Barry Stroud's 'Privileged Statements'

In his celebrated paper Transcendental Arguments' (Journal of Philosophy, Vol. 65, No 9 - May 1968), Barry Stroud talks about 'privileged statements' - statements which must be true in any language. These are exactly the statements I have drawn attention to as roots of recursive accounts (e.g. of concepts such as truth, accounts of which otherwise generate open question paradoxes).

Denial of a 'privileged statement' generates (as I've said elsewhere in this blog) a 'Moorean' paradox - a statement which undermines the possibility of its own intelligibility.

Curiously, Stroud doesn't seem to see the relevance of these privileged statements to accounts of necessity, reliable knowledge etc., even in much later writing (See 'Engagement and Metaphysical Dissatisfaction', Oxford, 2011). He seems to want to insist on a distinction between what we can say about the world and the way the world really is. A question than cannot arise in a playable language game, such as 'are we really able to talk to one another?' might still, in some sense, be an unsettled matter in the real world.

He says (in the original article):

“…from the fact that a proposition is a member of the privileged class it does not follow that it is a necessary truth, and so it seems that there are some propositions, such as ‘There is some language’, the truth of which is necessary for anyone’s ever asserting or denying anything, but which are not necessary truths. It could have been, and undoubtedly was, the case at one time that there was no language, and it probably will be again. Although it could not be truly denied, still it might have been, and might yet become false.”

There are lots of possible objections to this - the main one being that we do philosophy by talking to one another, and the idea of a philosophical position that cannot be intelligibly articulated is fairly odd. The temporal context issue is irrelevant, of course (I commented on this before), because we can talk about it now. 'Necessary' is a word we know how to use - it works in our language. If we can only use the word because there is something in the world that gives it sense, then we must assume that that something is there, because the sense is there. I'm inclined to think this supposed substrate causes more confusion than it addresses, of course.

The point is that the tail-chasing we get into trying to give 'necessity' a metaphysical ground arises just as much with 'metaphysical necessity'. What ground should we give to this?

Anyhow, I had a correspondence with Professor Stroud about this, and I don't think we understand each other any better. Well, I don't understand him, at any rate.  (!)

How is it possible to make a conception of metaphysical necessity, independent of 'privilege', philosophically intelligible? Stroud would need us to be able to say that a philosophical position that cannot be put into words might still be intelligible. That while no argument could every count against our being able to talk to one another, it might, nevertheless, not be necessary that we are able to talk to one another. What possibility is he pointing to here?

He might say: it is possible that the Universe might not have contained talkers. But this Universe needs to be describable - we need to be able to say of it 'it does not contain talkers' - in order for this example to make sense. In order to figure in a philosophical example, a something - including a universe - needs to be minimally articulable. Its intelligibility depends upon our making it intelligible here and now.

There is a kind of problem here, though - which Stroud describes as a 'metaphysical dissatisfaction' - an uneasiness about the link between what we are constrained to say (if we are to agree that we can say anything at all) and our own phenomenological condition, our visceral interaction with a world which is not, on the face of it, much constrained by words at all. This dissatisfaction cannot be articulated, of course - I might even ask why we call it dissatisfaction, except that I feel it connects with something sympathetic, something in us which wants to have something 'beyond language' that has, nevertheless, language-like powers of demonstration and calculation.

Open Meaning

In earlier posts (e.g. 'More Meaning' I've argued, perhaps sketchily, that there is an irreducible openness about meaning - that for a statement/expression/word to be meaningful, there must be something about it that is 'still to discover'; that once the use or a term is entirely captured in explicit rules, its meaningfulness, far from being elucidated, actually disappears.

This may seem a perverse position to take, and contrary to some of the main currents in the analytic tradition. However, that tradition has run up against intransigent confusions with the respect to what we might take to be 'the meaning' of a linguistic expression - Moore's paradox of analysis being one, and the persistent ambiguity of fundamental categories such as reference being another.

Yesterday, it occurred to me that the open question paradox associated with meaning really makes all of this inevitable. I've had a look through some earlier posts, and I think this idea is in them, but it's worth spelling it out:

When I explicitly ask you what you mean, I am generally looking for a contextual elucidation, not a complete account. That we can never give a complete account is a consequence of the systematic ambiguity of 'mean'. In other words, not only is the meaning of  'mean' undecidable or incomplete, this ambiguity is inherited by its uses - when I ask you what you mean by an expression, you can legitimately ask me what I mean by 'mean'.

