Experimental Free Will

Picture: Experiment. Shaun Nichols’ recent paper in Science drew new attention to the ancient issue of free will and also to the very modern method known as ‘experimental philosophy’. Experimental philosophy is liable – perhaps intended – to set the teeth of the older generation on edge, for several reasons. One is that it sounds like an attempt to smuggle into philosophy stuff that shouldn’t be there: if your conclusions can be tested experimentally they’re science, not philosophy. We don’t want real philosophy crowded out by half-baked science. It also sounds like excessive, cringing deference to those assertive scientists, as though some bullied geek started wearing football shirts and fawning on the oppressors. We may have to put up with the physicists taking our lunch money, but we don’t have to pretend we want to be like them.

Actually though, there doesn’t seem to be any harm in experimental philosophy. All the philosophy that goes by the name appears to be real philosophy, often very interesting philosophy; the experiments are not used improperly to clinch a solution but to help clarify and dramatise the problems. Often this works pretty well, and by tethering the discussion to the real world it may even help to prevent an excessive drift into abstract hair-splitting. Philosophers have always been happy to draw on the experiments of scientists as a jumping-off point for discussion, and there seems no special reason why they shouldn’t do the same with experiments of their own.

In this particular case, Nichols shows that there is something odd about people’s intuitive grasp of free will. Subjects were told to assume that determinism, the view that all events are dictated by the laws of physics, applied, and then asked whether someone would be responsible for various things. In the vaguest case they all agreed that in general, given determinism, people were not responsible for events. Given a specific example of a morally debatable act they were less sure; and when they were offered the example of a man who takes out a murder contract on his wife and children, most felt sure he was responsible even given determinism.

This is odd because it’s normally assumed that determinism means no-one can be responsible for anything. In order to be responsible, you have to have been able to do something else, and according to determinism the laws of physics say you couldn’t have done anything but what you did. It’s odder because of the distinction drawn between the cases. Where did that come from?

It could be that something in the experiment predisposed subjects to think they were required to make distinctions of this kind, or it could be that ordinary subjects are just not very good at coming up with strictly logical consequences of artificial assumptions; but I don’t think that’s really it. The distinction between the three cases appears to be a matter of who we’d blame – so it looks as if the man in the street doesn’t really grasp the philosophical concept of responsibility and relies instead on some primitive conception of blameworthiness!!!

But,  um – what is the philosophical concept of responsibility? It’s pretty clear when we cite the laws of physics that we’re talking about causal responsibility – but causal responsibility and moral responsibility don’t coincide. It’s clear that you can be causally responsible for an event without being morally responsible: someone pushed you from behind so that you in turn pushed someone under a train. Less clearly, in some cases it is held that you can be morally responsible for events you didn’t deliberately bring about: the legal doctrine strict liability, Oedipus bringing a curse on Thebes, poor Clarissa wondering whether having been raped is in itself a sin.  All of these are debatable; we might be inclined to see strict liability as a case of legal overkill: “we care so much about this that we’re not even going to entertain any discussion of responsibility – you’d better just make damn sure things are OK” . In the other cases we typically think the assignments of blame are just wrong (although Milan Kundera notably reclaimed the moral superiority of Oedipus in The Unbearable Lightness of Being). Nevertheless the distinction between moral and causal responsibility is clear enough: does the determinist case equivocate between the two, and were Nichols’ subject actually just too shrewd to be taken in?

It seems it might be so. No-one would suggest to a writer that he was not the author of his novel because it was all the result of the laws of physics, although in one sense it’s so. No-one would accept on similar grounds that I’m not responsible for a debt, however abstract and conventional the notions of debt and money may be compared with the rigorous physical account of events. So why should should the physical story stop us concluding that on another level of description we can be interestingly and coherently blameworthy?  That would be a form of compatibilism, the view that we can have our determinist cake and eat our free will, too. (I’d be a little uncomfortable leaving it there without some fundamental account of agency and morality, just as I’d be a bit unhappy to say that debt is a convention without some underpinning concept of money and economics – but that’s another discussion.) So perhaps Nichols’ subjects were compatibilists.

