It does seem rather clearly to refer to a particular marble, but that's not really true. It's more that when I talk to you, I'm trying to act on you in some way. The phrase doesn't simply refer to a particular marble. The phrase wouldn't refer to anything if it wasn't for complicated social conventions. The relationship between the phrase and any real marble is extremely complicated, and it's necessarily dependant on the primary relationship between me and you, the mutual language-speakers.
Well, part of the question is whether swampman is really participating in those conventions.
I think some of the attempts in this thread to maintain a causal history with swampman are missing the point, and wont let the argument get past the initial assumption: that swampman was uncaused, or at least not causally related to Don, and therefore its thoughts are 'isolated' from the world. Swampman's thought "the marble I saw yesterday" did have a direct referent in Don as a thought in Don's brain, and Swampman shares the same brain configuration, but Swampman never had the experience of the marble, and the fact that it thinks it is referring to a specific marble is just a memory illusion.
I agree with SRG in the general thesis of externalism, that the meanings of our words depend not only on causal history and internal states, but social conventions and relations among brains. However, it is still a question of whether or not Swampman is considered a part of those conventions. He can act as if he is part of those conventions; the question is whether or not that behavior needs to be caused in the right way.
How do people come to participate in social conventions? Through a long process of learning, training and exposure. Thgis is the socialization process we all went through to some degree; and so did Don. Swampman, however, went through no such process; it just so happens that he reaped the benefits of Don's social nature. But he never learned to gain these abilities, he just has them.
Lets imagine a small twist in the case. Say Don is playing poker when the lightening strikes the pond. It kills Don instantly, and fries anything he is touching (including his cards), but say Swampman happens to be created not only with exactly Don's internal configuration, but also Don's exact poker hand- a straight flush. Swampman runs back to Don's house, throws his newly formed cards on the table and yells 'I win!'.
Is swampman really playing poker? Is he legitimately participating in the game?
I'm not denying that "the marble I saw yesterday" is understood to refer to some particular marble. I'm arguing that this is an oversimplification of the way language works. It isn't like a formal system where one word simply stands for one fact of the real world. I guess I'm basically trying to argue Ludwig Wittgenstein's point of view in the Philosophical Investigations. It's a very hard and abstract argument to make, and he does it mainly through thought experiments, and I'm not sure if I can do it without derailing the thread, or even at all.
Basically the relevance to the thread's topic is that I don't think it's a real problem that when Swampman talks about "the marble I saw yesterday" there isn't a real marble he saw yesterday. That way of thinking about language is so computer-like, like if he uses a symbol without a real referent there will be a null-pointer error and the program will crash. I'm arguing that for "the marble I saw yesterday" to have real meaning, there doesn't have to be a real marble, because the meaning of the phrase is found more in the context of its use -- the phrase isn't simply a pointer to an object.
Think of the different ways I could use the phrase "the marble I saw yesterday." I could tell you "The marble I saw yesterday was purple." I could ask you "which one of these marbles was the marble I saw yesterday?" You could ask me "What's your favorite marble in the whole world?" and I could respond "The marble I saw yesterday." In all of these cases I mean to interact with you in a certain way by using the phrase, and in all of these cases I'm understood exactly the same whether I'm actually the person who saw the marble or a duplicate. The phrase "the marble I saw yesterday" fulfills its purpose in language irrespective of actual causal history.
I agree with you generally, but I think you are overemphasizing the role of social convention in language use. For words to have meaning, they must be part of a social practice, granted, but there are rules associated with that social practice, and if they aren't codified to some degee in the internal structure of the language user then they can't even begin to be part of the practice.
I dont think you deny that. But all this means is that when the swampman uses his words, we can all understand them as meaningful. And of course we can. We wouldn't detect any difference in the swampman from Don whatsoever. When Swampman says "The marble I saw yesterday", if he is speaking to english speakers, his interlocutors can all understand and recognize his words as in their own language, through deference to social convention, and consider him to use his words meaningfully.
But that is just recognizing how we might respond to the swampman. I am asking a question about the swampman himself- does he mean things by his words? What are his words grounded in? Are we warranted in allowing him into our social practices?
When Swampman says "the marble I saw yesterday", there is no referent to that sentences; nevertheless we can understand what he means because our understanding of the words doesn't rest entirely on causal history. No one denies that. The question is, what does the Swampman himself mean by his words, and is that enough to make him a legitimate language user?
Lets go bach to a much simpler, but relevant case. My ATM says 'Thank You' when I pull money out. Now, looking at those words from the perspective of a social practice where we say 'Thank You' in instances of gratitude, and I can understand the meanings of the words printed to screen. But clearly the ATM doesn't mean anything by those words. The ATM is just blindly printing stuff to the screen, that happens to mean something when read by an english speaker. Part of the reason it can't understand is because its a dumb ATM, and lacks any conceptual or psychological resources for making sense of its words. But part of it is that it was made to say those words, by design. Its not using the words at all.
Back to the swampman- just like the ATM, we can understand it as meaning things by its words from our social background. The question is whether the swampman himself means anything by his words. Swampman has the psychological resources to make sense of his words, of course, but they aren't grounded in anything- he says "The marble I saw yesterday" because thats how he was constructed. The lightening bolt might have struck and given him a memory of seeing a 20 foot tall cow instead.
Also, appealing to Parfit sort of examples (teleportation, me five seconds ago) misses the point. If it were true that me 5 seconds ago were somehow radically causally disconnected from me right now, that would give me pause and fill me with a sort of radical skepticism about the world; but that sort of skepticism is something we want to avoid. Its not something we want to appeal to in these swampmen cases as an intuitivly obvious starting point.
