that a Twin Earth does exist, or even that it could exist, but rather, that we have to be consistent in the use of the Twin Earth example. We cannot suppose that a Twin Earth exists in order to make a philosophical point, and then reject the existence of this world if it turns out that its existence poses problems for our argument. When we are consistent in the supposition of the existence of a possible world, we see that we have no more reason to think that water is H2O in all possible worlds than Twin Earthians would have to belief that water is XYZ in all possible worlds. Hence, Putnam has not shown that the term ‘water’ designate rigidly.
Hilary Putnam, “Meaning and Reference” in E.D. Klemke and Heimir Geirsson, eds., Contemporary Analytic and Linguistic Philosophies, 2nd edition (Amherst, New York: Prometheus Books, 2000), p 422.