As we said earlier in the class, some of our mental states have representational content.
The big ambitious question about these is:
Some smaller questions are:
Which states of a biological or mechanical system have some content, and how do we know that they do, and what content they have?
Does what content a state has supervene on its “inner” or “narrow” properties? That is, if two subjects were narrow duplicates, would their states have to have the same content?
Subjects are narrow duplicates (or “twins”) when they have the same neurophysiology inside their heads, and also the same mental images, sensations, feelings, and interior monologue, if any of that isn’t settled by their neurophysiology.
Presumably if we had an answer to the Big Question 1, that would give us answers to 2 and 3. But even lacking an answer to 1, we might still make progress on the more modest Questions 2 and 3.
Kim engages with each of these questions to some degree in his Chapter 8. His “interpretation theorists” think the answer to the How do we know? part of Question 2 is constitutively involved in the answer to Question 1. Other philosophers have more realist views about content. What that means is that, although hopefully we can know things about which states have which content, there are facts about content that are not constitutively dependent on our built out of what evidence we have about them. The other views Kim discusses are all more realist in this sense.
The causal/correlational semantics and teleological semantics that Kim discusses next are attempts to answer the big Question 1. These projects have some fans but they also have difficulties, and philosophers haven’t yet converged on agreeing that either of these, or any other existing answer to Question 1, is correct.
Our discussion will focus on Question 3.
In the early 1970s, the philosophers Putnam, Kripke, and Burge argued that the answer to Question 3 is No. They argued that subjects can be the same in their narrow properties yet differ in what the contents of their thoughts and/or language was.
There are two claims here:
Some theorists will agree that (i) is possible but deny that (ii) is possible. But many philosophers nowadays are willing to agree that both (i) and (ii) are possible. Putnam’s article in fact only talks about (i), but in this exposition I will tend towards talking about (i) and (ii) as a combined package.
There were two argumentative strategies that philosophers used to reach this result.
One strategy involved thinking about narrow duplicates whose physical environment differed. This is exemplified by Putnam’s story about “water,” and other natural kind terms like “lemon” and “tiger.” Putnam mentions all of these, but focuses on the first. (Other natural kind terms include “gold” and “elm” and “aluminum,” each of which Putnam mentions but to make other points.)
The other strategy involves narrow duplicates whose speech community differs. Putnam discusses this strategy too. It can be applied to natural kind terms (he does so with “elm” and “aluminum”) and also with disease terms like Burge’s “arthritis.” (Kim explains how Burge’s example works.) It could apply to a wide range of words.
There are cases where I have something specific in mind when I use a word — and if my usage happens to differ from my community, in ways I might not realize, still what I have in mind is what I mean by the word. For example, maybe I mislearned the word “enervate” and think that it means something like “energize.” Then when I say “This coffee is really enervating,” other people who understand “enervate” the more standard way would think I mean the coffee is weakening or exhausting (and may be confused why I would think that). But I could insist that what I mean is that the coffee is energizing. This is a case where I have the rule that “enervate implies energize” in mind and mean for it to be authoritative. Other people have different rules in mind. Then there could be a discussion of whether my words mean what I want them to mean, or mean what most English speakers would understand them to mean, or whether they’re ambiguous, or what have you.
That’s not a discussion we’re undertaking. We don’t want to think about cases of that sort. There are also cases where I don’t have complete rules in mind that I mean to be authoritative for a word. I mean to be deferring and using the word in the way everybody else does. I may have some opinion about whether the word applies in some situations, that can turn out to deviate from other peoples’ usage. But these aren’t supposed to be cases where I’d say “Well what I meant was…” Instead, they’re cases where I’d say, “Well I guess I was wrong…” Burge’s arthritis example is supposed to be of this sort; and so too Putnam’s elm/beech and aluminum/molybdenum examples.
We can also think about cases where the subject has no opinion whether the word applies in some situation, but trusts that their larger speech community has decided on rules that would settle the issue.
Putnam and Burge say we can imagine subjects of either of these sorts, and the subject and their duplicate live in different social environments, and this makes a difference to what their words end up meaning when they use them.
