Private Languages - A Summary

There are only two possibilities for a private language:

(1) That it is translatable into a public language, or

(2) That it is not.

If (1), then the language is not private in any relevant sense. Even if we do not require the translation schema to be spelled out, the 'owner' of the language must agree about the public translations.

If (2), we have no basis for calling the 'private language' a language at all. We identify languages on the basis of inter-translatability (Davidson, Quine, radical translation).

Whether or not this is a helpful restatement of Wittgenstein's exact position, it seems to be both conclusive and fundamental.

Necessity

My existence is necessary to the intelligibility of my contingency.

(Something Descartes didn't say ....)

Preconditions of Conception

The world does not have to have 'sameness' in it, far less something that 'sameness' is the same as, for 'sameness' to work, grammatically.

It is only necessary that the world embodies the preconditions of the grammatical category of sameness (if we wish to talk of a world at all - and I think we should). We need not even be able to fully articulate these preconditions, although if talk of the world is to be useful we must be able, find it useful, to articulate some of them.

When I talk (meaningfully, or I am not talking) of something's being the 'same' as something else, I am depending on this grammatical category - and am therefore committed to the world's being the kind of place that will sustain the category, and my use of it. I would be projecting the conditions of the use of sameness on to the world, not finding them in it.

It might be, of course, that to attribute anything at all to the world we must be able to attribute 'sameness' to 'sets' of its 'contents' - it may be such a fundamental grammatical concept that we can't do without it in our attributions. This still doesn't mean that we have found 'sameness' (or 'sets' or 'contents') in the world in a way which accounts for - validates, if you like - the concept.

I can draw conclusions about the world (that it is constructed in such a way as to allow me to use the concept 'same') from the fact that I can use the concept 'same'. I cannot demonstrate the validity of the use of 'same' from the contents of the world - this would be circular. Particularly if 'same' has the character of being necessary for quality/content attribution in general.

If I reject the validity of the concept 'same', then I no longer need to attribute these supporting conditions to the world. I also lose a very useful concept - possibly even an essential one, if we want to say that we talk to one another. This is because it is hard to think of how we would articulate this idea without some minimal 'sameness of meaning' construction: what I believe I say is (in some minimal sense) the 'same' as what you understand me to say.

If I found I could not render my sense of  'same' intelligible to someone else, then no amount of pointing at the world would enable me to resolve this.

I may become personally convinced that those with whom I share the concept see the 'same' world as I do - that they are comfortable with what our use of  'same' commits us to saying about it. The strength of this conviction is a product of the shared use, not the grounds of it. If the shared use breaks down, if we become unintelligible to one another in this respect, then the conviction that we share a world breaks down with it - unless we attribute error or dishonesty to them, although in what language might we make this attribution? (What world do I find their error or dishonesty in?)

Rules, Meaning and Machines

I did a search of this blog just now, looking for 'machine', and find that I've posted a lot more than I thought on the topic. I want to shine a different kind of light on some of the machine stuff by making a suggestion. Or perhaps focusing on it, if I've made it before...

The suggestion is that these two 'facts' are actually equivalent:

(A) We can talk about the world in the way that we do.
(B) We can construct machines.

If it isn't immediately obvious that these are equivalent, think about the following:

(1) It is always possible to re-state a 'natural' law as a kind of grammatical principle - an instruction about how to talk. If we talk in a way which radically ignores a 'true' natural law, we risk becoming unintelligible. By 'always' I mean that it is not possible to articulate a natural law which cannot be re-stated as a grammatical principle.

I am going to suggest that the 'grammatical principle' view of natural laws is the view that is the most epistemologically relevant.

(What I mean by a 'grammatical principle' here is something Wittgensteinian. Something like a 'rule about how to talk', taken very generally.)

(2) Meaning and prediction are strongly linked. Knowing what someone means allows us to make predictions about their behaviour (whether represented semantically or physically). I don't mean (obviously) that we can exactly predict their behaviour, just that if we couldn't draw some conclusions about it from what they appeared to be saying, then we couldn't be said to understand what they actually were saying.

(3) When we make predictions based on 'natural laws' we can be thought of as making predictions based on 'grammatical' rules. We are saying something like: 'If we understand each other now, then we also share some expectations'. If we deny the expectations, we undermine our present understanding. We discover that we were not as intelligible to one another as we thought.

Remember that there is no scientifically or philosophically relevant test which can distinguish between (i) the world having certain characteristics and (ii) our being able to talk as though it has those characteristics (both taken completely literally and generally).

It seems unexceptionable to say that physical laws permit the construction of machines.  What the above outline indicates is that we must conclude from this that grammatical laws (of the right kind) entail the possibility of constructing machines.  When we say 'if you want Y to happen, you should do X', we may as well be saying 'If your behaviour falls under description "X", then the statement "Y" will come to be true about the world'.

These hypotheticals can be enormously complicated, of course - we can construct very complicated machines. But encapsulated in the statements of these hypotheticals are the conceptual machines that we expect our 'real' machines to instantiate. If we are forced to concede that the 'real' machine does not behave according to the blueprint, then the hypothetical is, with qualification, falsified. We might experiment with the bounds of intelligibility here, but only to repair them - not to abandon them.

This process of law discovery - discovery of what it is intelligible to say about the world - and successful machine construction has left us with the 'machine world' metaphor. The metaphor is reinforced by the related 'discovery' that any intelligible description of the world gives it machine-like characteristics. It's hard to step back and see that it is the intelligibility that carries the machine metaphor into 'the world'.

No matter how successful our project of rendering parts of the world intelligible becomes, it gives us no grounds whatsoever for believing that we can render it completely intelligible. This is a corollary of the Goodman/Kripke paradox. The Open Question paradoxes generated by the hypothesis that the world is completely intelligible tell us that not only do we not have grounds for believing that it is, but that there is a sense in which it cannot be.

It's worth saying, here, that an intelligible 'mystery' implies its own sets of mechanisms, so this perspective creates no space for 'supernatural' processes.

It also doesn't permit contradiction or radical incongruence - the scope of intelligible theory does not include incoherence. (By definition).

The boundaries of intelligible theorising cannot be directly drawn, either. We may be able to show that there are incomprehensible numbers, but we cannot do arithmetic with them.

What it does mean, though, is that our effective judgements about these things are, in a sense, parochial. 'Intelligible' means something like this - 'can, in principle, be made intelligible to us now'. Some of our descendants may participate in interactions that we would find incomprehensible, and be able to do things that would seem magical to us. We might be tempted to say 'they speak a language we cannot understand', but this would only make sense if we had some hope of learning it. If we didn't, we'd have no idea what they were doing, or even whether they were properly human.

Any useful conception we might form of a machine world has this same parochial quality. This means that it cannot be 'absolutely' generalisable (in any meaningful way).

To suppose that the 'whole' world can be rendered intelligible in terms of law-like characteristics is to suppose that our intelligible rendering depends on these same characteristics that we are investigating. And this problem is metaphysically agnostic - no additional substances or processes can dissolve it.