That would be an interesting discovery but… I hate to say this… an interesting discovery in psychology. The fact that most people are instinctively compatibilists provides no particular reason to think compatibilism is true. For that, we still have to do the philosophy the old-fashioned way. Scientists may be able to gather truth from the world, like bees with nectar: philosophers are still obliged, like spiders, to spin their webs out of their own internal resources.

 

Perplexities of Consciousness

Picture: Eric Schwitzgebel. Eric Schwitzgebel, author of the excellent Splintered Mind blog (and Professor of Philosophy at University of California at Riverside) has a new book out, Perplexities of Consciousness, which sows doubt and confusion where they have never been sown before.  Like Socrates, Schwitzgebel wants to make us wiser by showing us that we know much less than we thought. It has often been thought that while we might easily be wrong about the world and the things in it, we weren’t prone to being wrong about how the world looks to us: Schwitzgebel seeks to show that in many respects we actually suffer from unresolvable confusions about even that, and worse, in some cases we’re demonstrably wrong. Subjective experience may be a matter of there being something it is like to see red or whatever; but we actually have no clear idea of what that something is (or what it is like).

Back in 2007 Schwitzgebel published a book Describing Inner Experience? Proponent Meets Skeptic with Russell Hurlburt which examined Hurlburt’s method of Descriptive Experience Sampling (DES). In DES subjects are asked to record their inward experience at random moments (prompted by a beeper) and are subsequently put through a fairly searching interview. This earlier book, as it happens, is currently the subject of a special issue of the Journal of  Consciousness Studies.  The book’s title sets up a confrontation, but in fact it’s clear that Hurlburt has a good deal in common with Schwitzgebel: the development of a special method for clarifying inner experience implicitly concedes that error is a serious possibility. The differences seem to be partly a matter of how well DES can really work, and partly a difference of agenda, with Schwitzgebel addressing the issues at a more radical and demanding philosophical level.

There is history to all this, of course: introspectionists like Wundt and Titchener once claimed that with careful training our inner experience could be described scientifically in great detail. The catastrophic collapse of that school of thought led on to the absurd over-reaction of behaviourism, which denied the very existence of inner experience. Only now, we might say, does it seem safe for Hurlburt to venture into the scorched territory of introspection and see if with a slightly different tack, a new beginning can be made.

Schwitzgebel’s new book gives us plenty of intriguing reasons to be pessimistic about that project. The book shows its origins in a series of separate papers, but it does cohere around  central themes, finding ambiguity and conflicting testimony just where we might hope for certainty.

First off it asks: do you dream in black and white? This used to be a common question and there was apparently an era when a large proportion of people thought they did. Nowadays no-one seems to think they dream without colour and in the more distant past the question never seemed to come up (dost thou dream in woodcut or tapestry?). Schwitzgebel very plausibly suggests that the idea of monochrome dreams is tied to the prevalence of black and white films and television. Perhaps if you asked, many people would now say they dream in 2D?  I’ve had dreams that apparently took place at least partly in the world of some video game.

You may feel that the question is actually meaningless and that dreams don’t typically either have or lack colour. Perhaps they are pure narrative, and asking whether they are in black and white makes no more sense than asking whether Pride and Prejudice was written in colour. Certainly if we extend the questioning and begin to ask whether dreams are in Technicolor, or what screen format or resolution they use (or in my case perhaps whether they were Xbox or PS2) the discussion starts to seem absurd; but could we deny that people might dream in black and white, at least sometimes? It doesn’t seem too difficult to imagine that you might. It is possible that the prevalence of monochrome moving images actually changed the way people dreamed for a while; but it seems probable we must admit that some people were in fact mistaken about the qualities of their own dreams, and we must certainly accept that there was a degree of uncertainty. So Schwitzgebel has his wedge in position.

