Buzzwords
No doubt those who have made their way through the nodes REFCORR and MODEL are now feeling an appropriate sense of revelation. Apparently, we ought to feel indebted towards these authors who have now decisively "won" the centuries of war against the "reflection-correspondence" theory, thereby illuminating as never before the tricky concepts of truth, meaning, thought and language. Let us remind ourselves how this was achieved... well you see, the new conception is *dynamic* where the old one was *static*
Sorry guys, but it's going to take more than buzzwords, strawmen and revisionist history to persuade us that there's anything really new or important in these ideas.
First off, it is quite silly to argue that if you construe the meaning of language in terms of correspondence of words and sentences to real objects and states-of-affairs, then you are inevitably resigned to believing that "any expression of our language which cannot be immediately interpreted in terms of observable facts, is meaningless and misleading". (This is a gross misrepresentation even of verificationism, never mind a more general truth-correspondence theory). Just because something is a real object or pattern or phenomenon, there is no need for it to be easily and unambiguously detectable. Let me elaborate:
The psychological state (beliefs, desires etc.) of a person is determined by a complete description of the state of the person's brain, and must therefore count as being in some sense a "real phenomenon", albeit a very high-level one. Analogously, the presence of a 'fork' in a particular chess position is a "real" and "objective" feature of the position, because once the raw position is determined, everything about the fork is also determined. However, whereas it is relatively unproblematic to translate from talk about forks and pins into talk about what moves are advisable for each player, it is evidently much harder to make clear and precise predictions about observables on the basis of a person's beliefs and desires. There is no reason, though, to deny that it makes sense for words and sentences to "refer to" or "correspond with" or "reflect" a person's psychological state (indeed, I couldn't very well have made this argument otherwise).
The alternative to "reflections", seemingly, is "models". This is motivated by raising some undeniably tricky and important questions:
"what does it actually mean that one thing "reflects" another. Also, how do we know that reflection takes place?"
These questions don't appear to me any easier or more difficult if they are recast as follows:
"what does it actually mean that one thing "models" another. Also, how do we know that modelling takes place?"
And I'm sorry but defining modelling in pseudo-mathematical terms is not a helpful way to go about answering them: What you get is a quick and easy (though it sounds "complicated" to an outsider) answer to the question "what is a model", but scratch the surface and it's apparent that all the important questions are left open. For instance: How do we determine the sets of states-of-affairs of a system and its model? When exactly is a model physically instantiated? Can there be such a thing as a model that is not physically instantiated (can we regard the addition of natural numbers, say, as a model of a certain property of collections of marbles)? In what contexts can it be said that a model affects the system that it models? How can we determine the reliability of the predictions of a model?
I believe that a framework that gives us satisfactory answers to all of these kinds of questions will have no need for vestigial flaps of mathematical formalism like that found in the "MODEL" node.