But our parochial concerns do have some general consequences.

One is that however our language practices might change, we will never talk in a way which renders the way we presently talk completely unintelligible. It must always be at least a limiting case. This isn't because what we say now somehow binds the future - it's just that we can only count something as a language if it can be rendered intelligible in the language within which we make the judgement: the language we are speaking now.

For this to be the case, we do not need to give some account of  'the language we are using now' much beyond pointing to it. If we want to talk about talking, we are pretty much committed to some conception of what we are talking with.

A slightly more disturbing consequence of this approach is to do with the relationship between conceptual machines and what we might call 'real' machines - the machines we build or find in the world. It's easy to see that the conceptual machine will always provide the normative standard against which the performance of the 'real' machine can be judged. This can't work the other way around - the 'real' mechanism, the constructed mechanism, cannot work as a normative standard. The Kripke/Goodman paradox makes this formally impossible, and the possibility of mechanical failures make it impractical.

This has an obvious consequence - that if we think of a person as a kind of biological machine, we cannot also think of them as providing a reliable normative standard of language use - of meaning. On the other hand, we must attribute something like this to someone we wish to count as a competent interlocutor.

A kind of way around this comes via the recognition that being a competent interlocutor includes being open to correction - being able to learn, when one has made a mistake. The normative standard is provided by the whole language community, in some sense - it is within the shared language that the practical issue of normative standards is resolved. And this doesn't give rise to any of the formal problems hinted at above, because the shared language is a context of adjudication, rather than a machine whose performance we might adjudicate on. We can think of a person, an interlocutor, as a kind of correctable machine; but we cannot think of the whole of the linguistic community in this way.

Dead Versions

In early posts, I provided links to what I thought might be a more organised presentation of the less speculative bits of my explorations.

These links no longer work, and I don't feel inclined repair them - I have the documents if anyone is interested in the history.

Since my friend Bob Tait died - Bob and I used to discuss, at some length, the kinds of issues I have raised here over the last few years - I have put together some notes to summarise our discussions. Perhaps, more correctly, to summarise them from my point of view ...

Making these notes has given me a framework that I find more helpful than the 'versions' framework, which was a very preliminary attempt at a book-length exposition. Each note introduces a single idea, and each is a spotlight on the whole. I've kept them brief and not bothered too much with explicit references - Google any external material you want to know more about, or drop me a note and I'll fill in the gaps.

Anyhow - the results are here. I'll update them periodically and try to remember to post note on this blog when I do.

Roger Scrutton

I heard Roger Scruton getting quite an extended interview on the ‘Today’ programme this morning. He was being given time to explain, at some length, how right-wing commentators were being misrepresented and silenced. A theme of his complaint was how his views were sometimes misunderstood because people didn’t attend closely to the meaning of what he was saying.

An example he used was a comment he had once made to the effect that ‘there was no such thing as date rape’. In the interview, he explained that what he meant, in the context of the reported discussion, was that there was no such crime as date rape, that it didn’t appear in the statutes. He then went on to say that we had to distinguish between cases where someone was attacked and raped and cases where a night out had ‘gone a bit wrong’ (possibly not an exact quote, but not materially misleading).

Words only have definite meanings if we agree (at least tacitly) about what these meanings are – in effect, about how these words are used. No real conversation starts by defining terms – we can only converse if we already share a great many semantic ‘hinges’. When we depend on one of these, but find ourselves lost in ambiguity or disagreement, it is not helpful to resort to definitions. This simple enlarges the arena of contention.

What we need to do, what we actually do in ‘real’ conversations, is stop, look at each other, and wonder ‘Well, what do you mean by ….?’

It is a technique of Scruton and other traditionalist commentators (Peterson, certainly) to toss ‘definition’ bombs into a conversation in order to derail it. “Aha, but you see what I mean by an ‘atheist’ is …”.

This is a kind of semantic authoritarianism, and it is often justified as a response to ‘relativism’. This dilemma – having to choose between stipulation and incoherence – is exactly the dilemma posed by open question paradoxes.

The ‘What do you mean by… ?’ question reveals its depth here. We may ask it explicitly (the case philosophers are inclined to focus on) or we may make tacit explorations – trying, by experiment, to discover a way of talking that allows the mutual conversation to continue. The conversation we are having now is the recursive root of the whole edifice of intelligibility. When we give it up – either by authoritarian stipulation or by reducing it to tribal signalling – we consign the whole to the flames.

Scruton can only have a conversation with someone about date rape if they are both talking about the same thing. Resorting to definitions and elisions may be a fun game he can report to his mates in the pub later, but that’s not really the relevant context …

Conversational trajectory ...

I think that all conversations end up on one of three trajectories:

• Towards confusion and disintegration
• Towards dogma, recitation, ritual
• Towards discovery, insight

The third is the only one that allows the conversation to continue. The first two either lose or eliminate the possibility of semantic content. It isn't possible to continue a confused conversation, and a dogmatic or ritualistic exchange cannot develop. It is possible that the first two options conflate, depending on how they are considered.

When we reject a particular way of articulating the truth (dogma), we do not relinquish objectivity. Objective truth can only be explored in a working language, however much we may wish to find a 'terminal formulation'. What we count as a working language cannot be captured in a specific formula either, but only reveals itself in our practice, in our actual conversations with one another.

Within these conversations, there are certain moves that would be terminally confusing. These moves, however, can only be characterised semantically - there is no 'extra-linguistic' formula for distinguishing them. This is a direct consequence of the indeterminacy of translation - it is only when we know what someone means that we can adjudicate on the sense (intelligibility) of what they are saying.

A Moore's paradoxical statement is only 'paradoxical' if we insist on decoding its representation canonically. And, of course, the insistence on canonical interpretation is also 'paradoxical', for the same reasons. If I insist that the classical example of a Moore's statement uses 'believe', for instance, in the 'normal' way, then my insistence is also a Moore's paradox.

Writing

A difficulty with writing is that my interlocutor must be imaginary. My apologies if you are reading this, and feel quite real.

There is a problem with talking to ourselves. Well, there are a pile of problems, really, not least being the private language argument.

One somewhat phenomenological aspect of this occurred to me this morning. When I ‘talk to myself’ it is generally in fragmentary phrases. The more anxious I feel, the more these resolve into definite sentences and short paragraphs, as though I am preparing a more thorough defence.

And yet, even while I do this, I am losing touch with the inner part of myself and paying attention to how I will seem to others. I am leaving my principal source of authenticity and insight behind and paying attention to coping strategies, to verbal ping pong, to ritual combat …

Our model of cognitive capacity as comprising a ‘rational’ functional element – an ‘executive process’ – is quite dangerous here. I think Jerry Fodor has most clearly fallen for the ‘internal monologue’ picture of rationality. It is a picture which is catastrophically self-referential: ‘I know that what I think makes sense because it looks like what I would say if I was making a good case. And I know this because that is how I report my thoughts to others’ … which is no more than to say that I know how to talk. How to be intelligible. Accounting for this in terms of some kind of ‘internal intelligibility engine’ is almost ridiculously circular.