Second, he asks: do things look flat? He cites sources who say that a coin viewed at an angle looks elliptical: not to me, he objects (at least not unless I view it very obliquely, when I can sort of see it that way); if coins look elliptical to you does the world in general similarly look flat? Schwitzgebel addresses the idea that the coin in fact looks like its projection on to a flat surface and raises some objections: in fact, he claims there’s no way to make the geometry of ‘flattism’ make sense. He suspects that here again people’s intuitions have been captured by ‘technology’ – in this case people are thinking of how objects would be represented in a drawing or photograph. He remarks that some theorists have claimed that stereoscopic vision involves systematic doubling of perceived objects; while it’s true that if we focus on something far away we can see two images of a finger held close to our face, Schwitzgebel finds it a very odd idea that most of the objects in our field of vision are normally doubled (I agree – I also agree with him that the world doesn’t look all that much flatter when viewed with one eye rather than two).

Now we move on to academic, questionnaire-based research into our mental imagery, and it seems Galton is to blame for first spreading the idea that this sort of thing worked. Curiously, Galton’s research found that the scientists in his sample were predisposed to deny the existence of mental imagery altogether, while the other subjects were more likely to accept it;  a result which no-one has been able to duplicate since. Perhaps back then people thought mental imagery was an airy-fairy poetic business which hard-nosed scientists should reject.  It turns out, moreover, that there is little or no correlation between reporting vivid mental imagery and being good at tasks which apparently require mental visualisation, such as comparing rotated 3d shapes. This is odd: why would evolution give us vivid mental imagery if it doesn’t even help with tasks that require vivid mental images?  Again it seems that our own reports are all but useless as a guide to what’s really going on in there. We may claim to visualise a table, but under questioning we usually turn out to be unable to provide details, or become confused, or pause to imagine up some more details.

The next chapter raises the interesting question of human echolocation. Nagel famously took bats as his example of a creature whose inner experience we could not hope to imagine because it had a sense we entirely lacked. With amusing irony Schwitzgebel’s case here is based on the fact that we do have some echolocation after all; we’re just not normally aware of it. With a little practice we can learn how to stop short of a wall just by making regular noises and picking up the echo (if you’re going to try this at home, I recommend taking some precautions to ensure that your nose isn’t the first part of you to detect the wall unambiguously).  Some blind people are well aware of this echolocating ability, but (a real score for Schwitzgebel) they misperceive it.  Typically they describe the experience as being about pressure on the face, and nothing to do with hearing: but experiments with blocked ears and covered faces show clearly that they’re wrong about their own experiences.

I mentioned Titchener earlier: Schwitzgebel gives an interesting account of his methods. Whereas these days one would typically try to capture the subject’s impressions as fresh as possible, uncontaminated by the experimenter’s own prejudices, Titchener and his contemporaries took the opposite view: until you were very thoroughly trained in discrimination of your impressions, your testimony was worthless. Schwitzgebel explains how Titchener’s researchers were trained to pick out ‘difference tones’, illusory notes heard under certain conditions.  (You can try it out for yourself on Schwitzgebel’s own page here. There is other useful stuff on his home page including abstracts and draft chapters.)  Some of these tones are debatable and only a minority claim to be able to hear them: are the others failing to hear them, or hearing them and failing to discriminate? Titchener apparently has no answer.

The next chapter returns to earlier concerns about whether experience is sparse or abundant:  are things outside the centre of our attention still constantly present in consciousness (abundant), or do they drop out (sparse).  Schwitzgebel previously used the terminology ‘rich’ and ‘thin’, and we discussed some earlier Hurlburtian experiments of his. It will come as no surprise to find that Schwitzgebel, who quotes radically opposing views from a variety of sources, regards the matter as unresolved and possibly unresolvable; but I must say that this time round I couldn’t see any reason why we shouldn’t just conclude that experience is sometimes sparse and sometimes abundant. There’s no doubt that sometimes when we focus narrowly on one thing, we lose track of everything else; and it seems hard to deny that at times we also pay vigilant attention to our surroundings in general. Couldn’t it simply be that consciousness can have a wide or a narrow beam, at least partly under our own control?

