The Nightmare Scenario

rosetta stoneMicrosoft recently announced the first public beta preview for Skype Translate, a service which will provide immediate translation during voice calls. For the time being only Spanish/English is working but we’re told that English/German and other languages are on the way. The approach used is complex. Deep Neural Networks apparently play a key role in the speech recognition. While the actual translation  ultimately relies on recognising bits of text which resemble those it already knows, the same basic principle applied in existing text translators such as Google Translate, it is also capable of recognising and removing ‘disfluencies’ –  um and ers, rephrasings, and so on, and apparently makes some use of syntactical models, so there is some highly sophisticated processing going on.  It seems to do a reasonable job, though as always with this kind of thing a degree of scepticism is appropriate.

Translating actual speech, with all its messy variability is of course an amazing achievement, much more difficult than dealing with text (which itself is no walk in the park); and it’s remarkable indeed that it can be done so well without the machine making any serious attempt to deal with the meaning of the words it translates. Perhaps that’s a bit too bald: the software does take account of context and as I said it removes some meaningless bits, so arguably it is not ignoring meaning totally. But full-blown intentionality is completely absent.

This fits into a recent pattern in which barriers to AI are falling to approaches which skirt or avoid consciousness as we normally understand it, and all the intractable problems that go with it.  It’s not exactly the triumph of brute force, but it does owe more to processing power and less to ingenuity than we might have expected. At some point if this continues, we’re going to have to take seriously the possibility of our having, in the not-all-that remote future, a machine which mimics human behaviour brilliantly without our ever having solved any of the philosophical problems. Such a robot might run on something like a revival of the frames or scripts of Marvin Minsky or Roger Schank, only this time with a depth and power behind it that would make the early attempts look like working with an abacus. The AI would, at its crudest, simply be recognising situations and looking up a good response, but it would have such a gigantic library of situations and it would be so subtle at customising the details that its behaviour would be indistinguishable from that of ordinary humans for all practical purposes. What would we say about such a robot (let’s call her Sophia, why not since anthropomorphism seems inevitable). I can see several options.

Option one. Sophia really is conscious, just like us. OK, we don’t really understand how we pulled it off, but it’s futile to argue about it when her performance provides everything we could possibly demand of consciousness and passes every test anyone can devise. We don’t argue that photographs are not depictions because they’re not executed in oil paint, so why would we argue that a consciousness created by other means is not the real thing? She achieved consciousness by a different route, and her brain doesn’t work like ours – but her mind does. In fact, it turns out we probably work more like her than we thought: all this talk of real intrinsic intentionality and magic meaningfulness turns out to be a systematic delusion; we’re really just running scripts ourselves!

Option two. Sophia is conscious, but not in the way we are. OK, the results are indistinguishable, but we just know that the methods are different, and so the process is not the same. birds and bats both fly, but they don’t do it the same way. Sophia probably deserves the same moral rights and duties as us, though we need to be careful about that; but she could very well be a philosophical zombie who has no subjective experience. On the other hand, her mental life might have subjective qualities of its own, very different to ours but incommunicable.

Option three. She’s not not conscious; we just know she isn’t, because we know how she works and we know that all her responses and behaviour come from simply picking canned sequences out of the cupboard. We’re deluding ourselves if we think otherwise. But she is the vivid image of a human being and an incredibly subtle and complex entity: she may not be that different from animals whose behaviour is largely instinctive. We cannot therefore simply treat her as a machine: she probably ought to have some kinds of rights: perhaps special robot rights. Since we can’t be absolutely certain that she does not experience real pain and other feelings in some form, and since she resembles us so much, it’s right to avoid cruelty both on the grounds of the precautionary principle and so as not to risk debasing our own moral instincts; if we got used to doling out bad treatment to robots who cried out with human voices, we might get used to doing it to flesh and blood people too.

