Sunday, May 20, 2007

Mind Is Everywhere

It has been a while, but I used to do philosophy of mind and artificial intelligence before moving more fully into political philosophy and ethics and twiddling with other things like institutional economics. I enjoyed its puzzles, but lost a sense that the field was really speaking to problems that I thought were interesting. Part of this was a matter of what I thought were real problems, and part of it was a matter of how I thought problems should be solved. It was never clear to me, for instance, that the problems in philosophy of mind were simply problems of language, which could be cleared up and, as a consequence, clear up problems in philosophy of mind (and elsewhere). It's nice to look at some of the more recent discussion, and fascinating to learn that the problems remain roughly the same.

The central problem was and still is the relation between the materiality of the brain and the seemingly non-material stuff that constitutes consciousness or experience. We can say that there is an observable correlation between brain activity and certain kinds of experiences, and we can say that we're pretty certain that the material brain activity is causally related to experiential activity. But we really don't have the language and/or knowledge to express the materiality of conscious experience. My typing out words on the keyboard, my looking at this glowing computer screen, the birds twittering outside with a jet passing overhead, the mild temperature, the taste of coffee on my tongue, various ideas and memories flitting in and out of the forefront of consciousness, my saturating sense of dread at finishing up grading papers today, and so on comprise the experience of here-now that I'm having here-now. That experience is not experienced as synapses firing or the interaction of various bits of gray matter and electrical charges. The question is: how can matter have experiences, how can matter be conscious? This is the so-called "hard problem."

Jerry Fodor discusses Galen Strawson's latest book in "Headaches Have Themselves" in the London Review of Books. As Fodor notes, apart from standard Cartesian mind-body dualism, the main going answers to the questions posed in the paragraph above are either that consciousness is an illusion or that consciousness is an emergent feature of neural processes. Neither of these answers are adequate. So, Strawson comes up with another response. The premises of the argument are eminently reasonable, but the conclusion is to say the least rather trippy, to use a technical philosophical term. Here's Strawson's argument, as portrayed by Fodor:
There are three philosophical principles to which Strawson’s allegiance is unshakeable. The first is that the existence of consciousness (specifically, of conscious experience) is undeniable; that we are conscious is precisely what we know best. (To be sure, we can’t prove that we are conscious; but that is hardly surprising since there is no more secure premise from which such a proof could proceed.) Strawson’s second principle is a kind of monism: everything that there is is the same sort of stuff as such familiar things as tables, chairs and the bodies of animals. This, however, leaves a lot of options open since Strawson thinks that nothing much is known about that kind of stuff ‘as it is in itself’; at best science tells us only about its relational properties. What is foreclosed by Strawson’s monism is primarily the sort of ‘substance dualism’ that is frequently (but, he thinks, wrongly) attributed to Descartes...

...Strawson holds that there isn’t anything about matter in virtue of which conscious experience could arise from it; or that if there is, we have literally no idea what it could be. In particular, we can’t imagine any way of arranging small bits of unconscious stuff that would result in the consciousness of the larger bits of stuff of which they are the constituents. It’s not like liquids (Strawson’s favourite example of bona fide emergence) where we can see, more or less, how constituent molecules that aren’t liquid might be assembled to make larger things that are. How on earth, Strawson wonders, could anything of that sort explain the emergence of consciousness from matter? If it does, that’s a miracle; and Strawson doesn’t hold with those...

So, then, if everything is made of the same sort of stuff as tables and chairs (as per monism), and if at least some of the things made of that sort of stuff are conscious (there is no doubt that we are), and if there is no way of assembling stuff that isn’t conscious that produces stuff that is (there’s no emergence), it follows that the stuff that tables, chairs and the bodies of animals (and, indeed, everything else) is made of must itself be conscious. Strawson, having wrestled his angel to a draw, stands revealed as a panpsychist: basic things (protons, for example) are loci of conscious experience...

