|A book by|
William H. Calvin
UNIVERSITY OF WASHINGTON
SEATTLE, WASHINGTON 98195-1800 USA
HOW BRAINS THINK
A Science Masters book, to be available in 12 translations (BasicBooks in the US)
copyright ©1996 by William H. Calvin
How should we approach explaining the unknown? It is well to keep overall strategy in mind,
especially whenever attractive shortcuts are offered as explanations by those whom the
philosopher Owen Flanagan terms ``the new mysterians.'' Using Dennett's epigrammatic
definition of a mystery, consider for a moment those physicists who are speculating about how
quantum mechanics might have a role in consciousness, might provide ``free will'' an escape
route from ``determinism'' via quantum mechanical processes down at the subcellular level, in
the thin microtubules that often cluster near synapses.|
I'm not going to take the space needed to do justice to their best-selling arguments (or rather the arguments of their best-selling books), but when you consider how little they actually encompass (let alone explain) of the wide range of themes involved in consciousness and intelligence (those previous two chapters), you might feel, as I do, that they're just another case of ``much ado about very little.''
Moreover, as chaos and complexity have been teaching us, determinism is really a nonissue, suitable only for cocktail party conversational gambits and hardly in need of a quantum mechanical escape clause. With some notable exceptions (I'll them ecclesiastic neuroscientists after the great Australian neurophysiologist John C. Eccles), neuroscientists seldom say similar things; indeed, we rarely engage in such word games about consciousness.
It's not for lack of interest; how the brain works is, after all, our primary preoccupation. Over our beers after a hard day at the neuroscience meetings, we tell each other that while we may not have wide-ranging explanations of consciousness yet, we do know what kinds of explanations don't work. Word games produce more heat than light, and the same is true of explanations that simply replace one mystery with another.
Neuroscientists know that a useful scientific explanation for our inner life has to explain more than just a catalog of mental capabilities. It also has to explain the characteristic errors that the consciousness physicists ignore -- the distortions of illusions, the inventiveness of hallucinations, the snares of delusions, the unreliability of memory, and our propensities to mental illnesses and seizures rarely seen in other animals. An explanation has to be consistent with many facts from the last century of brain research-- with what we know about consciousness from studies of sleep, strokes, and mental illness. We have numerous ways of ruling out otherwise attractive ideas; I've heard a lot of them in thirty years of doing brain research.
There are various angles along which to cut the cake of our mental lives. I tried focussing on
consciousness in The Cerebral Symphony. One reason that I'm going to hereafter avoid a
discussion of consciousness in favor of intelligence underpinnings is that considerations of
consciousness quickly lead to a passive observer as the end point, rather than someone who
explores, who adventures within the world. You can see that in the many ``consciousness''
connotations you'll find in a dictionary:|
Among scientists, there is a tendency to use consciousness to mean awareness and recognition; for example, Francis Crick and Christof Koch use consciousness when addressing the ``binding problem'' in object recognition and recall. But just because one word (in English) is used to denote these widely different mental facilities doesn't mean that they share the same neural mechanism. Other languages, after all, may assign one of the aforementioned ``consciousness'' connotations its own word. Crick's thalamocortical theory is most useful for thinking about object recognition but it doesn't say anything about anticipation or decisionmaking -- yet those are often among the connotations of consciousness, the word he uses. It's easy to overgeneralize, just by the words you choose. This isn't a criticism: there aren't any good choices until we understand mechanisms better.
By now, the reader might reasonably conclude that consciousness connotations are some sort of intelligence test that examines one's ability to float in ambiguity. Debates about consciousness regularly confuse these connotations with one another, the debaters acting as if they believed in the existence of a common underlying entity -- ``a little person inside the head''-- that sees all. To avoid this presumption of a common mechanism for all connotations, we can use different English words for different connotations, such as when we use ``aware'' and avoid ``conscious.'' I usually try to do this, but there are also pitfalls when you use alternative words. That's because of what might be termed back-translation.
Physicians, for example, try to avoid the C word by talking instead about the level of arousal that can be achieved with some shouting and prodding (coma, stupor, alert, or fully oriented to time and place) of the patient. That's fine, until someone tries to translate back into C-word terminology; yes, someone in a coma is unconscious, but to say that consciousness is at the opposite end of the arousability scale may be seriously misleading.