Option four.  Sophia’s just an entertaining machine, not conscious at all; but that moral stuff is rubbish. It’s perfectly OK to treat her like a slave, to turn her off when we want, or put her through terrible ‘ordeals’ if it helps or amuses us. We know that inside her head the lights are off, no-one home: we might as well worry about dolls. You talk about debasing our moral instincts, but I don’t think treating puppets like people is a great way to go, morally. You surely wouldn’t switch trolleys to save even ten Sophias if it killed one human being: follow that out to its logical conclusion.

Option five. Sophia is a ghastly parody of human life and should be destroyed immediately. I’m not saying she’s actuated by demonic possession (although Satan is pretty resourceful), but she tempts us into diabolical errors about the unique nature of the human spirit.

No doubt there are other options; for me. at any rate, being obliged to choose one is a nightmare scenario. Merry Christmas!

Actual Consciousness

honderich 3Ted Honderich’s latest work Actual Consciousness is a massive volume. He has always been partial to advancing his argument through a comprehensive review (and rejection) of every other opinion on the subject in question. Here, that approach produces a hefty book which in practice is about the whole field of philosophy of consciousness. There is a useful video here at IAI of Ted grumpily failing to summarise the whole theory in the allotted time and confessing with the same alarming frankness that characterised his autobiography to wanting to be as famous as Bach or Chomsky, and not thinking he was going to be. If you want to see the whole thing you’ll have to sign up (free); but they do have a number of good discussions of consciousness.

The theory Honderich is advancing is a further version of the externalism which we discussed a while ago; that for you to be conscious is in some sense for something to exist (or to be real, hence the ‘actual’ in Actual Consciousness). At first sight this thesis has always seemed opaque to the point of wilful obscurity, and the simplest readings seem to make it either vacuous (for you to be conscious is for a state of consciousness to exist) or just evidently wrong (for you to be conscious is for the object of your awareness to exist). He means – he must mean – something subtler than that, and a few more clues can only be welcome.

First though, we survey the alternatives. Honderich suggests (and few would disagree) that the study of consciousness has been vastly complicated by differing or inadequate definitions. This has led philosophers to talk past each other or work themselves into confusions. Above all, Honderich thinks virtually everyone has at some point fallen into circularity, smuggling into their definitions terms that already include consciousness in one form or another.

He sets out five leading ideas: these are not actually the five parts into which he would carve consciousness himself (he would analyse it into three: perceptual, cognitive and affective consciousness) but these are the ideas he feels we need to address. They are: qualia, ‘something it is like for a thing to be that thing’, subjectivity, intentionality, and phenomenality. More normally these days we divide the field in two initially, and at first glance four of Honderich’s views look like the same thing from different angles. When there is ‘something it is like’, that’s the phenomenal aspect of experience as had by a subject and characterised by the presence of qualia. But let’s not be hasty.

Having reviewed briefly what various people have said about qualia, Honderich notes that one thing seems clear; that it is always conceived of as distinct from, and hence accompanied by, another form of consciousness. Some people certainly assert that qualia are the real essence of consciousness or at any rate of the interesting part of it; but it does seem to be true that no-one proposes conscious states that include qualia and nothing else. That in itself doesn’t amount to circularity, though.

The next leading idea is something it is like to see red, or whatever. Nagel’s phrase is unhelpful but somehow powerfully persuasive. We all sort of know what it is getting at. Honderich notes that Nagel himself offered an improved version that leaves out the suggestion that a comparison is going on; to be conscious, in this version, is for there to be something that is how it is for you (to see red or whatever). What does this all really mean? Honderich suspects that it comes down to there being something it is like, or something that is how it is, for you to be conscious (of something red, eg), once again a case of circularity. I don’t really see it; it seems to me that Nagel offers an equation; consciousness is there being something it is like; Honderich pounces: Aha! But there being something it is like is being conscious! That just seems to be travelling back from the second term of the equation to the first, not showing that the the second term requires or contains the first. I’m simplifying rather a lot, so perhaps I’ve missed something. But so far as I can see while Honderich justly complains that the formula is uninformative, the only circularity is one he inserted himself.