The public-facing imagination is where we are most likely to find what Rogers and others have called ‘introjected values’ – the cut-and-paste patchwork of psychological grafts and donated thought processes that pinch and itch when we try to reconcile them with our lived experience. Values which not only do not come from ourselves, but which we are almost bodily aware of as inconsistent with any possibility of authenticity.

The more explicit and articulate my prepared inner monologue is, the more likely it is to be infected – to be a rationalisation built out of introjections.

In the public sphere, how do we distinguish between writing as case-making and writing as authentic exploration? Between self-justification and self-discovery?

Why am I writing this here?

Determinism, Prediction

The metaphor of universal determinism is based on our ability to predict and/or control the behaviour of some parts of the universe. We presume, from our knowledge of this ability, that the whole behaves like these parts.

I've hinted at why this must be wrong before, and also at some problems with the perspective that I've tried to take. What I want to do here is to draw attention to the the central incoherence of the metaphor.

First, it is a metaphor which is impossible, in principle, to completely unpack. It doesn't matter how much computing power we have, the machinery which implements it will be part of the universe we are trying to predict, and so would be subject to the halting problem (the open question paradox as it appears in algorithmic processing). The machine could not predict relevant aspects of its own behaviour.

Secondly, our predictive theories are captured in a semantic net which cannot, itself, be completely analysed. This is because this analysis must depend on semantic assumptions. Even the law of non-contradiction can only be stated in words.

I've mentioned the necessity of attributing agency to interlocutors, and the problems this creates for universal determinism. There is more to the semantic dimension than this, though. The Kripke/Goodman paradox shows us that there will always be competing semantic interpretations of any finite set of mechanical signals. And yet it is only mechanical processes (and signals) that a universal determinism can predict. No determinism can, therefore, predict the correct interpretation of, for example, a string of characters. This means that no determinism can absolutely establish the correct interpretation of any string of characters which might be taken to be a statement of universal determinism. There is no determinism which allows us to give an unequivocal semantic account of itself.

In some respects, this might be bundled in with other catastrophic implications of the Kripke/Goodman paradox - one of which seems to be that we cannot give an unequivocal account of anything at all. The worst of these implications can, however, be resolved within a recursive account of truth-telling, and a proper understanding of the irreducibly semantic character of whatever conversation we are having now.

This kind of solution is very unfriendly to the deterministic metaphor, however. It puts semantic, rather than mechanical, characteristics at the roots of our explanations and predictions. Also, because our conversations must take place 'in the world' the idea that there is a perspective from which a universal determinism could be demonstrated is incoherent. Not only is there no god-like perspective, but no being with such a perspective could have a private language any more than the rest of us. And certainly not one we could ever learn to speak...

The incoherence of the mechanical metaphor doesn't, at first, seem particularly relevant to the problem of logical determinism, which arises from the fact that the truth-value of a properly contextualised statement about the world - even if it is about the future - cannot vary. If it turns out to be the case that I live into my eighties, then our conventional understanding of the logic requires that this was always the case, but that we just didn't know until the relevant bits of the future had been properly examined. Letting the truth-values of properly contextualised statements vary breaches the rule of non-contradiction, which strikes at the very heart of intelligible semantic processing.

This seems to be a grammatical matter, however, and possibly less consequential for the attribution of agency, since that seems to be as necessary to semantic processing as the invariability of truth values.

The problems with the mechanical metaphor may have this relevance, mind you: that there are factual statements about the future whose truth values cannot in principle be determined from current facts. This gives the 'truth value' a very abstract flavour. It also means that the logical grammar will never undermine the 'use-meaning' of these statements, or of anything intelligible we might care to say about how we attribute truth-values to them, or about what we mean when we do. Preserving the rule of non-contradiction here can have no awkward practical consequences, and any time it seemed threatened re-contextualisation would remove the threat.

Choice

If I have a theory about why people choose things – if I can predict what they will choose – can I still call it choice?

Is their perception of this relevant? If it is, wouldn’t I say ‘They believe they are choosing, but really they are obeying a law of nature’. Can I claim both that they are choosing and that they are rule-bound?

There may be a long argument to be had here, but I don’t think we can. We do not attribute ‘choice’ to inanimate objects – even to complex objects such as machines or computers – if we have a mechanical account of their behaviour and interactions.

We attribute choice when it is convenient to attribute intentional states, and we must attribute choice when we must attribute intentional states - e.g. to interlocutors. This bounds any theory of choice, of course, since all our theories are grounded (ultimately) in the possibility of theorising – the possibility of language. Any attempt to ‘ground’ a theory of choice would find that it had to presume the possibility of choice in order to make theorising intelligible (since it is something that we do with interlocutors).

This is a slightly more circuitous version of the OQ paradox generated by the concept of ‘rational choice’ directly – that we cannot give an intelligible account of how we might rationally choose a method of making rational choices.

It has a novel coda with respect to the tired debate about mechanical determinism, however: The language in which our mechanical theories are constructed is only intelligible if we attribute the capacity to choose to our interlocutors.

Wittgenstein

Wittgenstein approached, and circled, the question of the language within which we do philosophy, but did not address it directly. He didn’t see what conclusions should be drawn from such an examination.

 

I think about this often. I’m not a W scholar, but have read a good deal of him, and about him. I don’t know whether he made a mistake, or whether I have. It is possible that, despite his therapeutic recommendations, he was simply unable to face what an enquiry in this direction might produce. Academics, and particularly academic philosophers, are deeply committed to a certain view of the generality of their disciplines and their methods.

 

He spoke a good deal about mysticism, and mentioned ‘that of which we cannot speak’, but avoided the boundary between this and what we actually say. And yet it is exactly the exploration of this hinterland, the discovery and rediscovery of its rules and conflicts, that lends meaning to the rest – that shows us, by discovery and demonstration, that we cannot give a syntactical or a ‘correspondence’ account of truth and meaning. We can only discover these things in the course of our conversational experiments.

 

The central limitation that philosophers find hard to accept is that there cannot be a theory about how to do philosophy properly. Their core activity cannot be rendered intelligible by explicating its validity rules.

 

And yet the desire for intelligibility – in ourselves, in our interlocutors, and in the world – seems like the whole ground of rationality. If this is not an intelligible desire, then what is?

 

We have to attribute minimal intelligibility to any potential interlocutor, of course. And it is obvious that to be an interlocutor one must be intelligible. But there does not seem to be a categorical imperative to be an interlocutor? Maybe its easier to think ‘whereof we cannot speak’ than ‘to whomsoever we cannot speak’. This is the picket fence that bounds the jurisdiction of argument, beyond which the dragons roam.

Bob Tait

Just before Christmas, a good friend of mine died unexpectedly. I knew Bob slightly for quite a number of years, but from about 2011 we had been meeting regularly - weekly - to talk about some of the things I ponder in this blog, and about how they related to his interest in wider cultural structures which are captured by, and in which are embedded, all out linguistic behaviour.