So the question is not whether we can understand the words swampman utters. The question is whether we are justified in including swampman in our social practices, and whether he is actually participating in them.
I don't see any compelling argument that Swampman might be excluded from our social practices just because he was created by a lightning bolt. There's good reason to consider the ATM machine excluded. It doesn't have the mental faculties to understand what it's saying. Its words mean something to us, but the understanding is strictly one-sided. Swampman, on the other hand, has the same mental faculties as Don. He is understood, and he himself understands. He might have inaccurate memories of his own history, but that doesn't suddenly exclude him from society.
He says "The marble I saw yesterday" because thats how he was constructed, yes. The same is true for Don. The only difference is that Don was constructed differently from Swampman, over time. Don himself might have had his memories altered, or he might just plain misremember. It doesn't make a difference. Don and Swampman equally know what they're talking about. Whichever one it is talking about the marble, he has a certain idea in his head, and vocalizes expecting to be interpreted in a certain way. The act is identical in either case.
The only argument I can see that makes Don's speech different from Swampman's speech is if you oversimplify and overemphasize the connection between some physical marble and the phrase "the marble I saw yesterday." It doesn't matter if such a marble really exists. It could be an imaginary marble. Everyone could be wrong about which marble Don saw yesterday. None of that "breaks" communication, as long as people are still exchanging ideas and influencing one another.
I just don't think this is a particularly formidable problem. You always seem to see problems where I don't.
You still aren't addressing my problem. You are focusing on the content of the words, and you recognize that 'communication' can take place whether or not the words have the same causal history. I am granting that.
But just because something is communicating information doesn't mean it is a genuine participant in an activity. The rolling boil of water conveys information about the temperature of the water as I am cooking; but the water isn't participating in the activity of cooking. It is just a thing that I use to cook with.
Now lets say I brainwash you into responding "I saw the marble yesterday" anytime you are prompted with the question "When did you last see the marble?" You respond that way entirely reflexively, without any hesitation. In such a case, we wouldn't be inclined to say you really mean anything by your words, even though I can still understand your words as meaningful. A 3rd person observer would see our exchange and think it was a genuine case of communication, that information was genuinely being exchanged, when in fact it was not.
This is of course a different case than if no one has messed with your brain, and I ask you "When did you last see that marble?" If you take a moment to reflect, and respond "yesterday", I am inclined to say that you do mean something by those words, and you are conveying information.
Your responses so far seem to deny there is any difference in the above cases, because in both cases the words gain their meaning by social convention. But thats not the point. In the former case, you aren't a genuine participant in our the linguistic activity, in the latter case you are.
The qeustion, then, is the swampman more like the former case or the later case. One might say the swampman is actually consulting his memory, reflecting a bit, and so on, before he answers. In other words, he is using his rich conceptual resources, unlike the brainwashed person, and so is more like the latter case.
The problem is that his 'rich conceptual resources' are entirely empty, entirely the result of construction, and have no genuine reference. Although he uses psychological resources to come to an answer, his psychological resources themselves are the result of construction. They were implanted in his mind, in just as designed a way as the brainwashing example. Note also that just because swampman is identical to Don, Swampman has no evolutionary or genetic history, and has no place in our genetic tree. Not only hasn't he learned our social conventions, it seems like he isn't even a person.
So sure, we can treat him as if he is a genuine participant in our activities, simply because we would never find out otherwise. The question is, is this treatment warranted?
The brainwashed man is like the ATM machine. He mindlessly outputs some phrase that, while it has meaning to others, does not have meaning to him. It's that fact that makes him different -- not the means by which he learned the behavior. It's not important whether the psychological resources necessary for participation in society were put there deliberately, randomly, or by some natural process. What's important is they exist.
This is my position, and I don't find the Swampman thought experiment troubling to it.
Now if the man was brainwashed in some incredibly complex fashion such that his "mindless" responses were novel and appropriate to various situations &c &c, then maybe they wouldn't be mindless anymore. But this is covering ground we've already covered in other threads.
Standard corellary example:
An ant is walking along a beach, and his path traces out lines in the sand. A man wanders by and sees the ant's path, and to him it looks like it spells out "To be or not to be..." in perfect Times New Roman font.
Now, obviously the ant did not mean to write that phrase. But the ant isn't entirely mindless- its behavior is a complex process that is the result of a living creature.
And obviously the phrase can have meaning independent of its origins/causal history. The man understands what it means, and recalls the Bard and all the social conventions tied up with that phrase, even if they were entirely independent of the intentions of the ant.
I think we agree on all the above, right?
The question is, to we count the ant's behavior as an instance of language use? That is, does the ant's behavior count as participating in our linguistic practices? As part of our conventions?
Well, in some sense the ant is using language, just like in some sense the ATM machine is using language. Language is being used. Words are being interpreted. In another sense, the ant isn't using language, because it isn't consciously directed toward using language. It's just a coincidence. It's hard to say the ant is an active participant in language use when it isn't trying to accomplish anything by using language. It's unlikely that the ant will continue to [appear to] participate in our social conventions based only on random chance. In any case, the ant is totally unlike Swampman, or even the ATM machine. It doesn't use language consciously, and it doesn't even use it predictably.
I don't think a thought experiment based on random chance is very applicable, because it's so unlikely. In reality most things will either be robust language-users or simple automations, and we'll classify the former as participating in our social conventions, and the latter as not really participating. There might be some grey area between the two categories, but it's not likely we'll find anything that appears to consciously participate when in reality it isn't directed toward that goal in any sense.