Putnam admits that words like “hunter” and “whole” may work in the way his traditional opponents said all words work. These may be words where to understand the word, you have to have certain rules in mind, and those rules completely fix what the term means and applies to. For natural kind words and words where we “divide the linguistic labor” and defer to our larger speech community, he’ll argue the traditional view is definitely wrong. Then there are words like “pencil,” “chair,” “bottle,” and “pediatrician,” which he seems to think could be argued to work either way. He’s not sure what to say about them.
We’ll focus here on the “water” example. Putnam will argue that narrow duplicates in different physical environments (one on Earth where there’s H2O, the other on Twin Earth where there’s XYZ instead) might not both be talking/thinking about the same thing when they say “water.” When we say “water,” we’re talking about something that’s essentially H2O, though that’s not something you need to know just to understand or have the concept “water.” Similarly, when we say “tiger,” we’re talking about a species that essentially has some genetic/physiological properties; but our narrow duplicates might use the word to talk about species that have different underlying properties.
Caution! When we talk about biological properties being essential, that can lead the discussion into issues about whether, for example, racial or sex properties are essential, and those issues can be politically charged. Some people enjoy controversial charged discussions, and some of those discussions need to take place, but others find them uncomfortable. However you receive the prospect of such discussions, note that they’re not discussions we’re undertaking here.
Nobody in the debates we’re considering are making any specific claims about which biological properties are essential, to tigers or humans or any other species. (Presumably some of our biological properties are accidental, even if one thinks others are not.)
Also, there’s a confusion that can arise when we consider claims like “Tigers essentially have genes so-and-so.” Suppose Tatum is born a tiger. One philosophical thesis would be that if so, then this individual tiger Tatum could not possibly exist without having the relevant genes. That’s not the kind of thesis we’re aiming to think/argue about and assess. We’re thinking instead about the thesis that Necessarily, when something is a tiger, it has those genes. That second thesis leaves it open whether Tatum is essentially a tiger or could possibly be a different kind of animal. It just says that, in any possible situation where Tatum is a tiger, Tatum has those genes. In other words, our issue is which properties are essential to the kind, tiger, rather than which properties are essential to individuals who happen to be (whether essentially or accidentally) tigers.
This kind of confusion doesn’t get started when we talk about “water,” because we don’t talk about “individual waters.” Terms like “water” and “gold” are mass nouns. The ambiguity described in the preceding paragraph only comes up when you’re using count nouns instead.
Before we get into the details of Putnam’s discussion, it can help to clarify some words he uses.
He talks about the extension of words like “water” and “tiger.” Some philosophers also or instead use the words “reference” or “denotation” here. (There’s a stronger tendency to use these latter notions when you’re talking about words that only apply to one thing, and a stronger tendency to use “extension” when you’re talking about words that can apply to many things. But these are only tendencies.)
Those notions contrast to a group of other terms philosophers use, such as intension, or “meaning,” or “concept,” or “content.” Sometimes these words are used interchangeably, other times philosophers give them different specific technical meanings. In the present discussion, we’re not going to distinguish them.
We already mentioned the extension/intension contrast earlier in the class. The basic idea here is that words like “lawyer in North Carolina” and “golfer in North Carolina” can have different intensions/meanings, even though it may turn out that they happen to apply to the same sets of people. That is, the words have the same extension even though they have different intensions/meanings.
Another notion Putnam talks about is methodological solipsism. Bare or metaphysical solipsism is the thesis that nothing in the world exists except for my own mind. Methodological solipsism is a much less radical view. It is the proposal to study mental states as though they don’t require or imply the existence of anything else in the world but the subject who has them. So the idea here is that if I believe that two is a prime number, then this belief state of mine is one that could be had no matter what the world outside me is like, or even if there is no world outside me. The phrase “methodological solipsism” is relatively new, but the idea behind it is the traditional, long-standing approach to thinking about the mind. Descartes was a prominent example.
Another way of expressing the idea of methodological solopsism is that mental states are “narrow” states, ones that supervene on a subject’s internal or narrow properties, what’s “in the subject’s head.” If you had a narrow duplicate of the subject in a different physical or social/linguistic environment, they should have all the same mental states.