Anyone with peculiar interests will know the joy and relief associated with discovering someone else similarly afflicted. I will miss Bob very much.

Bob and I both wondered, in different ways, whether the biggest threat facing humanity might be a kind of semantic catastrophe. He felt that we underestimate the confusion caused by our inclination to represent large scale cultural constructs as intelligibly intentional, and I could see our capacity for meaningful exchange being overwhelmed in a technological environment in which semantically moderated structures were less 'successful' than complex control mechanisms that depended only on signal propagation.

And, of course, he tried hard to make sense of cultural constructs, while I pursued semantic fundamentalism. So there.

Phenomenology - what it's like to be a language user ...

Traditional sensory empiricism is incoherent for a lot of reasons, but here's a particularly acute one:

If I have an empirical belief that I am particularly sure about, because I have had some experience which confirms it, the relationship between the belief and the experience must be of a certain kind - it isn't enough that the experience generates the belief. The way the experience generates the belief has to be convincing - it has to conform to our intuitions about the relationship between perceptions and what we conclude from them.

For instance: If I thought there was an ostrich in the room, and based my belief on certain sensory impressions, these impressions would have to have a certain kind of relationship with the presence of ostriches.  If I tried to describe the impressions to you, they might sound like the impressions of something a bit large, fluffy, and bad-tempered. They would be ostrich-like.

If I was wired up badly, so that when my foot was tickled I thought there was an ostrich present, I might be just as convinced as someone who had ostrich like sense impressions, but I would be much less convincing. I might say "I have an ostrich presence foot tickle".

Either that, or I would actually have ostrich like sense impressions when my foot was tickled.  Instead of wriggling and laughing, I would say 'I see something large, fluffy and angry'. I might discover, or be shown, that this had to do with my foot being tickled, but this would not make the sensation tickly - it would remain ostrich-like.

Now, suppose that, in the real world, I was hooked up to a machine in such a way that the presence of an ostrich would result in the ticklish sensation in my foot, but that I did not know this.  Now I have a sensory impression which is a reliable indication of the presence of ostriches, but has nothing ostrich-like about it. Even if the relationship between the tickle and the presence of ostriches was explained to me, I do not think I would be inclined to say that ostriches were tickly. At least not if I wanted to make any sense to someone who doubted my ostrich awareness capacities.

And what if we close the loop? What if I am hooked up to such a machine, but am also in the condition of having ostrich-like sensations when my foot is tickled? An interlocutor unaware of my physical circumstances (speaking to me on the phone, for instance) might find my ostrich description quite convincing, and be, correctly, persuaded that there was an ostrich in the room with me.

And this final case is, perhaps, the most realistic. We have no direct access to each others' perceptions, or to the machinery by which they are produced. We take for granted that this machinery is reliable because it makes us talk in ways which are intelligible - not because we either understand how it works or, further, whether that understanding could ever lead to epistemological insight.

Mechanism and Matrix - a Summary

I'm going to try to succinctly explain why we can never have a unified physical theory that (even in outline) 'explains' consciousness, intentionality etc. I have argued at length that we must attribute these to interlocutors (so we can't do without them), and have also given some slightly meandering accounts of how they cannot 'emerge' from a mechanistic world view (or any pseudo-mechanistic view, involving 'spirit' or 'mental substance', which does not surreptitiously presuppose them).

A mechanistic theory of the world must be articulable. Even if we can conceive of one which is not (I don't think we can, but am prepared to be, in a sense, agnostic) such a theory could have no place in science or philosophy, since both require theorising that can be rendered in language.

To be a unifying theory, of the appropriately fundamental kind, such a theory would have (at least in outline) to point to an explanation of how our linguistic mechanisms - the sounds and 'signals' that make up our exchanges - would work. Some hierarchy from physics, through chemistry, to biology, neurology and social science might be imagined.

My argument will be that no such hierarchy can be imagined, whatever assistance is imported by way of 'emergence' or complexity.

And it turns on a very simple thought: the mechanistic theory (however complex) would have to be true. This means that it would have to incorporate a method of implementing the discrimination of one sub-set of the noises, squiggles, clacks, and semaphores which are linguistic behaviour in a way which was isomorphic with our discrimination of true assertions. This is the open question paradox as applied to the matrix world: either our capacity for truth-telling is outside the world, in which case the matrix doesn't contain the whole of our world; or our capacity for truth-telling is part of the 'matrix illusion' - in which case we cannot hope to give a 'true' account of the mechanism which underpins this illusion.

So that's it.

Mathematics

When we talk about mathematics, we are talking about what Wittgenstein might call a 'grammar'.

I want to say something a bit more specific, but consistent with Wittgenstein's usage, about grammars.

If we find that a statement - of whatever kind - threatens the intelligibility (as opposed to the veracity) of a conversation, then we are dealing with an element of the grammar of that conversation. Clearly, logics and literal (linguistic) grammars largely fall within this characterisation. However, so do some more peculiar things - Moore's paradox and its variants being one.

Under some circumstances, empirical statements can have a grammatical quality. If you and I appear to agree about almost all the contents of a room we share, but you insist on adding a pink elephant to the list, I will find the intelligibility of our whole discourse threatened, not just the obviously difficult bit. I realise that this challenges some principles of empiricist epistemology, but that seems more conducive to insight than to confusion, as far as I am concerned.

Sometimes we find that groups of these grammatical statements can be generalised over - we can articulate grammatical rules and, even, grammatical theories. (These, of course, cannot be compounded into a theory of intelligibility in general, because this would generate an OQ paradox.)

Some of these theories are like literal (linguistic) grammars, in so far as they do not adjudicate on content. Others seem to include explicit or tacit statements about the world. (Kant, but without the metaphysics).

Mathematics falls into this latter group. It is a set of instructions about how to speak, on the one hand, but it allows us to make reliable measurements of, and predictions about, the physical universe.

As I have argued elsewhere, this is not an accident - the fact that we can talk about the world in the way that we do (an apparently slightly weaker epistemological claim than that it is the way we talk about it - although the stronger claim may not be intelligibly distinguishable from the weaker one), means that the world must have some 'machine-like' qualities that permit this reliable measurement and prediction. Another way of putting this is to say that if  our predictions were radically unreliable, our  'language' would be unintelligible - we would have no language.

There are, in short, some grammars that tell us about the way we can speak, and about the way the world is, as though these were the same thing. And, of course, they are. Any world we can have a conversation about must conform to the grammar of that conversation.

We cannot bring a literally indescribable world into our conversations. We may try to imagine one outside the net of words - perhaps in some private phenomenological space. But even to say we are doing this is to describe it to the extent required for it to be theoretically relevant.

(It is important that to bring apparently inarticulable - wordless - experiences into a space where they can be seen and empathised with is an important therapeutic process.)