Subjectivity for Honderich means the existence of a subject. The word, as he acknowledges, can often be used as more or less a synonym for one of the two senses already discussed: in fact I should say that that is the standard meaning. But it’s true that consciousness is tied up with the notion of an experiencing self or subject (and those who deny the existence of one are often sceptical about the other). Honderich suggests that it is implicit in the idea of a subject that the subject is conscious, and though we can raise quibbles over sleeping or anaesthetised subjects, he is surely on firmer ground in seeing circularity here. To define consciousness in terms of a subject is circular because to be a subject you have to be conscious. But nobody does that, or at least, no-one I can think of. It’s sort of accepted that you need to have your consciousness sorted out before you can have your conscious agent.

With intentionality we come on to something distinctly different; this is the quality of aboutness or directedness singled out by Brentano. Honderich bases his comments on Brentano’s account, which he quotes at full length. It’s only fair to note in passing that Brentano was not talking about consciousness; rather he asserted that intentionality was the mark of the mental; but there is obviously a connection and we might well try to argue that intentionality was constitutive of consciousness.

Honderich notes an intractable problem over the objects of intentionality; they don’t have to exist. We can think about imaginary or even absurd things just as easily as about real ones. But if we are not thinking about a real slipper when we think of Cinderella’s glass one, then surely we’re not really thinking about the actual one the dog is chewing in the corner, either; perhaps the real objects are just our ideas or images of the slipper, or whatever. If we don’t take that path, suggests Honderich, then this intentionality business is no great help; if we do suppose that thinking about things is thinking about a mental image, then we’re back with circularity because it would have to be a conscious image, wouldn’t it?

Would it? I’m not totally sure it would; wouldn’t the theory be that it becomes a conscious image when it’s an object of conscious thought, but not otherwise or in itself? Otherwise we seem to have a weird doubling up going on. But anyway, it’s too clear, in my opinion, that thinking about a thing is thinking about that thing, not thinking about an idea of it; we have to find some other way round the problem of non-existent objects of thought. So we’re left with the complaint that intentionality does not explain consciousness – and that’s true enough; it’s at least as much a part of the problem.

With phenomenality we’re back with a word that could be taken as meaning much the same as subjectivity, and referring to the same stuff as qualia or something it is like. Honderich draws on Ned Block’s much-cited distinction between access or a-consciousness and phenomenal or p-consciousness, and attacks David Chalmers for having said that qualia, subjectivity, phenomenality and so on, are essentially different ways of talking about the same thing. I’m with Chalmers; yes, the different terms embody different ways of approaching the problem, but there’s little doubt that if we could crack one, translating the solution into terms of the other approaches would be relatively trivial. Oddly, instead of recapitulating the claimed important distinctions, which woulod seem the natira;l thing to do at this point, Honderich seems to argue that if Chalmers thinks all these things are the same thing, they must in fact all be examples of something more fundamental, in which case why doesn’t Chalmers talk about the fundamental thing?

If there are the grammatical or subtle differences between the terms for the phenomena, the things do make up ‘approximately the same class of phenomena’. What class is that? To speak differently, what are these things examples of ? In fact didn’t we have to have some idea of that in order to bring the examples together in the first place? What brings the different things together? What is this general fact of consciousness? It has to exist, doesn’t it? Chalmers, I’d say, has credit for bringing the things together, but he might have asked about the general fact, mightn’t he?

This is strange; to assert that a number of terms refer to the same thing is not necessarily to assert that they are all in fact yet another thing. My best guess is that Honderich wants to manoeuvre Chalmers into seeming circularity again, but if so I don’t think it comes off.