Worse, equating ``conscious'' with ``arousable'' tends to be interpreted as ascribing consciousness to any organism that can experience irritation. Since irritability is a basic property of all living tissue, plant as well as animal, this extends consciousness to almost everything except rocks; some nonscientists are already talking about plant consciousness. While this is appealing to some people and appalling to others, scientifically it is simply bad strategy (even if true). If you throw everything into the consciousness pot and mix them up, it reduces your chances of understanding consciousness.
With so many major synonyms (aware, sensitive, awake, arousable, deliberate, and more), you can see why everyone gets a little confused talking about consciousness. One often hears the word's connotation shift in the course of a single discussion; were this to happen to the word ``lift,'' with one speaker meaning what hitchhikers get and the other meaning an elevator, we'd burst out laughing. But when we talk about consciousness, we often fail to notice the shift (and debaters even exploit the ambiguity to score points or sidetrack the argument).
And there's more: at least within the cognitive neuroscience community, consciousness connotations include such aspects of mental life as the focusing of attention, vigilance, mental rehearsal, voluntary actions, subliminal priming, things you didn't know you knew, imagery, understanding, thinking, decision making, altered states of consciousness, and the development of the concept of self in children -- all of which grade over into the subconscious as well, all of which have automatic aspects that our ``narrator of consciousness'' may fail to notice.
Many people think that the narratives we tell ourselves when awake or dreaming tend to structure our consciousness. Narratives are an important part of our sense of self, and not merely in an autobiographical sense. When we play a role-- as when the four-year-old engages in make-believe, playing ``doctor'' or ``tea party''-- we must temporarily step outside of ourselves, imagine ourselves in someone else's place, and act accordingly. (The ability to do this is one of the more useful definitions of a sense of self.)
But narratives are an automatic part of everyday life in our own skins. Starting around the age of three or four, we make stories out of most things. Syntax is often a junior version of narrative: just the word ``lunch'' in a sentence sends us looking for variants of the verb ``eat,'' for the food, the place, and persons present. A verb such as ``give'' sends us searching for the three required nouns we need to fit into roles: an actor, an object given, and a recipient. There are lots of standard relationships, with familiar roles for the players, and we guess from the context what goes into any unfilled gaps. Often we guess well, but dreams illustrate the same kinds of confabulation seen in people with memory disorders, in which bad guesses are unknowingly tolerated.
``Perception,'' it has been recently said, ``may be regarded as primarily the modification of an anticipation.'' It is always an active process, conditioned by our expectations and adapted to situations. Instead of talking of seeing and knowing, we might do a little better to talk of seeing and noticing. We notice only when we look for something, and we look when our attention is aroused by some disequilibrium, a difference between our expectation and the incoming message. We cannot take in all we see in a room, but we notice if something has changed.
A sense of self is thought to go along with a fancy mental life, so let me briefly address
the common notion that self-awareness (often called self-consciousness) involves
sophisticated, ``intelligent'' mental structures.|
How do you know which muscles to move in order to mimic the action of someone else -- say, in order to stick out your tongue in response to seeing such an action? Do you have to see yourself in a mirror first, to make the association between that sight and the muscle commands that will mimic it?
No. Newborn humans can imitate the facial expressions they see, without any such experience. This suggests that innate wiring connects at least some sensory templates with their corresponding movement commands, that we're ``wired to imitate'' to some extent. Such wiring might explain why some animals can recognize themselves in a mirror, while others treat their mirror image as another animal, to be coaxed or threatened. Chimps, bonobos, and orangs can recognize themselves either immediately or within a few days' experience; gorillas, baboons, and most other primates cannot. A capuchin monkey (Cebus are the most intelligent of the New World monkeys and the best tool users) with a full-length mirror in its cage may spend weeks threatening the ``other animal.'' Ordinarily, one animal would back down after a brief period, acknowledging the other as dominant. But in the case of the mirror monkey, nothing is ever resolved; even if the capuchin tries acting submissive, so does the other animal. Eventually the monkey begins acting so depressed at the unresolved social conflict that the experimenters must remove the mirror.
What might self-recognition involve? Actions produce expectations about what sensory inflow will result from them (so-called efference copy) and so the perfect fit of these sensory predictions during small movements with the inputs from your skin and muscles would provide a way of recognizing yourself in an image. The perfect fit of the mirror image's movements with internal predictions would certainly be unusual for facial movements in most wild animals, since they rarely see their own face.