Nor, having swallowed this really enormous camel, does Strawson propose to strain at the gnats. Consider, for example: he thinks (quite rightly) that there are no experiences without subjects of experience; if there’s a pain, it must be somebody or something’s pain; somebody or something must be in it. What, then, could it be that has the experiences that panpsychists attribute to ultimate things? Nothing purely material, surely, since that would just raise the hard problem all over again. So maybe something immaterial? But monism is in force; since the constituents of tables and chairs are made of matter, so too is everything else. So, Strawson is strongly inclined to conclude, the subjects of the experiences that basic things have must be the experiences themselves. Part of the surcharge that we pay for panpsychism (not, after all, itself an immediately plausible ontology) is that we must give up on the commonsense distinction between the experience and the experiencer. At the basic level, headaches have themselves.
Whoa. Basic material things are not only the sources of consciousness, as standard materialism has it, but the loci of consciousness. Everything, since everything is material, is at least potentially the subject of experience.

Now this conclusion (again, from reasonable premises) is so counter-intuitive that it's difficult to accept. How to deal with it? Fodor himself muses that,

Typical scientific explanations appeal to natural laws. Some natural laws are explained by appealing to others, but some aren’t; some of them are basic. So, roughly, the laws about molecules explain the laws about liquidity; and the laws about atoms explain the laws about molecules; and the laws about subatomic bits and pieces explain the laws about atoms . . . and so on down, but not so on down for ever. Eventually, we get to laws about whatever the smallest things are (or, perhaps, to laws about the fundamental structure of space-time); and there we simply stop. Basic laws can’t be explained; that’s what makes them basic. There isn’t a reason why they hold, they just do. Even if basic physical laws are true of everything, they don’t explain everything; in particular, they don’t explain why, of all the basic laws that there might have been, these are the ones there actually are. I don’t say that’s the right way to look at things, but it’s a perfectly respectable and traditional way. At a minimum, it seems that the various sciences form some sort of hierarchy, with physics (or whatever) at the bottom. That’s much as one might expect if the sort of view I’m discussing is at least approximately true.

Maybe, however, there’s something wrong with this view and we’ll finally have to do without it. Maybe the hard problem shows that not all basic laws are laws of physics. Maybe it shows that some of them are laws of emergence. If that’s so, then it’s not true after all that if Y emerges from X there must be something about X in virtue of which Y emerges from it. Rather, in some cases, there wouldn’t be any way of accounting for what emerges from what. Consciousness might emerge from matter because matter is the sort of stuff from which consciousness emerges. Full stop.

It would then have turned out that the hard problem is literally intractable, and that would be pretty shocking. The idea that the basic laws are the laws about the smallest things has been central to the ‘scientific world-view’ ever since there started to be one. On the other hand, as far as I can see, it’s not any sort of a priori truth. I suppose one can imagine a world where all the big things are made out of small things, and there are laws about the small things and there are laws about the big things, but some laws of the second kind don’t derive from any laws of the first kind. In that world, it might be a basic law that when you put the right sorts of neurons together in the right sorts of way, you get a subject of consciousness. There would be no explaining why you get a subject of consciousness when you put those neurons together that way; you just do and there’s the end of it. Perhaps Strawson would say that in such a world, emergence would be a miracle; but if it would, why isn’t every basic law a miracle by definition? I have my pride. I would prefer that the hard problem should turn out to be unsolvable if the alternative is that we’re all too dumb to solve it.

Perhaps we can discover via work in artificial intelligence that when we jiggle a few bits of material things here and there - hardware and software - that something like human consciousness emerges from material stuff. Perhaps we could figure out the "laws" behind this jiggling, and perhaps we could even discover that these laws look very different from ones we're accustomed to in physics and chemistry. But we're then still faced with the basic laws of material stuff and why these laws should be this way rather than that. Further, they would simply explain the jiggling, not consciousness or experience itself. Perhaps, then the study of consciousness would be closer to ethics rather than physics in that certainty really applies to very little in ethics of anything other than a trivial nature.


Murky Thoughts said...
This comment has been removed by the author.
Murky Thoughts said...