Schwitzgebel now feels ready for a direct assault on the doctrine that our knowledge of our own experience is infallible.  He warms up by questioning whether we know what emotion is, conceding (rather dangerously?) that his wife can sometimes judge his emotional state better than he can himself. What neutral yardstick he uses to confirm his wife’s diagnosis is not altogether clear.

Why is it, he asks, that scarcely anyone, even the most vigorous sceptics, seriously questions the infallibility of introspection on certain points? The core argument seems to be that we can be wrong about the way things are, but we cannot be wrong about the way they appear to us. But why not? Schwitzgebel claims the argument rests on equivocation between two senses of  ‘appear’ , one of them epistemic as in ‘it appears to me that…’.  I don’t know whether the argument actually rests so much on the word ‘appear’ , but it seems a valid and interesting claim that there are two levels at work here: our experience and our beliefs or claims about it, with no special reason to think that the latter must be magically veridical.

Now there is a kind of rock-bottom argument available here; I don’t think I’ve ever seen it used, but it may be in the back of the minds of those who argue for infallibility. This is that eventually you get below the level where truth and falsehood apply. If you pare experience down enough, you get to a point where it just is: it’s not actually that it’s infallible, more that it’s beyond the realm of fallibility or infallibility. To be fallible, to have truth values, there has to be an element of intentionality (and the right kind, too, with an appropriate ‘direction of fit’ – I don’t think desires can be false), but at the rock bottom level, this is absent. If I just experience without any thoughts about it at all, fallibility doesn’t come into it.

I dare say Schwitzgebel might accept that up to a point, although he could reasonably point out that our experience is in practice completely suffused with a kind of intentionality; unconscious parts of our mind do an awful lot of interpretation before reports from our senses reach us and arguably pretty well everything is presented to us ‘as’ something, not just as mere sense-data (the kind of intentionality involved, that lets part of our brain add implicit messages to the conscious part about ‘that there being a table’ and so on is interesting, probably very important, and totally obscure); though generally it seems we can ‘look through’ to the basic sense-impressions if we want.

The question then is perhaps whether there is some very simple level of intentionality that can be added to the rock-bottom experience without any chance of its being wrong yet without it having the kind of trivial self-validation (‘This is the sentence I wrote’) that Schwitzgebel rightly excludes. Could it be that along with the experience itself we have an accompanying belief which says something like “Yeah, that…” which can’t really be wrong?  I still find it hard to resist the idea that there’s something infallible in there.

It’s a great point though; a strong and well-founded attack on such an important and well-accepted dogma has to be of great value. It forces us into greater clarity even if in the end it isn’t accepted.

The book winds up with consideration of another engaging and interesting issue. What do you see when you turn out the light? (I can’t tell you, but I know it’s mine, as Lennon and McCartney argued in their seminal work.)  Apparently little attention has been given to what things look like when our eyes are closed in normal light, but the book unearths quite a sequence of views about what we can see when our eyes are closed in the dark. Grey bands feature strongly in the older reports and then seem to drop out of favour: specks of light turn up fairly regularly, but you won’t be surprised to hear that there is in the end no real consensus but a quantity of misplaced confidence. I think it’s perhaps a little surprising so few people seem to say that when their eyes are closed in the dark they see nothing.  Trying it myself I find I have to sort of insist on seeing and then get some very dim lines and grids, and flashes of amorphous shapes in brighter colour. This kind of thing might well have a good explanation in neurology and the more or less random firing of isolated neurons that respond to lines, grids, or patches of colour.  At times, strangely, I had to do something hard to describe to stop my imagination intervening in what I was experiencing and livening things up.  Overall, it doesn’t look too promising a field for research to me, but, as Schwitzgebel suggests, why not see what you think (or see what you see)?

So in conclusion, what’s the verdict? I think Schwitzgebel’s case is essentially successful; his contention that there’s more to deal with here than we may have realised seems hard to deny.  This is salutary and also interesting; and since the subject is engaging and much of the discussion is readily accessible, I think the book deserves a wider readership outside the philosophical village.