Since we can never have a 'complete' grammatical theory, there must be an element of experiment as well as calculation in the way we progress. The discovery that some radical move - zero, negative numbers, complex numbers - does not result in chaos is  characteristic of mathematics. As is the discovery that some 'obviously true' principles - e.g. mathematical induction - can be resistant to direct proof. Gödel wanted to slip a theory of forms in here, I would prefer to say that for our language to be presently intelligible, some future discoveries must come out the right way ...

Hilbert, of course, believed that generally intelligibility would turn out to depend on mathematics being formalisable - and he turned out, not surprisingly, perhaps - to be exactly wrong. It is because it is not completely formalisable that it is mathematics. We cannot predict how our language will develop, except to say that the future must preserve - perhaps as a special case - the intelligibility of how we speak now. (To say otherwise is to claim that the way we speak now is potentially unintelligible, which is incoherent.)

We will discover new ways of talking. We will discover what rules we must follow to remain intelligible.

Matrix Metaphysics and 'Machine' Worlds

If you lived in a 'Matrix World', and another person in that world said to you: "Do you know that we live in a Matrix World"?, how would you be able to work out what they meant?

If the world you appeared to be in was a perfect simulacrum of a "Real" world, there would, ipso facto, be no test that could reveal this. If 'We live in a Matrix World' and 'We do not live in a Matrix World' cannot be distinguished on the basis of some kind of validity criteria, then they also cannot be shown to be semantically distinct. We resist this idea in fictional narrative of course, because there always is a semantic distinction - even if it rests on criteria external to the narrative.

(This may seem to have a 'positivist' flavour, but I am not making any statement about the kind of validity criteria that would be relevant - I am just saying that there should be some.)

In the film, of course, and in other science fiction explorations of this theme, there is a 'way out' of the Matrix World. There is a way to test the simulacrum. But there is no way to test whether the proposed test is a cunning device of the simulacrum, whether the red pill simply triggers another set of 'reality' programs, designed to deal with certain kinds of inconvenient metaphysical speculation.

There are plenty of cases, of course, of metaphysical diseases turning out, after all, to be diagnostically distinguishable - see any physics textbook that refers to Democritus. In these cases, we often find that the archaeological interpretations turn on disputable semantic equivalences. Would Democritus recognise modern atomic theory as a descendent of his speculations? (The answer is, almost definitely, "no" ...)

I don't want to get lost in the semantic woods, however, because there may be a more interesting aspect of the 'Matrix World' metaphysic that has some very deep and general consequences.

A computer is a machine which can manipulate the physical representations of logical symbols in a way which mimics the syntactical rules which determine the use of these symbols in a logical script. In other words, it is a machine in which purely mechanical interactions are organised so that they mimic purely syntactical ones.

It can do this so well, that it can appear to have a kind of 'semantic awareness'. (Soon, we will have machines which will appear to make this claim about themselves, just as we do.)

If we set aside the possibility of demonstrable emancipation (a 'red pill' exit), then we must imagine a possibility in which everything is simulated (including our discovery that we have real biological bodies that are being kept in a giant incubator.)

I'm going to duck a lot of complicated issues here by making a proposal: that the central thing that needs to be simulated, in such a world, is our conviction that we are talking to one another about a real world that we all inhabit. In other words, our experience of ourselves as language users. This may seem to leave out a lot of private phenomenology, but then so do our actual linguistic exchanges - our beetles stay in their  boxes. For the purposes of exploring some of the issues raised by the Matrix World hypothesis, I'm going to think of the language we are using as a kind of script that could, in principle, be generated by a computer following syntactical rules.

I might as well say, here, with a nod to Liebniz, that all that is really needed is a simulation that convinces me that I live in a world with other language users, and that I am having this conversation with (one or two of) them. I'll pass over this possibility, since it generates the same issues as a more general (and perhaps more friendly) simulation would.

I'm also going to pass over all of the problems associated with simulating 'consciousness'. I have two reasons for doing this - one is that we don't yet have a clear enough understanding of what we mean by 'consciousness' to be able to use the word in this context, and the other is that I plan to drop 'meaning' altogether for a while, and consider the model in purely syntactical terms. This renders 'consciousness' as a syntactical construct - a string or characters, if you like - whose use is determined by the way the Matrix-support machine operates. I don't think I'm doing any harm to my central argument here, although - again - I realise I'm walking past some doors that look open ...

So: lets think about the Matrix machine, which I will now represent as an 'adequate' syntax manipulator, because it feeds me strings of terms, and responds to the terms I feed it, in a way that convinces me that I am having a conversation with someone (as in John Searle's 'Chinese Room' experiment). It doesn't matter here that I am 'part' of the simulation (this is one of the meaning ambiguities of 'we live in a Matrix World'.) What matters is that there is something going on that works as a convincing simulated conversation. This must be possible since we have to have some vehicle for our speculations - a language in which we can have philosophical discussions about Matrix Worlds, among other things.

Obviously, a minimum requirement for implementing a language simulation is a syntax simulator. And this is all that can be expected of a Matrix computer. Since such a machine 'contains' the whole world, even a simplistic reference semantics would be reproduced by the machine's internal processes, and so reduced to syntax.

Now it doesn't really matter very much what kind of ontological status we give the underlying machine - whether we think of it as a physical machine (with causal interactions), or as some kind of abstraction (per Max Tegmark) or as comprising some 'non-physical' support substance (a la DesCartes). In all cases, we have one kind of interaction (causal, formal, or spectral) being organised to mimic (to be isomorphic to) another, strictly syntactical, one. We have a 'something in the world' which is able to convince me that I am having an intelligible conversation with someone, when the only 'evidence' for this conversation is a string of physical symbol representations, a string of symbols (as an abstraction), or a more mysterious 'mechanism of the mind' that directly produces this conviction in me.

Another open door I have to acknowledge here is that all of these routes to explanation generate open question paradoxes. I've talked about this at some length already in this blog. No substrate can guarantee the truth of any statement it generates, since our articulation of  the substrate as a reliable truth-generator is a  computation in the language we wish it to validate. The substrate simply becomes part of a paradox-generating 'theory of truth'. I am walking past this door for the moment because the open question paradox it generates is a semantic paradox. It depends upon a semantic concept of truth.

The real power of Gödel's paradoxes is that they are syntactical paradoxes. They are generated by the capacity of an entirely formal system (in so far as such a thing can be conceived) to model its own theorem determination processes. They appear, for example, when we use arithmetic to determine which of two sets a Gödel number should be allocated to, when the number encodes the arithmetical mechanism for allocating itself to one of the sets in such a way as to frustrate any such allocation. (This is why it comes out as a version of Russell's paradox).

Which is also, I guess, why Gödel's proof theorems and Alan Turing's computability results are equivalent. My Matrix generating Turing machine which passes the 'Turing test' (for intelligibility, perhaps, rather than intelligence) must also generate paradoxes when it is set to compute certain characteristics of its own behaviour, because these computations must be possible to represent in the 'language' that the Matrix denizens speak. This is at the centre of what I'm trying to do, however coarsely.