Strawson seems not to have a handle on what a "thing" is even. Consciousness may be a thing like a wave is a thing--an abstraction of an identifiable character that travels within its medium according to some intelligible rules. The tree outside my window is an abstraction too, if by it I mean the same thing that was there last spring or back when it was a sapling. I bet many of the atoms in the chairs and tables on which Strawson rests such faith are swapping places with others in the environment, some irreversibly, resulting in oxidation and decay. Nails rust. Dinosaur bones once fossilized (dinosaur bones no longer exist, unless you mean those of birds). I thought philosophers more or less had this down after whoever it was made the "same river twice" observation. Anyway, Edelman and Koch say consciousness is nerve circuits resonating, and Ramachandran among others says that all qualia are synesthetic--a matter of anatomy. Patch a bundle of auditory nerves into the visual cortex in a deaf person's head and she will see sound. Why is the experience of color like it is? Because we do color in the visual cortex and not elsewhere, such as in the auditory cortex, or in chairs. If there ever were a hard problem of consciousness, it looks liable to be pretty straight forward from here on out. I think that's more or less the Churchland's view.

helmut said...

Maybe I'm not clear in my discussion by talking about experiences of things. Strawson doesn't appear to be saying that consciousness is a "thing." Consciousness/experience just is.

Still, I don't think you're saying anything different, MT. You're just focusing on the materialist premise (which is indeed widely assumed), rather than Strawson's conclusion. Strawson starts with that materialist premise, just like any reference to nerve circuits, synapses, anatomy. The point is that we can observe a causal relation between the physical activity (synapses) and certain kinds of experiences, feelings, etc. But consciousness/experience itself is not one of the electrical firings, experience is not experienced as flashes of neural electricity. Further, if consciousness just is a particular configuration of physical properties and relations (that a rusty nail doesn't have), then we should be able to say that an artificially replicated version is conscious or the subject of experience. In other words, conscious machines. This assumes that consciousness is a matter of computational procedure, however. While this might be possible at some point, we're pretty far from it right now, and mostly because we don't know how to go from the right physical configuration to calling something "conscious."

Consciousness/experience is something much more complex. How we get from point A (physical processes and causality) to point B (complex experience) is the mystery.

Murky Thoughts said...

if consciousness just is a particular configuration of physical properties and relations (that a rusty nail doesn't have),

I was describing a hypothetical particular person's consciousness as an abstraction defined in terms of how it changes in time, as for a wave. We can reproduce the tidal wave that hit Sumatra in a scale mock up in the lab if we agree on what the essence of the happening was. Fluid dynamics specialists and Sumatran villagers are bound to disagree, as will natural philosophers and priests about whether consciousness is reproduced by a non-biological machine--or a dog or indigenous tribe person for that matter. But to the satisfaction of a first-year physics student, you might be considered to have re-instantiated a particular wave in a new phenomenon over some new time by showing that the deflections and dimensions obey the same equation--or kind of equation. That's by convention in first-year physics. Conventionally a pile of rust is not a nail, but a nail rusts according to principles of chemistry (and stochastic environmental details) and a resulting pile of rust is indeed "identifiable" with the original nail if we choose to do so. You couldn't do the identifying from an arbitrary configuration from one point in time (perhaps not even in principle, depending how important the stochastic environmental details are). The identity is in the history, seemingly just as it is for persons and cadavers and the tree outside my window. I might have machines so incredible that I can copy my mother's mind in the evening and test out a variety of different lies will work on her in the morning, yet still some will object that I've not been experimenting on my mother or even on the person my mother was on the evening I copied her mind. I'd say that's being overly sentimental. But imagine how hard it will be to convince some people that a machine is conscious before we figure out how to copy his or her mother--which is the likely sequence of events.

I guess what we're hoping for is that neuroscientists will settle on formulas describing circuit behavior in time which seem essential to consciousness in the way the wave equations seems essential to waves. But what we're liable to get first is human-like behavior emerging from machines in which we've implemented a mix of elements from brain measurements and hypothesis, and which wires itself in response to the environment we provide, and whose operation we might not be able to get a great handle on afterwards.

Ramble ramble.

Murky Thoughts said...

