I am very grateful to Linas for his comments on our mental concepts of chairs and tables. He knows that it is 'something more' than what has been said in the discourse. I am not so sure. There have been many acts of what I call primordial speech situations in which chairs and tables have been pointed out to me. For me, such things are props that make up an extension of the speech situation. The act of ostentation make them part of the communication. If I had a dog it wouldn't know what a chair is. It would not even have noticed it had been part of such a speech situation. I recall discussions with people (not with dogs, though) in which we attempted, for instance, to negotiate if barstools with a back rest should be referred to as stools ("Could you bring two more stools over, please?"). Does the result of such speech situations and negotiations mould my mental concept? I think it does. Does each of us have the same (perhaps innate) concept of a chair (as Fodor sees it)? Rather unlikely. Does everybody have their individual concepts? If yes, how do we understand each other, except by assuming that meaning is in the discourse and not in our heads?
How do we learn what is the 'true' representation of "the catch"? Do I really have to have the experience to learn what it means? I have absolutely no idea about rowing. Can I find out about sculling and the catch? Google points me to Joe Paduda's The Art of Sculling. There I read: "The other critical part of the stroke is the catch. While the release is important because you want to get the blades out of the water without sacrificing any of the work you did, .. on the drive, at the catch you want to ensure that you don't slow the boat down when you reverse direction. If you must be quiet about the release, you must be absolutely silent at the catch. Think about the catch for a moment. ..." And so it rambles on, for three pages. Isn't this the kind of stuff people are told when they learn about sculling? No one, I believe, who hasn't been told about it will experience it as "the catch."
I do not at all deny that people have memories and the ability to make use of them. But memories as such cannot be inspected. My problem as a linguist is that I have to get my assumptions accepted by other linguists. When I say that people have broad mental representations of things like chairs and tables in their minds, then I might be asked if I can produce evidence for this claim. This I cannot. The only evidence is what people tell me. I have no idea what there is in their minds. All I have to work on is their utterances. There might be mental concepts (even though we don't know anything about their nature), but a table is what people say a table is. No more, no less.
John reminds us of the way language and experience are connected in Halliday's and Matthiessen's account. I find these ideas inspirational; they make us ponder, and it doesn't matter that they are rather hypothetical and cannot be tested. My problem is that we hardly ever come across what is pure, unmediated experience. It is a bit like being in a state after the second bottle of wine when we still act quite 'normally' but have ceased to reflect on what we are doing and saying. The only pure experiences are those of which we have no conscious memories. Dogs may have unconscious memories; they may bark at someone who has abused them years ago. I sometimes walk along a street and I register a smell that brings up a long-forgotten, very faint and unpindownable memory of my childhood. But it is being aware, being conscious of an experience that turns a non-symbolic 'feel' into something that can be memorised and verbalised, that can become something I can tell others. It works the other way around, too. Having learned what a "catch" is I will eventually experience it. In the same way we experience passionate love. In societies in which it is not a discourse topic, in which it is not shown and not acted out, it doesn't exist. People don't experience it. (Back to Wittgenstein: there is no private language.)
As amsler points out, our mental concepts of 'vehicle' (if we have them; at least the way we use the expression 'vehicle') will change one we begin negotiating their meaning. There is no point to model the mind and the way it processes language by dealing with lexical items if there are data, real language data, that show us how it is done. Geoffrey reminds us that we do not have to look into a translator's mind to find out how translation works. All we have to do is to analyse parallel corpora, evidence available to all of us in the room. This is how translators learn what to do with their translation units: they look what others have done. Translation is providing an interpretation of a text in a target language. As long as it is a natural language text, and not a controlled language text, only human beings are good at it. And only people can argue with me why this translation is better than another. Linguistics is, I believe, concerned with the interpretation of meaning, language engineering is good for many things, including the translation of texts written in controlled language.
Linas also says that computers can 'grapple' that 'the sky is blue' means the sky is blue. I am afraid, my clock, when I last talked to it, still hasn't grappled the meaning of time, or even of a quarter to seven. That doesn't keep it from ringing then. But I doubt if it really knows what it is doing. Very clever computers with access to a lot of real language data may even tell us when to translate 'blue' as 'ble' and when as 'azurro'. Computers will be programmed to do all kind of fascinating things. But they don't know, and don't understand, what they are doing. The same data crunched by the same software will deliver the same results, time after time. The same poem will never be interpreted by the same person in the same way. There is no meaning without intentionality. But intentionality can never be observed inside people's minds, only as what is happening, in terms of symbolic exchanges, between people.
The cognitive sciences cannot tell us a lot about an individual's intentionality. Neither can corpus linguistics. But it can tell us a lot about the intentionality of a discourse community.