Externalism about Mental Content
Meaning just ain't in the head
Externalism is the thesis that mental content is not solely constituted by what is in the head but at least in part by things outside it (non-philisophical E2 readers bear with me and hopefully the explanation will make some sense). One explanation of this is put forward by Hilary Putnam with his Twin Earth thought experiment. He tries to show that the meaning of your words, and even further, what you are thinking, is constituted by things outside of your head as well as inside it. The thought experiment is supposed to illustrate this by giving an example of two people with identical brain states who are having different thoughts. We are supposed to conclude that mental content "ain't in the head". This view, as we will see, depends on a certain basic intuition, and I will discuss the plausibility of this, before going on to look at a slightly different challenge to the externalist thesis as put forward by John Searle. Searle believes that two people can have identical brain states, yet be having different thoughts, but he does not believe that this means mental content ain't in the head.
Putnam's case for externalism: Twin Earth
In Putnam's thought experiment, we have a protagonist, Oscar, who makes the assertion "water is wet". Oscar lives on Earth in the year 1750, but there exists a Twin Earth, almost identical to our own, including an identical Oscar, who we will call Twin Oscar. Oscar and Twin Oscar are in identical physical states, and are having identical phenomenological experiences. On Twin Earth, the surface properties of water are exactly the same, but it is not H2O, it is XYZ. The residents of Twin Earth call XYZ 'water', but we will call it 'twater'. Twin Oscar, then, making the concurrent assertion that "water is wet", is talking about something different from Oscar. If we have them both saying "water is H2O", it is even more evident that they mean different things, as the truth value is different (Oscar is right but Twoscar is wrong).
Putnam's intuition is that what we mean when we say 'water' is whatever is identical microstructurally (at such close range that we might not ordinarily be able to perceive) with what we are looking at when we point at some liquid and say "that's water", as opposed to merely whatever is colourless, tasteless, and wet (etc). If we accept this intuition, then Oscar and Twin Oscar do mean different things when they think about 'water'. Putnam believes he has demonstrated that this seeming paradigm of a non-indexical thing does in fact work in a similar way to an indexical (an indexical could be a word like 'you' or 'tomorrow') - in other words that the meaning of water is relative to relationship between the speaker and the environment. As Oscar and Twin Oscar's meanings differ, they have differing mental contents. Remembering that their brain states are type-identical (i.e. physically exactly the same), this is supposed to lead to the conclusion that content 'ain't in the head', and this is the externalist thesis.
Do we define kind terms indexically?
If we want to deny Putnam's case for externalism, we may well have to deny that content differs between the two twins. The most obvious way of doing this is by rejecting Putnam's intuition that microstructure determines what we mean by 'water'. Searle identifies what he calls a traditional intuition. This is that with regard to natural kind terms, prior to scientific knowledge, we do not define them as whatever is identical in structure to a paradigm example, as Putnam asserts, but by a 'cluster-of-concepts' as Searle (1983: 204) terms it. So in medieval Earth and Twin Earth, water and twater are the same thing. This seems to be what Segal (2000: 55) means by 'motley' definitions – water is whatever is clear, liquid, (etc.). If medieval 'water' was just whatever is clear, colourless and wet, then there would be no justification for externalism, because Oscar and Twoscar would be having the same mental content (about the same thing) despite different external realities. They would simply both be having thoughts about clear, colourless liquid.
Putnam (1983: 73) gives another example:
Consider the kind lemon. There are lemons which are green rather than yellow; and there could well be a citrus fruit which was yellow and not a lemon. What makes something a lemon is having the same nature (e.g. the same DNA) as paradigm lemons, and not fulfilling some set of criteria (yellow color, thick peel, tart taste,…) laid down in advance. (italics author's)
So he is appealing to the existence of some essence which makes a lemon a lemon, in this case
DNA. The example suggests that if I had indeed only seen yellow lemons but once came across a green one, although I may initially be lead to believe it was a different fruit, when I found out the DNA was the same, I would accept it was a lemon. This and the Twin Earth
thought experiment suggest that just as surface features can vary but the microstructure remain the same, microstructure can vary but surface features remain the same. In medieval Earth, if we are using advance criteria rather taking the definition of water to be indexical to some essence, these would correspond to surface features. Putnam seems to believe the nature of a natural kind cannot be constituted by its basic surface features. If it were, then the advance criteria would be identical to the nature, and we could say content was in the head.