The issue of self-consciousness in the animal literature could revolve about something as simple as the attention paid to predictions about facial sensations. That's part of consciousness considerations, surely, but hardly the pivot that some would make it. Self-recognition surely involves both Horace Barlow's guessing right and Jean Piaget's sophisticated groping, but I'd put it on the list of things that intelligence isn't. Self-recognition is surely more to the point than quantum fields, however.
Do the enigmas of quantum mechanics really have something to do with such conscious aspects of our mental lives? Or is the invocation of QM in the consciousness context just another mistaken instance of suggesting that one area in which mysterious effects are thought to lurk -- chaos, self-organizing automata, chaos, fractals, economics, the weather -- might be related to another, equally mysterious one? Most such associations certainly conflate the unrelated, and when the two areas are at opposite ends of the spectrum of enigmatic phenomena, the argument is particularly suspicious.
Reducing things to basics-- the physicists' rallying cry-- is an excellent scientific strategy, as long as the basics are at an appropriate level of organization. In their reductionist enthusiasm, the consciousness physicists act as if they haven't heard of one of the broad characteristics of science: levels of explanation (frequently related to levels of mechanism). The cognitive scientist Douglas Hofstadter gives a nice example of levels when he points out that the cause of a traffic jam is not to be found within a single car or its elements. Traffic jams are an example of self-organization, more easily recognized when stop-and-go achieves an extreme form of quasi-stability-- the crystallization known as gridlock. An occasional traffic jam may be due to component failure but faulty spark plugs aren't a very illuminating level of analysis-- not when compared to merging traffic, comfortable car spacing, driver reaction times, traffic signal settings, and the failure of drivers to accelerate for hills.
The more elementary levels of explanation are largely irrelevant to traffic jams -- unless they provide useful analogies. Indeed, packing principles, surface-to-volume ratios, crystallization, chaos, and fractals are seen at multiple levels of organization. That the same principle is seen at several levels does not, however, mean that it constitutes a level-spanning mechanism: an analogy does not a mechanism make.
Quasi-stable levels make self-organization easier to spot, especially when building blocks such as crystals emerge. Since we are searching for some useful analogies to help explain our mental lives, it is worth examining how levels of explanation have functioned elsewhere. The tumult of random combinations occasionally produces a new form of organization. Some forms, such as the hexagonal cells that appear in the cooking porridge if you forget to stir it, are ephemeral. Other forms may have a ``ratchet'' that prevents backsliding once some new order is achieved. While crystals are the best known of these quasi-stable forms, molecular conformations are another, and it is even possible that there are quasi-stable forms at intermediate levels-- such as microtubule quantum states where the consciousness physicists would like the action to be.
Stratified stability refers to stacking up such quasi-stable levels. Life forms involve piling up quite a few of them; occasionally they collapse like a house of cards and the higher forms of organization dissolve (which is one way of thinking about death).
Between quantum mechanics and consciousness are perhaps a dozen of these persistent levels of organization: examples include chemical bonds, molecules and their self-organization, molecular biology, genetics, biochemistry, membranes and their ion channels, synapses and their neurotransmitters, the neuron per se, the neural circuit, columns and modules, larger-scale cortical dynamics, and so on. In neuroscience, one is always aware of these levels because of the intense rivalry between neuroscientists working at adjacent levels.
An occasional alteration in consciousness is due to widespread failures in certain types of synapses. But a more appropriate level of inquiry into consciousness is probably at an level of organization immediately subjacent to that of perception and planning: likely (in my view) cerebral-cortex circuitry and dynamical self-organization involving firing patterns within a constantly shifting quiltwork of postage-stamp-sized cortical regions. Consciousness, by any of its varied connotations, certainly isn't located down in the basement of chemistry or the subbasement of physics; I call this the Janitor's Dream ( leaping from the subbasement of quantum mechanics to the penthouse of consciousness in a single bound).
Quantum mechanics is probably essential to consciousness in about the same way as crystals were once essential to radios, or spark plugs are still essential to traffic jams. Necessary, but not sufficient. Interesting in its own right, but related only distantly to our mental lives.
Yet, because mind seems ``different'' from mere matter, many people still assume --
despite all the foregoing -- that this means some spooky stuff is needed to explain it.