Now, Putnam doesn’t directly challenge this view in the selection we read. Whenever he talks about “concepts” or “psychological/mental states,” he always uses those words in the sense understood by the methodological solipsist. Sometimes he says this explicitly, and writes for example “psychological states (in the narrow sense).” Other times he leaves it implicit.
Using words that way, Putnam’s argument in our reading is that what your words mean is not a matter of what your psychological states are (in the narrow sense).
One could use the same style of argument, to argue, more boldly, that the mental states we ordinarily experience and ascribe to ourselves and others, and guide our behavior and thinking by, should be understood in a “wide” sense, that is, as having contents that don’t narrowly supervene. You might be thinking that water is delicious, and thereby be having thoughts about stuff essentially made of H2O, while your narrow duplicate has thoughts about a different liquid. This would be externalist not just in sense (i) but also in sense (ii):
Putnam doesn’t do that in this selection, but he was glad to do it in other places. As I said, I will tend towards talking about (i) and (ii) as a combined package, just to streamline our discussion. But it’s worth noting that when Putnam says “concept” and “psychological” in this reading, he always means “as the methodological solipsist understands those notions.”
Putnam identifies two theses or assumptions that he says philosophers have traditionally accepted both of.
This has the consequence that any two subjects who are narrow duplicates will grasp/understand the same meanings/concepts.
If these two assumptions are accepted, it follows that if two subjects are narrow duplicates, then their words must have the same extension.
Putnam’s argument will try to tell a story where two subjects are narrow duplicates but their words have different extensions. He writes (p. 584):
[These assumptions entail that] two speakers cannot be in the same [narrow] psychological state in all respects and understand the term A differently; the [narrow] psychological state of the speaker determines the intension (and hence, by assumption 2, the extension) of A.
It is this last consequence of the joint assumptions 1, 2 that we claim to be false. We claim that it is possible for two speakers to be in exactly the same psychological state (in the narrow sense), even though the extension of the term A in the idiolect of the one is different from the extension of the term A in the idiolect of the other.
Putnam’s story involves comparing a subject currently on Earth to another subject currently on Twin Earth. On Twin Earth there’s no H2O, but instead XYZ in all the same places. XYZ superficially looks and behaves a lot like water, but there are fundamental and important differences in the underlying chemistry.
At a certain point in the story, Putnam invites us also to consider our ancestors in 1750, and their counterparts back on 1750 Twin Earth. This date is chosen to be before the development of modern chemistry, and anyone knew what “oxygen” was. Current people on Earth and Twin Earth may not be narrow duplicates of each other, because on Earth people know that water is H2O, but on Twin Earth nobody knows or even thinks that. They instead have thoughts about the stuff they call “water” being made of XYZ. But our ancestors back in 1750 could be narrow duplicates of each other. (Imagine that it’s irrelevant that the Earthlings have H2O in their brains and bodies but the Twin Earthlings have XYZ there.)
Putnam’s first step is to argue that as currently used on Earth, “water” includes in its extension H2O everywhere in the universe (it would even include any H2O on Twin Earth, if there were any), and excludes XYZ everywhere. So if I say “There’s lots of water on Twin Earth,” what I say would be false. Whereas if a Twin Earthling says, using his words, “There’s lots of water here,” what he says would be true.
That can sound natural. What views is Putnam opposing? One alternative would be to say that my word (and also the Twin Eartling’s word) applies to both H2O and XYZ everywhere. They are both water, just chemically different kinds of water. A different alternative would be to say that my word applies to H2O on Earth and XYZ on Twin Earth. (So if there’s any H2O on Twin Earth, it wouldn’t count as “water.”) That would also be a meaning that the Twin Earther’s word could have too. If my word had either of these meanings, then when I say “There’s lots of water on Twin Earth,” what I say would be true.
But Putnam thinks it isn’t true. He thinks in my mouth, “water” just picks out H2O, not XYZ. Regardless of what location or situation I’m talking about. Even a possible situation. In every possible situation I describe using “water,” I’m always talking about H2O. That’s why (what I call) water turns out to essentially be H2O.
