What’s the definition of a scientist? Someone who looks for a black cat in a
dark room. And a philosopher? Someone who looks for a black cat in a dark room
where there is no black cat.
By sheer coincidence, this drollery appeared in the horoscope section of the
Tucson Weekly newspaper just as 800 scientists and philosophers gathered for the
second biennial Tucson conference, “Towards a Science of Consciousness”. The cat
the scientists are looking for is an explanation of how processes in the brain
create conscious awareness. They haven’t captured their cat yet but occasional
sightings make them believe they will one day.
Whether the cat the philosophers are after—the so-called “hard problem”
of consciousness—exists at all is much more in doubt. If it does, the
scientists are in deep trouble. Salvos between them and the believers in the
hard problem dominated the opening day and reverberated throughout the
conference. Whoever is right, one thing is certain—consciousness remains
the first and last of the great human mysteries.
Advertisement
So what kind of problem is it? The philosophers of the hard school think that
consciousness is in a league of its own. Conscousness, they argue, has
absolutely unique properties: it is private, subjective, peculiar to the
individual, and cannot be directly observed by a third person.
As David Chalmers of the University of California, Santa Cruz, and the
hardest of the hard school of philosophers, summed it up after the last Tucson
conference: “When we see, we experience visual sensations—the felt quality
of redness, the experience of dark and light, the quality of depth in a visual
field. Other experiences go along with perception in different
modalities—the sound of a clarinet, the smell of mothballs . . . Then
there are bodily sensations from pains to orgasms—mental images that are
conjured up internally, the felt quality of emotion, and the experience of a
stream of conscious thought. What unites all of these states is that there is
something it is to be to be in them. All of them are states of experience.”
The hard school believes that understanding how the brain works does not
automatically mean we will understand consciousness. They accept that we will be
able, for example, to trace the visual processes that help us to discriminate
colour, starting with cells in the retina that respond to different wavelengths
of light. But really explaining consciousness, explaining why these neural
processes should be accompanied by a feeling of “what it is like to be me”, is a
completely different kind of problem, says Chalmers.
Indeed, he has suggested that consciousness might turn out to be an
irreducible property, in the same category as time and space, and understanding
it may force us to rewrite everything we know abut the Universe. Others think
that consciousness can be explained only by turning to a field such as quantum
mechanics, where normal laws of causality seem not to hold.
This is all nonsense for those on the other side. At Tucson, Daniel Dennett,
from Tufts University, author of Consciousness Explained was first to attack.
Facing an audience mostly sympathetic to the hard problem stance, Dennett said
he felt like “a cop at Woodstock”. But this didn’t stop him from absolutely
dismissing the hard problem.
For Dennett, there is no mysterious process required for the brain’s
information processing capabilities to become “conscious”: the brain is a kind
of hypothesis-making machine, constantly throwing up new “drafts” of what is
going on in the world.
“Mental states,” explains Dennett, “do not become conscious by entering some
special chamber in the brain, nor by being transduced into some privileged and
mysterious medium but by winning the competition against other mental states for
domination in the control of behaviour.”
Those who think that brain processes cannot explain our first-person
experience of consciousness have the question all wrong, according to Dennett.
“It presupposes that what you are is something else—in addition to all of
this brain-body activity. But what you are, however, just is the organisation of
all this competitive activity between this host of competencies which your body
has developed. You automatically know about these things going on inside your
body because if you didn’t, it wouldn’t be your body.”
Later in the week, Patricia Churchland from the Institute for Neural
Computation at La Jolla, California, weighed in with Dennett. Setting conscious
experience on a pedestal as the hard problem may be counterproductive, she
said.
“It suggests that we can already see that the hard problem is going to have
to have a real humdinger of a solution—that it’s going to have to be
really radical, that it’s going to have to come from somewhere really neat like
quantum mechanics, that it can’t just be a matter of a complex, dynamical system
doing its thing. Well, I can’t actually see that,” concluded Churchland.
Language gives us our clearest view into the consciousness of other people
through our myriad social dealings.
Some researchers have tried to peek through the same window into the minds of
other species, and many have come away with the distinct feeling that these
other animals may also be conscious. This viewpoint used to attract scorn, but
recently the evidence has become much stronger that humans are not alone in
using language and in forming abstract concepts.
“The mind of the ape cannot be that much different from our own,” says Sue
Savage-Rumbaugh from Georgia State University, who is one of the most passionate
believers that apes, at least, have a well-developed consciousness.
The first attempts to teach an ape to “speak”—in sign language, since
apes lack the vocal apparatus to speak aloud—proved disappointing, she
admits. A female chimp named Washoe was never good enough at communicating to
convince sceptics that she was actively using language.
But Washoe was only taught to speak, not to listen—a crucial omission,
says Savage-Rumbaugh. She taught a pair of chimps called Sherman and Austin
together and their language abilities burgeoned as they learnt to listen to one
another and used language to cooperate to their mutual benefit.
More recently, Savage-Rumbaugh reared a pair of bonobos, or pygmy
chimpanzees, in the company of humans who spoke English to them and pointed to
symbols on a board. While the bonobos, Kanzi and Panbanisha, never received any
explicit training in language, they picked it up anyway.