But the mind should be seen as something like a crystal-- comprised of the same old
matter and energy as everything else, just temporarily organized in some complicated
way. This is hardly a new idea, witness Percy Bysshe Shelley several centuries ago:|
It has been the persuasion of an immense majority of human beings that sensibility and thought [as opposed to matter] are, in their own nature, less susceptible of division and decay, and when the body is resolved into its elements, the principle which animated it will remain perpetual and unchanged. However, it is probable that what we call thought is not an actual being, but no more than the relation between certain parts of that infinitely varied mass, of which the rest of the universe is composed, and which ceases to exist as soon as those parts change their position with respect to each other.
The traffic flow patterns in brains are far more complicated than those in vehicular movement; fortunately, there are in music some similarities that we can exploit for analogies. Understanding consciousness and intelligence will require better metaphors and actual mechanisms, not steps backward into word games or spooky stuff.
Ghosts are another version of spooky stuff, and for our analysis of creative mental life it's worth looking at what has happened to the ghost concept. Ghosts illustrate the other essential creative aspect of mind, the role of memory.
The very presence of the word ``ghost'' in most languages suggests that quite a few people have needed to talk about inexplicable things they've heard or seen. Why have so many people considered ghosts to be real? Is this where the notion of an incorporeal spirit world got started?
We now know that ghosts appear real because of mistakes made in the brain. Some are trivial, everyday mistakes and others arise from abnormalities in dreaming sleep; a few are stirred up by small epileptic seizures or the pathological processes seen in psychosis. We call them hallucinations; they involve false sounds more often than false sights. The people and pets that they feature are often scrambled a bit, just as they are in the jumble of our nighttime dreams.
Remember that the seemingly stable scene you normally ``see'' is really a mental model that you construct-- the eyes are actually darting all around, producing a retinal image as jerky as an amateur video, and some of what you thought you saw was instead filled in from memory. A hallucination merely carries this mental model to an extreme: memories stored inside your brain are interpreted as current sensory input. Sometimes this happens when you are struggling to wake up, when the paralysis of the muscles during dreaming sleep hasn't worn off as fast as usual. Dream elements appear superimposed on the image of real people walking around the bedroom. Or you might hear a dead relative speak to you with a familiar phrase. Half the brain is awake, and the rest is still dreaming away. With any luck, you realize this and don't try to place a more exotic construction on it. Each of us, after all, experiences nightly the symptoms of dementia, delusions, and hallucinations in the course of our dreaming sleep; we're accustomed to discounting such things.
Yet hallucinations can also happen when you are lying awake at night, even when you are working during the day. I suspect many of these ``ghosts'' are just simple cognitive mistakes, like one that recently happened to me: I heard a distinct crunching sound in the kitchen, which was repeated a moment later. Ah, I thought as I continued typing, the cat is finally eating her dry food. It took another two seconds before ``Oops, let's play that again.'' The cat, alas, had been dead for several months, and had had a long period of being fussy about her food. What I had faintly heard turned out to be the sound of the automatic defroster on our refrigerator -- it's somewhat more subtle than the racket made by icemakers-- and I had routinely made a guess about what the sound meant without fully considering the matter.
We are always guessing, filling in the details when something is heard faintly. A squeaking screen door, blown by the wind, may sound enough like the I-want-food whine of your dear departed dog for you to ``hear'' the dog again. Once this memory is recalled, it may be very hard to replay the actual sound you heard -- and so the fill-in of details from memory becomes the perceived reality. This isn't unusual; as William James noted a century ago, we do it all the time:
When we listen to a person speaking or [we] read a page of print, much of what we think we see or hear is supplied from our memory. We overlook misprints, imagining the right letters, though we see the wrong ones; and how little we actually hear, when we listen to speech, we realize when we go to a foreign theatre; for there what troubles us is not so much that we cannot understand what the actors say as that we cannot hear their words. The fact is that we hear quite as little under similar conditions at home, only our mind, being fuller of English verbal associations, supplies the requisite material for comprehension upon a much slighter auditory hint.This fill-in from memory is part of what's known as categorical perception; we just call it a hallucination when we are unaware of what triggered it. Unless a sound repeats, we may not be able to compare our filled-in perception of it to the original; fortunately, where visual phenomena are concerned, we can often take a second look and catch the mistake before getting committed to ``the apparition.''