I think it was Russell who said that when acting on a vigorous mind, scepticism produces new energy rather than despair: so can we add any positive conclusions to the overwhelmingly negative ones in the book?  I’m left with two main thoughts. First, there is a philosphical case to be answered in the central assault on infallibility.  Second, I think a great deal of the ambiguity and contradiction exposed by Schwitzgebel comes from the sheer complexity of the task. We ourselves, the experiencing entities, are pretty complex, with different levels of conscious and unconscious thought interacting in a variety of ways.  Second, the ways we can experience and think about things is limitlessly varied.  I don’t think it would be difficult to list fifty significantly different ways of  ‘thinking about’ a table, with and without explicit imagery.  Accordingly in many cases I think simple misunderstandings over what type of experience we’re talking about are more than half the problem. Perhaps Galton’s scientists thought ‘mental imagery’ meant what I would call ‘voluntary hallucination’; perhaps for those monochrome dreamers ‘black and white’ just meant dreams where colour wasn’t specifically salient. Taking all these problems together with a certain natural human variability, I think we might find explanations.  It wouldn’t be trained subjects we need, just better terminology and a more clearly developed common understanding.  I say ‘just’ – in fact it’s quite a tall order but not, I think, hopeless.

Mystery

Picture: shrouded in mystery.
Picture: Blandula. I’d like to say a word for mystery. I think Haldane summed it up:

…the universe is not only queerer than we suppose, but queerer than we can suppose.

Picture: Bitbucket. I hate that quote so much!  The complacent fake modesty; the characteristic Oxford attitude of mingled superiority and second-ratism:  don’t you go thinking you can apply your mind to these weighty issues, little man; the best you can do is listen reverently to the words of our mighty predecessors. Footnotes to Plato! Footnotes to Plato!

Picture: Blandula. Good grief, what a reaction! Well, then, let me quote someone you ought to like better; Leibniz:

…it must be confessed that perception and that which depends upon it are inexplicable on mechanical grounds, that is to say, by means of figures and motions. And supposing there were a machine, so constructed as to think, feel, and have perception, it might be conceived as increased in size, while keeping the same proportions, so that one might go into it as into a mill. That being so, we should, on examining its interior, find only parts which push one another, and never anything by which to explain a perception.

It’s interesting, incidentally, that Leibniz should have picked a mill. In these days of computers, it’s natural for us to talk of thinking machines, but it must have been a much less obvious metaphor then; especially a mill, which doesn’t produce any very complex behaviour… though Babbage called the central processor of the Analytical Engine the ‘mill’ didn’t he… and of course Leibniz himself designed calculating machines, so perhaps a mechanical metaphor was more natural to him than it perhaps was to his readers… Anyway! The point is, this is a good example of a recurrent theme where someone holds up for our examination a mental phenomenon – in this case perception – and holds up as it were in the other hand the physical world, and says it’s just obvious that the latter cannot account for the former.

Here’s Brentano.

Every mental phenomenon is characterised by what the Scholastics of the Middle Ages called the intentional (or mental) inexistence of an object, and what we might call, though not wholly unambiguously, reference to a content, direction towards an object (which is not to be understood here as meaning a thing), or immanent objectivity. Every mental phenomenon includes something as object within itself, although they do not all do so in the same way. In presentation something is presented, in judgement something is affirmed or denied, in love loved, in hate hated, in desire desired and so on. This intentional in-existence is characteristic exclusively of mental phenomena. No physical phenomenon exhibits anything like it. We could, therefore, define mental phenomena by saying that they are those phenomena which contain an object intentionally within themselves.

This time we’re not talking about perception as such, but about intentionality: though Brentano claims it’s characteristic of every mental phenomenon.

Then again, Nagel.

If we acknowledge that a physical theory of mind must account for the subjective character of experience, we must admit that no presently available conception gives us a clue how this could be done. The problem is unique. If mental processes are indeed physical processes, then there is something it is like, intrinsically, to undergo certain physical processes. What it is for such a thing to be the case remains a mystery.