We may be able to criticise his view with an example of a kind that has nothing we can identify as an essence or nature deeper than its surface features. For example, if we found a dog with no DNA, but who was in all superficial respects dog-like, would we really want to say it was not a dog? However, all natural kinds have some kind of chemical composition, so it is not certain that we could find or indeed plausibly hypothesise a kind that was impossible to classify by microstructure. Segal discusses empty concepts, which he thinks are problematic for those externalists who argue that natural kind concepts need to correspond to something in the world. He points out that they would have to say the concept of a ghost was meaningless, or 'motley', composed of general surface features. He says the former is clearly an unacceptable consequence, and that the latter would be ad hoc and out of keeping with the direction of the theory thus far (Segal, 2000: 55-6). This need not be a problem for Putnam. We may imagine he would say that as water and lemons have microstructures, this constitutes their essential nature, whereas there is no paradigm of a ghost that includes microstructure, only surface features, so in this case the advance criteria and the essential nature are the same thing, leaving us with no problem.
Searle (1983: 203) and Segal (2000: 32-3) mention the idea that we could travel between Earth and Twin Earth, find out Earth 'water' is H2O whereas Twin Earth 'water' is XZY and simply choose whether perhaps to rename XYZ 'twater' or to allow both substances the same name. Segal (2000: 132-3) thinks the latter is more likely. Indeed it certainly does not seem we would automatically find ourselves compelled to decide XYZ does not constitute water. Intuitively it seems somewhat plausible that discovering more about the nature of water would cause us to change the predetermined criteria we have for deciding if something is water rather than demonstrating that what we always meant by water is something different. In Putnam’s favour, if we do not count microstructure of some sort in determining natural kinds, then there is excessive arbitrariness. For example, if Putnam is wrong, then if we had called a fox a feline because of its pointy ears, prior to the zoological discovery that genetically it is a canine, we could not say we had previously been mistaken.
If Oscar and his twin have different mental content does this have to mean that content ain't in the head?
Searle argues that we can accept that a kind term like 'water' is indexical even though it may originally have seemed as if it were not. But he does not see this as a challenge to internalism, the theory that content is indeed in the head. When Oscar and Twin Oscar say "water is wet", they are having different thoughts, but Searle believes it does not follow that this means the content is outside their heads. Just because we do not know the microstructure of water, does not mean our heads do not determine what we mean by water – they do exactly this, indexically.
According to Searle (1983: 207), Oscar believes something like "water is whatever is identical in structure to that", while Twin Oscar would be uttering the same phrase. They are different because they refer to different things. But the mental content is constituted by the thought that "water is whatever is identical in structure to that", "that" being causally related to a different thing in each case, but not by the actual microstructure of water.
In each case, we have type-identical experiences, type-identical utterances, but in fact in each case something different is meant. That is, in each case the conditions of satisfactions established by the mental content (in the head) is different because of the causal self-referentiality of perceptual experiences. (Searle, 1983: 208)
We have been working on the premise that what is needed to give something meaning, and thus content, is a full set of
truth conditions. Searle (2004: 187) explains the indexical nature of the propositions allow the truth conditions to be contained in the head, because it has to be relative to the speaker. So we gather that it is a matter of fact about the world whether or not truth conditions are met, but it is not up to the external world what the truth conditions are.
Many would argue that it is simply not logically conceivable that the twins' heads remain the same, their contents differ, but exist solely in the head. Yet what it is expressing is actually quite intuitive. Searle's argument is attractive, because it seems obvious that I should be able to have a thought about something outside of me, something I do not fully understand, yet be the sole owner of the content of the thought. With causal self-referentiality, Searle shows us how we can indeed unite the intuitions that the twins' contents differ, yet exist in their heads, by explaining how it is inner content can be about something even if we do know the precise nature of what the thing is.
Conclusion
Putnam offers a convincing argument for the indexical nature of kind terms at least in some cases. However Searle's argument elegantly shows how this need not lead us to externalism. This is made all the more compelling by the problems faced by externalism, for example the case of empty concepts with relation to some kinds of externalism. Furthermore, Searle makes an appealing case that externalism is based on an intuitively unappealing view, that says intentionality can be constituted by incorporating the external world into mental content rather than explaining how it is the external world can relate to the mind. This debate is very active within current philosophy of mind and there are many other considerations and arguments I have not looked at. My personal feeling, however, is that externalism is not highly plausible.
References
Putnam, H. (1983)
Realism and Reason. Cambridge : Cambridge University Press.
Searle, J.R. (1983)
Intentionality – An essay in the philsophy of mind. Cambridge : Cambridge University Press.
Searle, J.R. (2004)
Mind: a brief introduction. Oxford : Oxford University Press.
Segal, G.M.A. (2000)
A Slim Book about Narrow Content. Cambridge, Mass. ; London : MIT Press.