There is nothing that can count as a computation, but (at the same time) not be possible (in principle) to represent in 'our' language (the language of the conversation I am having with you). We know this from the general nature of a Turing machine, which can emulate any other computing machine, and so emulate any set of syntactical rules. Since we can model a Turing machine using our language, we can also (in principle) model any set of syntactical rules in our language - including the rules of the 'language/world' composite that the Matrix machine must instantiate.

If the Matrix machine must (as a minimum) be able to reproduce the syntax of a language/world combination in a way which is semantically convincing to the 'language user' components (us) of its simulated world (to be a bit brutal with the metaphysics), it must do this by following machine rules which mimic syntactical structures which can be represented and reproduced within the language of these simulation components (us, in other words). This is the only way that its mechanisms could count as mechanisms of a Matrix machine. We may 'explain' our own existence by referring to a mysterious deity (perhaps having a dispute about whether this is an explanation), but we cannot call the Matrix machine (essentially) 'mysterious' without immediately conceding that it is not a machine.

Where this leads us is to something like this: the Matrix hypothesis, in order to be intelligible, requires the machine which generates the simulacrum to work according to intelligible rules; in fact to syntactical rules which can, in principle, be formulated in human language - perhaps in many hundreds of billions of lines of computer code, explicitly mapped to described physical states of the machine, plus a description of its physical processes that demonstrates its reliability as a syntax emulator. And this must be reconciled with the fact that the central illusion that the Matrix World maintains is exactly that we have a language in which we can speculate about things like whether we live in a Matrix World. Remember that it must be true that we can speculate thus, because that's what we're doing now.

This might be a long-winded way of saying something like this:

The hypothesis that we live in a Matrix world contains the hypothesis that this world is supported by a humanly (in principle) intelligible machine - i.e. a machine whose operations could (again, in principle) be described in human language. The exchange of tokens which comprises this language, and is the syntactical substrate of our semantic hypotheses, is determined by the operations of the machine. These operations, on the other hand, are only intelligible if they can be represented within the language we 'speak' by exchanging these tokens ...

Where does this leave us?

When I first started to think about this, I thought I would be able to show that the Matrix World hypothesis, and by extension, all other 'metaphysical substrates' would turn out to be not only semantically incoherent (as various forms of the open question paradox show them to be) but also syntactically incoherent, because the rules of the matrix machine would have to be 'representable' in the physical language tokens that its inhabitants exchanged when they spoke to one another - including when they said things like 'We live in Matrix World ...'.

But now I'm stuck, and it's because there isn't a method of producing this representation itself in purely syntactical terms. There is no equivalent of  the 'Gödelisation' of arithmetic for language in general, because the Gödelisation of arithmetic depends upon a semantic framework. While we can formulate the rules of logic in arithmetic in a way which allows us to calculate arithmetical isomorphs of logical theorems, I don't think we can represent the fact that we can do this in the logical or arithmetical scripts. And this is what would be required to show that the Matrix world hypothesis is not only semantically paradoxical, but also syntactically incoherent.

If anyone out there thinks this makes sense, and can either see a way forward or confirm my suspicions, please get in touch.

Counselling Conversations

No scientific theory, however 'abstract' or 'objective' makes sense independently of its being part of a conversation, and a conversation requires more than one participant. Our intuitions about the independence of theory is possibly a consequence of our phenomenological encounters with reality, and its intransigence in the face of our common desires for it to be other than it is. And perhaps also our habit of writing things down, so rendering them, apparently, and misleadingly, independent of any specific interlocutors.

It is because of this that an intense therapy session can feel like a productive seminar - new ways of talking are being explored in both. Also, in both circumstances, our private phenomenological spaces and our capacity to interact with one another seem to meet, or to engage most fluently. We acquire important insights and misleading metaphors here.

What has this to do with validity? (For which we always seem to seek an 'external' validation, of all things ...)

Well: Validity is grounded in the playability of the game, and 'reality' restricts the games that we will, or can, find playable. I'm in inclined to think that this is the only epistemological role that some conception of reality has (or needs to have). That it is, in Bob's words, 'parsable' - though not, obviously, in any completely determinate way.

So, in our intense conversations, we find three 'worlds' (as Karl Popper might have agreed, though about little else here ...) - the world as we talk about it, the private phenomenological world which we feel drives our talk and actions, and 'reality' - the invisible and intractable generator of barriers and trips that, sometimes unexpectedly, appear upon our field of play.

And all this is fine, and, as I have argued, not only supplies an intelligible epistemological groundwork, but is the only way such a groundwork can be made intelligible.

What else does it have to say, though?

There are normative aspects of this - not only with respect to the attitudes we must take to interlocutors (for them to be interlocutors) - e.g. that we regards them as, broadly (and in context), honest and competent. This is not an easy norm to render as a set of rules, of course, so some of what follows may be implied by it anyway. However, I'm going to articulate a few thoughts I have:

Rogerian counselling (much mocked by the 'Eliza' caricature) is based on attitudes of positive regard, empathy, and congruence. The 'counsellor' takes the 'client' completely seriously, tries to see their world as they see it, and exhibits a kind of holistic honesty of practice that is more fundamental than literal truth telling. It is the relationship between this counsellor and the client that does the therapeutic 'work', and this relationship requires the whole presence of the counsellor.

Rogers drew his conclusions from what he saw in effective therapy, but believed that the 'person-centred' attitudes suggested a powerful approach to social and political conflicts.

I have worked as a person centred counsellor, and would make the following observations:

(1)

A conversation in which one participant is the arbiter of sense quickly becomes unintelligible to all. This is because 'I am the arbiter of sense' can only make sense in its own terms, which are systematically opaque (we have to keep asking the arbiter how to make sense, to the point where even our questions do not make sense independent of the arbiter's adjudications ...).

In a person-centred counselling relationship, the counsellor is not an 'expert' - he or she does not diagnose, prescribe, or advise. (If there is an 'expert' in the room, it is the client - who knows much more about his or her own internal world than the counsellor does.) Instead, the counsellor is an interlocutor - a real person with whom the client can have a real conversation, which includes tacit negotiation of meanings and linguistic rules. Potential failures of intelligibility - if they are consequential for the relationship - are negotiated mutually. The participants explore new ways of talking.

(2)

Bertrand Russell is meant to have said that you should only criticise the strongest version of your opponent's position that you can construct. (There is some dispute about whether he followed his own advice, but sometimes it is hard to distinguish malice from cognitive failure.)

In a real conversation, we must engage with the motivations and emotional concerns of our interlocutors. This cannot be empty - condescension would be damaging to the relationship. We have to do this, because without some understanding here we do not know what they mean by the things that they say.

A good public example of what goes wrong here is the vacuous debate between fundamentalists on both sides of the 'god' debate. If each tried more seriously to understand what was going on for the other, there would be less stone-throwing and more discovery. (I am reminded of Desmond Tutu's remark that 'God is the potential for goodness that exists in humanity'.)

Even (!) among philosophers, it is clear that some positions (both positive and negative) have an appeal that is hard to justify in theoretical terms alone. Some empathic recognition of the emotional needs which drive these preferences might really help to clear the air.