It would be easy to find similar sources in which people contrast the pinkish-grey jelly in the skull with the sparkling abstract mental life it apparently sustains. These claims tend to have two things in common. The first is, they are essentially ostensive. There isn’t really an argument being offered at this point, more a demonstration; we’re just being shown. Look at this; then look at that;  see?

Second, the claims are emphatic: they insist. It’s obvious, they say, just look: no-one could think that this was that.  These things have nothing whatever in common.

Picture: Bitbucket. Of course they’re emphatic: they want to hurry us on past the sketchy part before we pause and ask why this shouldn’t be that. They’ve bundled the idea into its coat, thrust its hat into its hands, and are shoving it out the front door because they’re afraid that otherwise we might entertain it for a while.

Picture: Blandula. If they didn’t want us to entertain the idea, why would they write about it? No. The absence of argument here isn’t a weakness; on the contrary, it’s a demonstration of the case. The very fact that we can’t give arguments for the existence of our mental life proves its utter distinctness; we can’t prove it, we can only notice it. But, and this is the reason for the emphasis, what noticing!  ‘In your face’ doesn’t begin to get it; phenomenal experience is inside your face; it’s so emphatically there you could fairly say it defines the word.  

What I’m saying is that these claims have a special quality; simply to pay careful attention to them is itself to experience their validity. The reality and distinctness of the mental world really deserve for once that much-misused term ‘self-evident’.

Picture: Bitbucket.  It’s going to be very convenient if the absence of argument is taken to make a claim self-evident. Proving six impossible things before breakfast will be child’s play.

Actually, I think you are giving us an argument, but it’s so feeble you prefer it not to be recognised: the argument from bogglement. It runs: I can’t see how this could be that; it follows that it’s somehow cosmically distinct. The falsity of the argument, once stated is too obvious too require further comment, but let me just remind you of all those people who couldn’t imagine how the earth could possibly be moving.

Dennett has pointed out in a similar context that mysterians rely on passing off complexity as ineffability. Of course we don’t know all the details of how particular physical events in the world trigger neuronal events in the brain, let alone how that vastly complex series of functions constitutes experience. Even if we had the information, we couldn’t hold it all in mind at once. But our inability to do that, and any bogglement which may arise, does not in any way tend to show that there is no complete story.

Now Dennett would say that if only we could hold in mind all the details of the physical account, all this fantastically complex stuff, then the bogglement would vanish. But I’m not sure about that. Let me confess something: I too, boggle at the task of understanding the incomprehensible complexity of mental phenomena. But I boggle at other things too. Take computers. Now I think I can say without claiming to be an expert that I know how computers work.  I’ve played around with one or two high-level programming languages; I’ve dabbled in machine code; I’ve run up a couple of routines for imaginary Turing machines. In short, I know how it works. And yet, it still sometimes looks like magic; my mind still boggles sometimes at what I see computers doing. Now if I can get bogglement from something I understand quite well and know is a 100% physical process, it follows that my boggling at the brain and what it does shows nothing. 

Picture: Blandula. It’s not the bogglement that matters; it’s the undeniable direct experience. It’s not because we don’t understand experience that we say it’s something distinctly non-physical; it’s because we can see it’s distinctly non-physical.  To me I confess it seems to need some kind of educated perversity to deny this. Colin McGinn has pointed out that conscious experience is non-spatial: it has no position or volume.  I don’t know quite what we should say about that – abstractions like numbers are non-spatial too in what seems a different sort of way (if I mention Plato, will you start shouting again?); but it captures something about the obvious – patent – irreducibility of the mental.

I mean, just give it fair consideration; lift your eyes out of that two-dimensional world you’re cycling round and round and just notice that there’s another whole aspect to the world. To think it possible is in this case to realise it’s true, I believe.

Picture: Bitbucket. You know, you’re right.

Picture: Blandula. You see it?

Picture: Bitbucket. No, you’re right that with you there is no argument.