Honderich goes on to an extended review of the territory and what others have said, but I propose to cut to the chase. Cutting to the chase, by the way, is something Honderich himself is very bad at, or rather pathologically averse to. He has a style all his own, characterised by a constant backing away from saying anything directly; he prefers to mention a few things one might say in relation to this matter, some indeed that others have at times suggested he himself might not always have been committed to denial of, that is to say these considerations might be ones – not by any means to characterise exhaustively, but nevertheless to bring forward as previously hinted – perhaps to be felt to be most significantly indicated or at any rate we might choose, not yet to think, but to entertain the possibility, of considering as such. Sometimes you really want to kick him.

Anyway, to get to the point: Honderich is an externalist; he thinks your perception of x is something that happens out there where x is real and physical, not in your head. There is an extension to this to take care of those cases where we think about things that are imaginary or otherwise non-physical; in such cases the same thing is going on, it’s just that the objects perceived are representations in our mind. In a sense this is externalism simply redirected to objects that happen to be internal. Of course, how anything comes to be a mental representation is itself a non-trivial issue.

Honderich says that for you to be conscious of something is for something to be real, to exist. This puzzling or vacuous-seeming formula is underpinned by the eyebrow-raising idea of subjective physicality. This is like a kind of Philosopher’s Stone; it means that what we perceive can be both actual and physical in a perfectly normal way, yet subjective in the way required by consciousness. How can we possibly eat our cake this way and yet still have it? It’s kind of axiomatic that the actual qualities of physical things don’t depend on the observer (yes, I know in modern physics that’s a can of worms, but not one we need to open here), while subjective qualities absolutely do; my subjective impressions may be quite different to yours.

How is this trick to be pulled off?

The general answer to the question of what is actual with your perceptual consciousness, putting aside that in which it may issue immediately, is a part, piece or stage of a subjective physical world of several dependencies, out there in space, and nothing else whatever. Your being conscious now is exactly and nothing more than this severally-dependent fact external to you of a room’s existing…

It looks at first sight as if this talk of worlds may be the answer. The room exists subjectively for you and also physically, but in a world of its own, or of your own, which would explain how it can be different from the subjective experience of others; they have different worlds. This would be alarming, however, because it suggests a proliferation of worlds whose relationships would be problematic and whose ontology profligate. We’re talking some serious and bizarre metaphysics, of which there is no sign elsewhere.

I don’t think that is at all what Honderich has in mind. Instead I think we need to remember that he doesn’t mean by subjectivity what everyone else means. He just means there is a subject. So his description of consciousness comes down to this; that there is a real object of consciousness, whether in the world or in the brain, which is unproblematically physical; and there is a subject, also physical, who is conscious of the thing.

Is that it? It seems sort of underwhelming. But I fear that is the essence of it.  Helpfully, Honderich provides a table of his proposed structure:

honderich table

Yes, that seems to me to confirm the suggested interpretation.

So it kind of looks as if Honderich has used a confusingly non-standard definition and ended up with a theoretical position which honestly sheds little light on the real issue; yet these were the very problems he criticised in earlier approaches. I can’t deny that I have greatly simplified here and it might be that I missed the key somewhere in one of those many chapters – but frankly I’m not going back to look again.

Early Qualia

Locke with flowersThe problem of qualia is in itself a very old one, but it is expressed in new terms.  My impression is that the actual word ‘qualia’ only began to be widely used (as a hot new concept) in the 1970s.  The question of whether the colours you experience in your mind are the same as the ones I experience in mine, on the other hand, goes back a long way. I’m not aware of any ancient discussions, though I should not be at all surprised to hear that there is one in, say, Sextus Empiricus (if you know one please mention it): I think the first serious philosophical exposition of the issue is Locke’s in the Essay Concerning Human Understanding:

“Neither would it carry any imputation of falsehood to our simple ideas, if by the different structure of our organs, it were so ordered, that the same object should produce in several men’s minds different ideas at the same time; e.g. If the idea, that a violet produced in one man’s mind by his eyes, were the same that a marigold produces in another man’s, and vice versa. For since this could never be known: because one man’s mind could not pass into another man’s body, to perceive, what appearances were produced by those organs; neither the ideas hereby, nor the names, would be at all confounded, or any falsehood be in either. For all things, that had the texture of a violet, producing constantly the idea, which he called blue, and those that had the texture of a marigold, producing constantly the idea, which he as constantly called yellow, whatever those appearances were in his mind; he would be able as regularly to distinguish things for his use by those appearances, and understand, and signify those distinctions, marked by the names blue and yellow, as if the appearances, or ideas in his mind, received from those two flowers, were exactly the same, with the ideas in other men’s minds.”

Interestingly, Locke chose colours which are (near enough) opposites on the spectrum; this inverted spectrum form of the case has been highly popular in recent decades.  It’s remarkable that Locke put the problem in this sophisticated form; he managed to leap to a twentieth-century outlook from a standing start, in a way. It’s also surprising that he got in so early: he was, after all, writing less than twenty years after the idea of the spectrum was first put forward by Isaac Newton. It’s not surprising that Locke should know about the spectrum; he was an enthusiastic supporter of Newton’s ideas, and somewhat distressed by his own inability to follow them in the original. Newton, no courter of popularity, deliberately expressed his theories in terms that were hard for the layman, and scientifically speaking, that’s what Locke was. Alas, it seems the gap between science and philosophy was already apparent even before science had properly achieved a separate existence: Newton would still have called himself a natural philosopher, I think, not a scientist.

It’s hard to be completely sure that Locke did deliberately pick colours that were opposite on the spectrum – he doesn’t say so, or call attention to their opposition (there might even be some room for debate about whether  ‘blue’ and ‘yellow are really opposite) but it does seem at least that he felt that strongly contrasting colours provided  a good example, and in that respect at least he anticipated many future discussions. The reason so many modern  theorists like the idea is that they believe a switch of non-opposite colour qualia would be detectable, because the spectrum would no longer be coherent, while inverting the whole thing preserves all the relationships intact and so leaves the change undetectable. Myself, I think this argument is a mistake, inadvertently transferring to qualia the spectral structure which actually belongs to the objective counterparts of colour qualia. The qualia themselves have to be completely indistinguishable, so it doesn’t matter whether we replace yellow qualia with violet or orange ones, or for that matter, with the quale of the smell of violets.

Strangely enough though Locke was not really interested in the problem; on the contrary, he set it out only because he was seeking to dismiss it as an irrelevance. His aim, in context, was to argue that simple perceptions cannot be wrong, and the possibility of inconsistent colour judgements – one person seeing blue where another saw yellow – seemed to provide a potential counter-argument which he needed to eliminate. If one person sees red where another sees green, surely at least one of them must be wrong? Locke’s strategy was to admit that different people might have different ideas for the same percept (nowadays we would probably refer to these subjective ideas of percepts as qualia), but to argue that it doesn’t matter because they will always agree about which colour is, in fact yellow, so it can’t properly be said that their ideas are wrong. Locke, we can say, was implicitly arguing that qualia are not worth worrying about, even for philosophical purposes.

This ‘so what?’ line of thought is still perfectly tenable. We could argue that two people looking at the same rose will not only agree that it is red, but also concur that they are both experiencing red qualia; so the fact that inwardly their experiences might differ is literally of no significance – obviously of no practical significance, but arguably also metaphysically nugatory. I don’t know of anyone who espouses this disengaged kind of scepticism, though; more normally people who think qualia don’t matter go on to argue that they don’t exist, either. Perhaps the importance we attach to the issue is a sign of how our attitudes to consciousness have changed: it was itself a matter of no great importance or interest to Locke.  I believe consciousness acquired new importance with the advent of serious computers, when it became necessary to find some quality  with which we could differentiate ourselves from machines. Subjective experience fit the bill nicely.