Physical Constraints

It is a (literally) undeniable fact that the world is the kind of place that we can talk about in the way that we do. This is not quite the same as saying that it is the way we say that it is, but the two conditions are, at least, methodologically indistinguishable.

How does the world guide our words, and so our intelligible thoughts?

It isn't possible to give a complete answer to this question, because that answer itself would have to be a statement, or thought, that the world has guided. We cannot give a (constructive) justifying account of our ability to give justifying accounts.

With respect to 'external reality', we can say a bit more. There is clearly a (somewhat vaguely circumscribed) set of statements that we think of as 'empirical', and which we also want to think of as 'obviously true'. Among these we would probably include the kinds of statements about immediate perception which traditional epistemological empiricism focused on. I've spoken about these, and the errors that our intuitions about them have led us into, but I feel I could be more explicit about what I think about them, if I don't think of them as epistemological fundamentals (which I don't).

The most obvious thing, of course, is that philosophers have tended to focus on the reliability of our perceptions themselves rather than on the intelligibility of our statements about them. They have been vaguely aware of the difficulties about 'shared perceptions', but have tended to think of these in terms of puzzles about perspective.

Wittgenstein's repeated demonstrations (via the 'beetle in the box' parable, and in many other places) that the meanings of our words could not depend on internal processes which others did not have access to should have put a stop to all this, but the tanker of prejudice has proved hard to turn. It should have been obvious to anyone who took his arguments seriously that the 'shared experience' was a projection of the shared meaning and not a determinant of it. We project the contents of our boxes from our shared use of 'beetle'. The question of what is 'really' in each box cannot even arise - and so is strictly meaningless. So long as 'beetle' works in the language, we have as much access to the 'shared content' as we need. We can only ask for more by inventing more words - by moving the goalposts, so to speak - and so generating new hidden beetles.

If we disagree about some 'matter of fact' (such as the existence of a table in a philosophy classroom), we do not resolve this by investigating each other's perceptual apparatus. We resolve it by trying to find out whether we can make such a disagreement intelligible - whether we can play a language game around 'this table does not exist', for instance. By doing this, we find out interesting things about the words and concepts, and about the limits of intelligible conversation. By discovering how much we have to throw away if we don't agree about the table, we also discover something about the universality of basic physical concepts. It's very hard to talk about anything without them ...

It is possible that scepticism about the reliability of our agreements about our immediate physical environment strictly entails a scepticism about our ability to talk to one another. (One route to such an argument might be through the fact that agreement about the physical coding of our language seems to be a necessary preliminary to agreeing about the content of what we say, but I'm not sure about this.) If this entailment was reliable, then our agreements about our immediate physical environment - our agreements about our empirical perceptions and intuitions - would be as reliable as our ability to talk to one another. (They would be among Barry Stroud's 'privileged statements', whose importance he seemed determined to ignore...)

What is clear is that it is, in practice, very hard to have a conversation with someone who seems to see an entirely different world from the one that we do. A table sceptic is hard to take seriously as an interlocutor (as we discover in our philosophy experiments ...)

I have been arguing, in this blog, that we don't need - that it is incoherent to ask for - a metaphysical underpinning to our scientific theorising; that 'privileged statements' about the possibility of language and their consequences provide a completely adequate foundation for this.

Meaning is internal to our linguistic practices, and as a consequence of this so is any world that we can describe to one another. Our experience of a private 'conscious' world outside the scope of articulation, of public theorising - our conviction that the 'real beetles' have some role in establishing the validity of that theorising - is misleading here; is, in fact, quite false.

In order to make sense of  'real necessity,' we must be able to give some kind of account of it - and to agree about the meaning of that account. Its nature is entirely captured in our shared account. Even if we could project it onto the world in an undeniable way - if we could show that any describable world must exhibit real necessity (as we accounted for it), for instance - we would be showing that the existence of real necessities was a consequence of our capacity to theorise, and not a justification of it.

If we want to say 'X must be true of the world', we need to know what 'X' means. We discover this meaning through our experiments with the way we use 'X'. When we 'discover' that X is true of the world, we cannot distinguish this from discovering that 'X is true of the world' is, in the most general way, intelligible. It's meaning is exhausted by its role in our linguistic interactions, and our tests of its truth depend upon our capacity to agree about how to talk about it.

It is an undeniable part of our human experience that there seems to be more to it than this, but this 'more' is both outside the scope of our public theorising and irrelevant to its validity.