(3)

The capacity for congruence is acquired through practice. It is acquired experientially, just as the capacity to speak is acquired experientially. We become honest and competent interlocutors through participating in conversations.

There are no external criteria of congruence, although incongruence is likely to reveal itself as a kind of 'bad faith'. That is, as an attempt to achieve some subterfuge - not always as clear cut as a lie, as it can often be, in a sense, 'unconscious'. Incongruence can arise from unprocessed conflicts in the counsellor (e.g. suppressed shame related to issues which the client wishes to explore). An important danger of incongruence in a counselling context is that it can lead to the counsellor responding to the client in a way that is not connected to the client's distress, but to the counsellor's.

In fact, of course, there cannot be any external criteria if congruence is closely related to honest competency, for exactly the same reasons that there can be no absolute criteria for successful translation (Davidson and Quine) or for rule-following (Kripke). Similarly, the possibility of congruence cannot be dispensed with without dispensing with the possibility of honest and competent interlocutorship (and so, unintelligibly, with the possibility of language).

---------------------------

There will be more on this. There is some relationship between the productivity of counselling conversations and the establishment of the semantic basis for formal reasoning that is worth exploring further.

Consciousness and Metaphysical Dissatisfaction

We each have an ‘internal world' - a world of something like 'raw phenomenological states' -  which seems to inform our shared public judgements. Or so I think, partly because many people seem to share my public judgement about this (!) More convincingly, I find that when I try to talk to someone about this internal world, there seems to be a good deal of congruence between aspects of my internal world and aspects of theirs.

As this kind of conversation between us progresses, however, two things become clear:

(1) That we cannot distinguish between (a) sharing our perceptions of our phenomenological state and (b) agreeing about how to talk about our phenomenological states.

(2) That there are some irreducible differences of perspective - I always see the world from within my 'sensorium', and my interlocutor always sees it from within theirs. I cannot see through their eyes, and they cannot see through mine. As Thomas Nagel has famously observed, it isn't even clear what my seeing their world as they do might mean.

And these two are not unrelated: The 'private perspective' issue seems to open a crack that allows a kind of linguistic scepticism to become intelligible: how do I know that you and I are not systematically confused when we describe, to each other, our phenomenological conditions? How do we know that our apparent agreement is not a product of this confusion?

This particular scepticism cannot be addressed recursively, because it does not threaten our general confidence that we can be intelligible to one another. We could be quite mistaken about phenomenological congruence, but still make sense otherwise. Somewhat bizarrely, for epistemological empiricists, agreement about how to talk about our phenomenological states is all that is required - it really doesn't matter whether this agreement is 'grounded' on some 'real' similarity.

(There are, of course, other problems with rendering an 'internal' scepticism about language intelligible - we need to imagine someone who understands how human language works, but 'sees', in some way, that human beings are systematically deceived by it. Their conclusions obviously cannot be represented in any human language, or in any language that can be translated into a human language - and so we find ourselves in Davidson and Quine's 'radical translation' maze. How can we establish that an 'untranslatable language' is a language?

If we cannot represent a belief in any conceivable language, how do we know what the belief is?

How can I possibly know what kind of sense I am making when I say 'George believes that language does not work'?)

The ‘internal awareness’ features of consciousness are, in so far as they are referred to in our shared language, ‘beetle in a box’ concepts. The grammar of the concept does not require that we all have the same experiences, in some independently verifiable way – it only requires that we have experiences which allow us to learn the shared grammar. In fact, issues of ‘sameness’ do not really arise here, as no independent public judgements about private sameness are possible. Even the question whether I have the ‘same’ private experience on different occasions is difficult to separate from my public engagement with the relevant grammar, and runs into ‘private language’ problems that cannot be resolved.

Consciousness fulfils a number of semi-explanatory roles, e.g. as a repository of self-awareness, and particularly of self-aware intentional states; as a collection of cognitive mechanisms revealed through certain kinds of intelligible behaviour, perhaps especially behaviour which seems to reveal an awareness of others’ 'cognitive states'; as the 'subject space' of perception; as a property of the 'soul substance'. And others. Not all mutually congruent…

Since some time in the 17th century, however, we seem to have been convinced that the 'internal world' somehow supports and validates what we say about the public world. If I tell you that the kettle is boiling, then this is because some features and mechanisms of my internal space have led me to be convinced that this is the case, and that I can be sure I am not mistaken. It is from this period that we began to think that even the meaning of what I say about the kettle is somehow tied up with this hypothesised inner theatre.

I suppose it should be obvious to any reader of this blog that I don't think anything remotely like this is true. I am, however, interested in why this kind of picture is appealing, and what (apparently unconnected) errors it might lead us into. Perhaps I am also interested in discovering other ways to think here.

Among the errors I include a certain kind of metaphysical conviction which is impossible to coherently articulate: that there is a substrate to the way we speak that is, somehow, represented in our speech; that warrants it, that makes it 'true'. And that there are respects in which this substrate 'looks' like what we see when we look out through the windows of our minds onto the landscape of the world. This picture, I think, underlies most 'realisms'.

It is incoherent, because, of course, anything we say about it must also be a representation - e.g. of the relationship between what we say and the substrate - and so would require another substrate. Unlike mathematical correspondence, which is an articulated relationship between pairs of objects which are represented equally; aletheial correspondence is between an object which is a representation and one which is not; between a description, a name, or a theory and the 'thing' which is described, named, or hypothesised about, but which is only philosophically accessible through that same describing, naming or hypothesising.

But what is the source of this model? Of the 'metaphysical dissatisfaction' which Barry Stroud finds so uncomfortable - and which I sense, but really want to give a very different kind of account of?

One very simple confusion has contributed to it, and that is between discovery and justification: failing to distinguish between how we become convinced of the truth of a statement, and how we justify it. Seeing may be believing with respect to the same subject, but my seeing is not your believing. If we both 'see something at the same time', justification may not be required, but these occasions are uncommon. And although justification may not be required, it turns out, notoriously, to be difficult to provide when it is. The request for justification immediately raises a doubt about whether we did 'see the same thing'. What starts out as a case of simple shared experience turns out to be quite difficult to give an account of.

I suppose that, for each of us, our sensoria are so viscerally part of our experience of our own cognitive processing that it is difficult to imagine that, from the point of view of shared theorising, they are epistemologically irrelevant. The feeling of certainty associated with a direct experience does not play a role in how we demonstrate the truth of what we say about it, unless that role is agreed with our interlocutors. It cannot be shown to have a necessary role unless no agreement is possible without it - and it is easy to show that this position is either false or question begging. I have no way of testing your account (to me) of your internal states that does not depend on my trusting your accounts of this kind...

In the film 'Bladerunner', a central plot device is Rick's use of psychometric testing to distinguish 'replicants' from 'real' human beings. The romantic plot begins when he tests Rachel - a replicant who, unlike the others, is not trying to deceive him. She completely believes (should I say 'believes'?) in her own humanity, but fails his tests.