We negotiate these things tacitly, and either arrive at a satisfactory way to go on, or silently leave the arena of enquiry.

Empiricism and Analyticity

While 'analytic' truths might seem to be independent of any facts about the world, the possibility of an analytic engine is an empirical matter. That some machine can be made which does analytic computations is a fact about the world - that it contains this possibility.

Arguably we, either individually or communally (I'm thinking of a language community here), are just such analytic engines. I say 'arguably' since anyone who has followed the general trend of my arguments so far will see that I think a claim to the contrary would turn out to be unintelligible - any argument for such a contrary claim would rely upon the very analytic 'machinations' which it would want to render questionable. If I can't produce reliable arguments, then I can't reliably argue that position either.

Does this tell us something about analyticity itself? I'm not sure. I have a moderately (though not completely) Quinean view on this - what we count as conceptual seems to be at least partly contextual. There is this, though:

To say that 'we' are an analytical engine is, I think, to say that we can produce reliable arguments for something being true on the grounds that some other things are true. We can deduce one truth from another, and we can show that this deduction at least makes sense, and, in some cases, is unavoidable. Scepticism about this must entail a scepticism about sceptical arguments as well, and so about its own intelligibility.

Taken this way, taking the possibility of argument (entailed by the possibility of language use), as the context, the possibility of analytical argument depends on the possibility of the engine - an empirical, not a logical (or analytic) possibility.

Can we imagine the world being otherwise? Not if we imagine an orderly world which we can talk about - i.e. one which we can both imagine and describe. A world with definite characteristics is a world whose parts are related in a grammatical way, so to speak. And if they are related in such a way, they comprise, to a minimal degree, an analytical engine.

That fact that we can build computers as well as produce arguments tells us this about the world - that it's character can be rendered grammatically, and so that it can be reorganised into a grammatical computation and prediction engine of sorts. We are such an engine.

Can we imagine a world without us in it? Of course, to an extent. But not one that we aren't imagining - that we haven't brought into our linguistic context. (How does the sentence 'I am presently imagining a world that cannot be described' work? And is 'a world that cannot be described' a description? - OQ paradoxes again ...)

Positivism

Rudolph Carnap, and others, believed that a question that could not  be answered empirically or analytically must be a nonsensical question. The realisation that this condition of meaning was, by it's own account, meaningless, pushed 'positivist' approaches to verification and meaning out of the limelight.

The question whether there is any condition of meaningfulness that could avoid this trap has been explored to some extent. In general, however, the statement M: 'A statement is meaningful if it satisfies criterion X', can only (by its own account) be meaningful if it satisfies criterion X. This can only be unequivocally established if we know what 'X' means in M, and for X to mean anything at all, M must be meaningful. As I've argued in other posts, accounts of  meaning will always run up against open question paradoxes.

However, one positivist insight that should not be lost is this: Although a general closed account of what it is for a statement to be meaningful will always turn out to be wrong or question-begging, systematic scepticism about meaningfulness is also incoherent. A sceptic cannot claim both (a) that a question is meaningful and (b) that no conceivable approach to answering it can succeed.

I think Austin was onto something like this with 'real'. If we want to be epistemological realists, we must have some conception of what this means - and this conception must be relevant in some way to how we arrived at, and might defend, our metaphysics.

A question which cuts itself completely adrift (consequentially speaking) from the rest of our conversation is not part of that conversation. It is OK to ask questions that are hard to answer ('Is there life elsewhere in the galaxy?), or even questions which we think may never be possible to answer. But a question which is couched in such a way that any answer is necessarily going to be inadequate (as philosophically sceptical questions are) must be suspect. Nothing we say can have any consequences for the nature of the answer, and so nothing can be consequent upon the answer.

If I ask 'is there a reality beyond what I can perceive?' in such a way as to systematically block any approach to giving an answer, then the answer can have no consequences.

If it has no consequences, can it have meaning? Can I claim that a statement which commits me to nothing has a meaning?

Since it could have no determinate role in a conversation, I doubt if I could.