O.K., I can experience the alleged conundrum better if I suppose the tree outside my window is conscious. Ignorant as I am about botany, it seems insurmountable to account for what it's like to be a tree. And yet if I imagine that I've identified some phloemal and xylemal correlates of consciousness, my inner scientist feels I'll have gotten somewhere--not that I've gotten nowhere and that I'm still trapped in physicalism. If I found twelve correlates for distinguishable behaviors by twelve branches, I'd like to think I'd be at least tempted to conclude that the vegetable was, in some essence, a coalition of consciousnesses. It would help that I'm not a tree, because in the analogous case of me being a person, I'm liable not to experience the the enjoyment that I hypothesize will derive from such information about "the nature of my existence" --correlated happenings in my vestibular system or my amygdala, for example. Is that just a matter of perspective though. I think it's both perspective as well as a pernicious bias, which comes from our false but eminently pragmatic working assumptions about ourselves and others. Regarding the coalitional tree, if I noticed also that its branches sometimes acted without external input simultaneously and faster than signals could travel from one to the other, I'd have to hypothesize a central "will" or volitional phenomenon of some kind, but there's not so much of that kind of evidence for ourselves as we'd like to imagine, and I think some that suggests the converse--volitional awareness lagging behind action. We identify with our legs even though they are optional (you don't get a name change when you get your leg amputated) and our brains are in our heads. I think we identify with "decisions" to a degree whereever in our brains they originate. To me the scientifically more natural question of consciousness is how does this lie come about, and whereabouts do we look to see it happening?

Murky Thoughts said...

By talking about a "more natural question" than the traditional "hard problem" or whatever about consciousness I mean to imply something stronger, I now realize, so I'll say it more explicitly: Prima facie the traditional problem doesn't deserve inquiry. The "burden of proof" (really a burden of motivation) falls to advocates of a psychologically and scientifically naive theory of consciousness. Western philosophy moved beyond ontological arguments for a "God" of the Church's description when the physical sciences got going, so we might expect that as the "life sciences" get up to speed philosophers would start to see once pivotal seeming questions about life as being archaic. Anselm and Augustine were in error regarding their choice of research program, and so ultimately was Descartes (hence neuroscientist Damazio's choice of title "Descartes' Error"). If physics and cosmology aren't obliged to explain "God," then biology isn't obliged to explain duality or "souls."

Anonymous said...

Thanks for this post, helmut. It ties in (vaguely, you may think) with some reading I've been doing lately on particle physics.

My thought is, crudely put: what if God created this universe randomly, not according to consistent laws, as the physicists like to think? What if their particle arrangements and group theorying break down because there is an apparent law behind all that, but it's just not consistent?

The connection of mind to matter could be the same sort of thing: it just is, and the more you try to tie it down to physical laws, the less it fits.

I suspect my idea could be put more rigorously, with God left out. (Although the picture of her laughing at all those physicists is priceless to some of us.) What if the universe doesn't in fact run on the sort of laws the physicists imagine? This would explain why they have to keep adding more and more preposterous epicycles to their theories; quantum mechanics on top of Newtonian mechanics, and string theory on top of that, just as the ancients had to keep adding epicycles to the circular (had to be, to express God's perfection, you know) orbits of the planets to describe what they saw.

Alternatively, perhaps we've just started with a wrong paradigm that makes all those physical laws not quite fit and not quite work for something like mind.

I recognize, however, that either of my ideas puts a lot of physicists and perhaps philosophers out of work.


helmut said...

You've given me a lot to ponder, MT (by the way, glad to have you back again).

I do think that the "hard problem" has always been the problem. It was originally, with Descartes, a body/mind problem (or body/soul, solved miraculously through the pineal gland). But it isn't any longer. Again, Strawson here takes the materialist premise as a given, as I do. I agree - we're not obliged to explain "souls."

But I can't dismiss the distinction between experience and neural activity either. With 100 billion neurons in a single brain and 100 trillion connections between neurons, it may very well be the case that we're simply too scientifically (and technologically) unsophisticated currently to figure out the nature of consciousness given these kinds of exponential relations. I grant that that remains an open possibility. But at the same time, this is the purely physical question once again. Consciousness is not experienced as "physical" in the same sense.

Maybe Cheryl is right and the laws of physics are, in the end, terribly limited or its "paradigm" is. Maybe we're then at the cusp of some radical thinking about the basis of science. I have no idea on this.

But I like the idea that, even if neuroscientists were able to replicate entirely the physics and processes, the question of how to describe experience or consciousness remains. I mean, we'll have to give the machines the vote, but we basically end up with an other minds problem at that point.

Anonymous said...

And how will we be able to tell if the machines are conscious?

If they tell us so?