Putnam’s second step is to acknowledge that current Earthlings and current Twin Earthlings are not narrow duplicates, because the former have thoughts they’d express using the concept H2O, but the latter do not. Nowadays, we all know that water is made of H2O.
If we consider our 1750 ancestors though, they didn’t know that. Nor did the 1750 Twin Earthers know that the stuff they called “water” was made of XYZ. So these people might be narrow duplicates of each other.
The third step is to decide what to say about the relation between our use of the word “water,” and the use of our 1750 ancestors. Certainly we know more about the stuff water than they did. But does our word “water” have the same meaning and extension as theirs?
Putnam thinks it does.
The alternative view here would be that for us “water” has come to mean and apply only to H2O, and that’s now part of what someone who has the concept of water needs to know. But back in 1750, it had a broader meaning, that applied to XYZ too. So if our 1750 ancestors had said “There’s water on Twin Earth,” those words in their mouth would have been true. When we discovered that all the clear, transparent stuff in our rivers and faucets is made of H2O, we changed our meaning/concept of “water.”
Putnam opposes that. He argues that “water” picked out H2O (and only H2O) all along, and that our ancestors meant for “water” to pick out a certain underlying microstructure, that they hadn’t yet identified. (Remember back in his discussion of “polio” in the paper on behaviorism, Putnam argued that we also intended for “polio” to pick out an underlying disease agent, even if we didn’t yet know it was a virus or which virus it was.)
Now, if Putnam is right that the extension of my word differs from that of current Twin Earthers, and that my word didn’t change (get more restricted) in its extension between 1750 and now, and neither did the Twin Earthers, then it follows that our 1750 ancestors were narrow duplicates of people on Twin Earth, but their words have different extensions. This is the result Putnam was aiming for.
It also follows that people like our 1750 ancestors can use and understand a word “water,” that in fact applies to all and only H2O, even in any possible situation they talk about. So it would be true for them to say “water is necessarily H2O” (even though they didn’t yet know that it’s true). At the same time, since they don’t know that water is made of H2O, they can reasonably and coherently imagine water being made of other chemicals. For all they know in having the concept, it could turn out that water isn’t H2O. Hence, that’s something they can coherently imagine but it’s in fact not metaphysically possible. (See Putnam p. 590.)
Similarly, necessarily when something is a tiger, it has such-and-such genes. Even though, if you didn’t know this about tigers, you might still have the concept “tiger” but be able to coherently imagine them having other genes, or no genes.
We encountered this divergence between imaginability and possibility earlier in the term, when we were discussing arguments for dualism. These are the discussions where the contemporary separation of these notions started.
Remember that Putnam ascribed two assumptions to his opponents:
He’s now argued that (at least) one of these has to be rejected. Which one?
For some words, Putnam thinks that Assumption 2 should be rejected, but philosophers already knew these words needed special handling anyway. These are words we call “indexicals,” and include “I,” “now,” “this city,” and so on. With these words, it’s very plausible that when I use them on Tuesday in Greensboro, and you use them on Wednesday in Carrboro, they have the same meaning. We both speak the same language. But in my mouth, these words pick out a different speaker, time, and city, than they do in your mouth.
What about words like “water,” “tiger,” and the rest? One option would be to treat them along the same lines as indexical words. Then we could say that “water” has the same meaning in our mouth and the Twin Earther’s mouth. It’s just that when we use the word, that meaning picks out all (and only) H2O, but when they use it it picks out XYZ. Some philosophers have argued for this view.
But Putnam at the end of his paper says he thinks it isn’t the right way to go with “water” and “tiger.” For these words, he prefers instead giving up Assumption 1. He’d say that our word “water” and the Twin Earthers’ word have different meanings, even though we are narrow duplicates, and so our (narrow) psychological states are the same.
As I said, other philosophers and Putnam in other places are happy to continue even further down this path, and say that not only is the meaning of our words different, but also we are thinking thoughts with different (wide, that is, not narrowly-supervening) contents. When we and our 1750 ancestors think to ourselves, “The garden needs more water,” our thoughts are about a stuff that’s essentially made of H2O not XYZ. Our ancestors just didn’t know that yet.