We now know that suggestibility (it doesn't even take hypnosis) and stress (it doesn't even require grieving) can augment our natural tendencies to jump to conclusions, allowing memories to be interpreted as current reality. If I'd been stressed out over something, I might not have searched for an alternative explanation until it was too late to walk into the kitchen and find the sound's true source. Later, upon recalling that I'd ``heard'' the dead cat, I might have fallen into the common nonscientific explanations: ``It was a ghost!'' or ``I must be losing my mind! Maybe it's Alzheimer's!'' Both possibilities are frightening, and both are highly unlikely. But if they're the only explanations that occur to you, you may make yourself quite unhappy.
Have the scientific explanations eliminated ghosts from our culture? At least for those at the educational level of juveniles, the whole notion of ghosts remains a cheap thrill (for exactly the same reason that dinosaurs are so popular with children: they're the potent triple combination of big, scary, and safely dead). Temporal-lobe epileptics, before a physician explains their hallucinations to them, don't think ghosts are funny at all. Grieving relatives may wish, in retrospect, that someone had warned them about meaningless hallucinations.
In this case, science (for those whose education includes it) can eliminate what was once a frightening mystery. Science doesn't merely empower us, as in seeding better technologies; it also helps prevent trouble in the first place. Knowledge can be like a vaccine, immunizing you against false fears and bad moves.
There's a second neuroscience ghost story: the philosopher Gilbert Ryle's lovely
phrase ``the Ghost in the Machine'' refers to the little-person-inside manner in which
we commonly refer to the ``us'' inside our brains. It has led some researchers to talk
about the ``interface'' between ``mind'' and brain, between the unknowable and the
knowable. Is this just Descartes's pineal gland proposal dressed up in modern clothes
by the new mysterians?|
We're now making good progress in replacing such pseudospirits with better physiological analogies -- and, in some cases, with actual brain mechanisms. Just as an earlier generation of scientists usefully eliminated the external ghosts, I like to think that our currently evolving knowledge about the spirit substitutes will help people think more clearly about themselves and interpret their experiences more reliably, and will help psychiatrists to interpret the symptoms of mental illness.
The consciousness physicists, with their solution in search of a problem, surely aren't intending to tell yet another ghost story. They're just having a good time speculating, in the manner of science-fiction writers. (Still, consider how odd it would be for neuroscientists to speculate about the enigmas of physics, even those neurophysiologists -- and there are many -- who once took several courses in quantum mechanics). But why do these physicists take themselves so seriously, when a dozen levels of organization outside their own specialties? Specialization itself is perhaps part of the answer, and it demonstrates one of the hazards of intelligence.
Specialization in science is all about asking answerable questions, which requires
focusing on the details -- and that takes a lot of time and energy. None of us really
wanted to give up those wonderful debates we carried on as undergraduates about the
Big Questions. We cared about those questions. They're what attracted us to science in
the first place. They're not obsolete, like the ghosts. But the subsequent intellectual
development of working scientists sometimes reminds me of what's it like to be in a
canal lock as the water level drops.|
At least in Seattle, that's like being in a giant bathtub with a view of waterfront, fish ladders, mountains, and spectators. Once the plug is pulled, your boat sinks, and your attention is captured by the formation of the whirlpools in the lock, which are bouncing the boats about. They're fascinating. If you stick an oar in one, you can spawn off secondary whirlpools. Self-similarity theories suggest themselves, and so begins a digression into fractals.
Should you look up from your experiments and your theorizing in this oversized bathtub, the view of your surroundings will have become a rectangular patch of sky. Now you're looking out from inside a big wet box, whose walls are one or two stories high. In the patch of sunlight on the north wall of the box are some shadows of the people standing topside. As in Plato's Cave, you start to interpret the shadows on the walls, making imperfect guesses about what's really happening up there. What appears to be two people slugging each other turns out to be nothing more than one person standing in front of the other and gesturing while carrying on a conversation.
Specialization can be like that -- no more big picture, unless you come up for air occasionally and admire the scenery, see the fuller context.
