In Chapter One, we give three statements of the main tenets of the Correspondence Theory.
Correspondence Theory of Truth
Correspondence Theory of Meaning (Classical)
Correspondence Theory of Meaning (Modern)
All of these seem reasonable--after all, if we are going to talk about the world, say things of it which are true (and sometimes false), there must be something which connects the vehicle of expression of these thoughts, propositions, sentences, in short language, to that world. In other words, language must refer to that world. In turn, the world must be ``mind-independent''. Otherwise, what we say will be true only relative to some set of concerns, some internally constructed system, what our words mean will itself be relative to the system within which we attach meaning to them. Hence, the term ``Referential Realism'' to refer to this set of views.
In Fact and Meaning, Jane Heal defines ``minimal realism'' as follows. First, it must acknowledge the possibility of incompatibility between claims, though this incompatibility is not necessarily a contradiction (12); second, it must have a commitment to "mind independent" states of affairs (16). The ontology of Word and World certainly meets these conditions. There is a third principle which follows from the first two. "Taking a realistic stance towards some subject matter involves supposing that when people differ about which one of a set of incompatible moves should be made, further investigation will produce either agreement on the question or recognition by both parties that no firm judgment should be made" (18-19).
All the versions of Correspondence Theory given meet these requirements. The cost of failing to accept them is, it seems, Relativism of the most Rampant sort, including meaning relativism or, as it's more aptly called ``meaning scepticism.'' This is the view that it is simply impossible ever to hope to be able to say anything that might have any bearing on anything outside our minds.
Modern philosophy of language has been dominated or--as Wittgenstein would say--``captured'' by this picture, whether as presenting an unobtainable ideal or as the foundation of meaning, reference &c. This is not to say that this is the only meta-theme in philosophy of language, only that it is an important and influential theme.
The key to Referential Realism is that whatever the primary content bearing entities in language turn out to be be (words, sentences, theories), their ``existence and nature owe nothing to convention'' (16).
This doesn't mean that there aren't parts of language which are entirely conventional. For example, it seems undeniable that the fact that use this or that sound or mark to signal some content is entirely a matter of convention, and that linguistic entities which do not function to bear content are wholly conventional. What is important here is that if some linguistic entity is content bearing, it must find its significance, its content, outside language itself, where language is seen as some (at least partly) human invention.
What do we mean here by ``content bearing''? Quite simply, ``conveying or carrying information about, or purportedly about, the world outside language.'' How does Referential Realism carry this out in practice?
Take two statements.
There is a sense in which both can be seen as true. This is the sense in which we look to some group, some set of practices to see if the terms have a meaning which allows them to be strung together to make well-formed sentences which, in turn, function within the accepted system of belief. Thus, for modern scientists, (1) is true, and for members of various societies, (2) is also true.
But there is another sense in which we want to say that while (1) is true, (2) is simply false. Referential Realism gives us an account of this.
But the first option, a, is really a false option. The problem with (2) is not comparable to the problem with (3) The speed of light is 200,000,000 m/s (metres per second), which is about the speed of light but just happens to get it wrong. The problem with (2) is that it's about nothing at all, it's nonsense at best, but it's not even clearly nonsense since its problems seem to go deeper than that. The second alternative,b, is a more accurate portrayal of what the problem really is: (2) just doesn't get us out of our minds because it can't correspond to anything. (1), by contrast, takes us out of our minds because it does, and (3) takes us out of our minds because it could. It is the possibility of corresponding with reality, this central tenet of Referential Realism, that allows us to escape the Prison-House of the Mind, the Prison-House of Language.
Seen in this way, the task of the philosopher of language is to show how this correspondence, this reference is possible, how it happens. In Chapter Two, we outline three varieties of Referential Realism.
Some of the reasons for this classification are obvious, others are surprising.