Under duress, we could all find ourselves in Rachel's predicament - in fact, it is easy to see her in any person emotionally abused to the point where their cognitive capacities are undermined. What is very plain about people in these circumstances is that - so far as can be reasonably determined - their internal world is scrambled by their social circumstances. There is no protected reservoir of sense and meaning that they can retreat to.

This is one problem with (slightly heavy handed) 'swamp man' and 'zombie' examples - from any socially realistic point of view, a creature judged to be of this kind would be driven quickly mad. Once again, science fiction shows more insight than philosophy ...

These examples are generally set up as problems for a certain kind of 'physicalist' or 'mechanist' metaphysics, but - of course - they are problems for any 'substance' based substrate. Or for any non-recursive substrate, for that matter. If we need 'consciousness' to give an account of sense or intelligibility, then we have no way of giving a sensible or intelligible account of consciousness.

I must be really clear here: I am not arguing that consciousness is some kind of illusion, or 'epiphenomenon', or must be given an account of in some other kind of spooky way. There are two anchors I think we can rely on: (1) believing that we are conscious is not irrational (2) we do not need to give (and indeed cannot give) a reductionist account of it (because any such account will generate open question paradoxes).

(I see, and accept, that these have the consequence that not everything we rationally believe in is tractable to reductive scientific investigation. What others might not see is that they also have the consequence that introducing some kind of 'non-scientific rationality' - e.g. an irreducible supernaturalism - will be a blind alley. No substrate works here: not physicalism, not information theory, not mathematics, not theology, not sociology of knowledge. No substrate whose validity can intelligibly be questioned will work. My apologies to all those theorists of consciousness who work in these fields - many of them are doing interesting and useful work, but it is not about consciousness.)

The central reason why believing in consciousness is not irrational is that we must attribute it to an interlocutor as a condition of attributing intentional states. It is the attribution of intentional states which distinguishes an interlocutor from some other device which we might engage with using language-like processes. I do not believe that I talk to my phone, or that a tape recorder can speak, exactly because I do not attribute intentional states to these devices.

I cannot say 'X says it is raining', claim that X is an honest and competent interlocutor, and also say that X does not believe that it is raining, or that X does not know that X believes this. (Or, worse still, say that it is not possible for X to have intentional states at all). Saying 'It is raining' cannot be reduced to producing a string of characters, or making certain noises, or engaging with a signalling process. (I can, of course, wonder whether X is an honest and competent interlocutor - but I cannot resolve this by coming to some explicit agreement with X about it.)

The language of intentional states also seems to be useful when describing creatures without language when we want to attribute actions to them (as opposed to behaviours). Freud, notoriously, introduced the the possibility of  'unconscious' intentional states to explain the actions of his patients, with a similar motivation. In both of these cases, the attribution of intentional states is essential to the theory, but potentially dispensable (if some other theory were available). To regard someone as an interlocutor is, necessarily, to attribute conscious intentional states to them.

But what about Stroud's metaphysical dissatisfaction? And mine, for that matter ...

I might tritely say that we all have our private miseries, but this would not be fair. After all, these miseries have, explicitly or implicitly, influenced our theoretical discourses. I might even, I think, find someone who did not have these anxieties slightly disturbing - even deluded. (Zombie-like?)

Giving them a general characterisation - 'existential loneliness', for example - doesn't seem to help much, either.

I think it goes something like this: it is very hard for me to comfortably believe that I understand your intentional states, if I must also - however abstractly - recognise that your phenomenological world, your sensorium, may be radically - Nagellianly - different from mine. Somehow, accepting each other as honest and competent interlocutors seems to carry this baggage.

And I suppose the way I address this - basically emotional - issue is by keeping the question abstract, while giving you a nudge and saying 'can you see what I mean here?'

A note from the sea ...

Our signals cannot decode themselves, and the result of linguistic decoding is not just a new signal, re-representing, re-encoding, some content or consequence of the original.

A 'theory of everything' would, somehow, have to decode itself. It would have to contain a theory of its own meaning. Perhaps a theory like this goes beyond what physics might aspire to, but I am not sure about this. Already we approach certain problems with attempts that few can understand...

(What good is it to me if the truth can only be expressed in Martian? Or perhaps what I write here is equally opaque ...)

When we contradict ourselves, we destroy meaning - but we do not guarantee meaningfulness just by avoiding contradiction. We cannot guarantee meaningfulness at all. We imagine that we 'talk' by exchanging these tokens, and find that even identifying and circumscribing the 'tokens' turns out to be a murky business. We think we can, with the tokens, write down 'rules' about how the tokens should be used. We have imagined that these rules can form an exhaustive account, or structure, or, perhaps machine ... but they cannot. As we tidy each room and photograph it, the lights go out - insight evaporates, rather than deepening.

We try new tokens: 'meaningfulness', 'intelligibility', 'philosophy' ... the pile gets higher, but we see no more from its summit than our own ability to build it, from when its content is derived.

Representation and Content

(S) "The statement S describes C"

This is a common construction, and it suggests a relationship between 'things' (events, objects) in the world and the assertions which represent them.

The statement that a list of such statements can be constructed must be a member of the list, since it is a statement about the way the world is - given the picture of 'description' that is suggested here, the possibility of such a list would be a fact about the world.

Also, and somewhat intractably, any statement about why this was the case would also have to be a member of the list. This obviously generates an OQ paradox, since membership of the list is a kind of truth test.

Finally, since any general account of representation could be used to generate such a list (to determine whether a statement was a member of it or not), any general account of representation generates an OQ paradox.

Taking representation - somewhat loosely, but, I suspect, validly - as a proxy for content, I think that any account of content must generate the same problem. This includes supernatural accounts (mind, emergence) etc.

Since we can't intelligibly deny that our statements have content (since such a denial would render itself unintelligible), we are left with only the possibility of a recursive account - an account which makes the nature of content, and the validity of attributing content, depend upon the possibility that some things we say (including our discussion of content) must have content.

This applies to our most 'basic' statements about the world - the kinds of simple statement of fact that a naïve materialist might wish to take as metaphysical fundamentals. Since we can only guarantee their content by a recursive account, they cannot be the kinds of fundamentals that are required.

And this also applies to basic statements about the tools of language.

If we think of these tools just in terms of strings of text and rules for manipulating them, we find that statements about the natures of the strings (what they contain) and the consequences of the rules can only be validated recursively. The illusion of simplicity here arises from the 'obviousness' of the basic presumptions - an obviousness which is rooted in, but not validated by, something like our 'language instinct'. Human beings just find some things more 'obvious' than others, and persist in trying to give this metaphysical or epistemological significance.

Imagining the world as mechanism - and intelligence as complex signalling and manipulation machinery - may be a useful heuristic, especially for those scientists engaged in enquiries which seek to produce models of this kind. But the usefulness of the heuristic, like the 'obviousness' of some 'atomistic' statements of fact or rule, is just epistemologically misleading.