The price of progress is often an unfamiliarity with other levels of organization, except for those just above or below that of your specialty. (A chemist might know biochemistry and quantum mechanics, but not much neuroanatomy.) When you've got no data but those supplied by your own mental life, it's easy to give fanciful interpretations of the shadows on the wall. Still, sometimes that's the best you can do, and Plato and Descartes did it very well in their day.
But when you can do better, why be satisfied with shadow boxing? Or continue to play word games? A word itself, one eventually realizes, is a very poor approximation to the process it represents. By the end of this thin book, the reader will, I hope, be able to imagine some neural processes that could result in consciousness -- ones that can operate rapidly enough to constitute a quick intelligence.
Describing our mental lives has a well-known hang-up, the old subjectivity snare
associated with point of view, but there are two other whirlpools we will also need to
The passive observer, poised in the middle between sensation and action, is a point of view that leads to all sorts of needless philosophical trouble. Partly, that's because sensation is only half of the loop, and we thereby ignore sensation's role in preparing for action. Some of the more elaborate couplings of sensation to action are called ``cortical reflexes'' but we also need to understand how thought is coupled to action in an intelligent manner, when we grope for a novel course of action. Ignoring the mental middle, as the behavioral psychologists did a half century ago, is not a long-term solution. What neuroscientists often do is to investigate the preparation for movement; that gets us somewhat closer to the thought process.
We often talk of our mental activities as being subdivided between sensing, thinking, and acting phases. But trouble arises because few things happen at one point in time and space. All of the interesting actions in the brain involve spatiotemporal patterns of cellular activity-- not unlike what constitutes a musical melody, where the space is the keyboard or musical scale. All our sensations are patterns spread out in time and space, such as the sensation from your fingers as you get ready to turn to the next page. So too, all our movements are spatiotemporal patterns involving the different muscles and the times at which they are activated. When you turn the page, you are activating about as many muscles as you use in playing the piano (and unless you get the timing just right, you won't be able to separate the next page from the rest). Still, we often try to understand mental events by treating them as if they actually occurred at one place and happened at one instant.
But what's in the mental middle is also a spatiotemporal pattern-- the electrical discharges of various neurons-- and we shouldn't count on it being funneled through one point in space (such as a particular neuron) and a decision made at one point in time (such as the moment when that particular neuron discharges an impulse), as if a perception or thought consisted of playing a single note, once. I know of only one such case in vertebrates (occasionally nature makes things convenient for neurophysiologists): it's an escape reflex in fish, conveniently channeled through a single large brain-stem neuron, whose discharge initiates a massive tail flip. But higher functions inevitably involve large overlapping committees of cells, whose actions are spread out in time, and that's a more difficult concept. Understanding higher intellectual function requires us to look at the brain's spatiotemporal patterns, those melodies of the cerebral cortex.
In addition to the navigational hazards, we will need to select our building blocks
with care so that we don't simply replace one mystery with another. Premature closure
is the most obvious hazard to selecting building blocks, where we stop surveying
candidate mechanisms too early, as when explaining via spirits or quantum fields.|
We must also beware of several other hazards having to do with end points of an ``explanation'': the new age everything-is-related-to-everything and the reductionistic explanations at an inappropriate level of organization (what the consciousness physicists and ecclesiastic neuroscientists do, in my not-so-humble opinion).
Explaining mental life is a big task, and you may have noticed that this is a reasonably thin book. As noted, instead of further exploring consciousness connotations, I'm going to cut the cake differently, focusing on the structures of our mental lives that are associated with intelligence. Intelligence is all about improvising, creating a wide repertoire of behaviors, ``good moves'' for various situations. A focus on intelligence covers a lot of the same ground as does a focus on consciousness -- but it avoids many of the navigational hazards. Most important, the good-moves repertoire is an end point very different from the snapshots of passive contemplation. Certainly it's easier to find a continuity between ourselves and the rest of the animal kingdom by addressing the subject of intelligence, compared to the muddle we generate when we try to talk about animal ``consciousness.'' And so the next task is to take a brief look at where good guessing might have come from, in evolutionary terms.
The paradox of consciousness -- that the more consciousness one has, the more layers of processing divide one from the world -- is, like so much else in nature, a trade-off. Progressive distancing from the external world is simply the price that is paid for knowing anything about the world at all. The deeper and broader [our] consciousness of the world becomes, the more complex the layers of processing necessary to obtain that consciousness.
You are reading HOW BRAINS THINK.|
The paperback US edition