“These conditions are all that is needed for apes to acquire understanding of
language at least equal to a three-year-old child,” says Savage-Rumbaugh. For
example, the bonobos can respond correctly, even on first hearing, to new
sentences such as “Can you find the pine needles in the refrigerator?”
Kanzi and Panbanisha clearly understand even more complex concepts, says
Savage-Rumbaugh. For example, Panbanisha watched as a human secretly substituted
a bug for some sweets in a box. When a second human tried to open the box, the
first human asked the bonobo: “What is she looking for?” Panbanisha replied that
the human was looking for the sweets. “To answer a question as sophisticated as
this, Panbanisha needs a concept of what thinking is, and that other people’s
thinking is different from her own,” says Savage-Rumbaugh.
Even more strikingly, Panbanisha added that the first person was being “bad”
to play such a trick—the same comment that the researcher’s four-year-old
daughter made.
Wild bonobos that have never been exposed to human language may also use
language of a sort to communicate with one another. Bonobos hang out in the
treetops in large groups of between 60 and 100 individuals, but when they move
from roost to roost, they travel across the ground in smaller groups.
On a recent trip to Zaire, Savage-Rumbaugh noticed that bonobos had carefully
broken off plants of the same species just before and after two trails crossed,
apparently as a way of marking which fork to take for the groups following them.
Intrigued, she began to search the forest for more markings. She found 96 places
where plants had clearly been broken off by bonobos. All but a handful served as
some sort of trail-marking.
Apes undoubtedly show the clearest evidence of conscious thought among
nonhuman animals. Similar intelligence might be much harder to recognise in,
say, dolphins, simply because they are so different from us. Humans and some
apes use their hands to fashion tools, a sign of intelligence. “How do we look
for intelligence in a non-handed animal?” asks Diana Reiss of Rutgers University
in New Jersey.
On land, humans know that trail-marking is clever, but what takes its place
in the ocean? Humans can talk to apes, and the apes can sign back, but how could
we communicate with a dolphin?
Despite these difficulties, Reiss sees clear glimpses of an active
intelligence. “I often walk away thinking there’s somebody in there—or
maybe I should say, there’s some mind in there,” says Reiss.
For example, the dolphins she studies at Marine World Africa in Vallejo,
California, blow bubble rings, just as humans blow smoke rings from cigarettes,
and then play with the rings as they rise to the surface.
She has even seen them drop various items, such as bits of fish or seaweed,
into the centre of a bubble ring and watch how the turbulence buffets them. “It
looks like intelligent, goal-directed behaviour,” she says. “I felt like I was
watching a bunch of scientists testing contingencies.”
If researchers have a hard time measuring intelligence in a dolphin, they
find it still harder to crawl inside the brain of a bird. Yet here, too, at
least one researcher sees glimmerings of what may be consciousness. For twenty
years, Irene Pepperberg of the University of Arizona has studied the mental
capacities of a grey parrot called Alex, who listens to questions in English and
responds aloud with English words.
“Alex is no Einstein. We think he’s an average parrot,” says Pepperberg.
Nevertheless, Alex can count objects up to six, recognise shapes and colours,
and perform simple comparisons such as same/different and larger/smaller. Alex
can also ask for objects, and he will correct his trainer if she gives the wrong
response. If Alex says “wanna grape”, for example, and is given a piece of
banana instead, about three times out of four he will say “no”, then repeat his
request. Pepperberg won’t go so far as to claim that this behaviour shows that
Alex can consciously compare his expectations to reality, but she does believe
that “there’s certainly a `there’ there”.
What kind of a “there” might it be? Its tempting to see consciousness as an
all or nothing phenomenon but that’s a mistake. A parrot may be conscious of
what is going on around it but, to paraphrase Dennett, it probably can’t wonder
whether it’s Friday and even whether it’s a parrot.
If anyone can be considered the grandfather of the hard problem school of
consciousness it is René Descartes, born 400 years ago this year. His
meditations on the unique unity of consciousness, which convinced him that
“mind” and “body” were separate, were quoted at length at Tucson by Michael
Lockwood, from Green College, Oxford.
“When I consider the mind,” wrote Descartes, “that is myself in so far as I
am merely a conscious being. I can distinguish no part within myself. I
understand myself to be a single and complete thing. Nor can the faculties of
feeling, will, understanding and so on be called its parts, for it is one and
the same mind that wills, feels and understands.”
Descartes may have thought his consciousness was a unity, but neurologists
today would not agree. There is, they say, no more graphic evidence of the way
consciousness is “assembled” from different neuronal processes than the bizarre
way that brain injuries can tear them apart.
Perhaps most startling of all is “blindsight”, which violates our
common-sense view of consciousness. Here, damage to areas of the primary visual
cortex removes all sensation of light or colour from corresponding areas of the
visual field. Patients with this damage appear totally blind in one part of the
visual field. If asked whether they can see an object in this area, the answer,
obviously enough, is no.
But, astonishingly, if the patients are forced to guess where this object
they cannot see is located, they often point at it quite accurately. Although
they have lost all conscious sensation of “seeing”, at some level they are still
able to see. “Consciousness” and the brain’s information processing thus appear
split.
What can it be like to have blindsight? Most patients simply say they are
totally blind and cannot understand why experimenters ask them to “guess” where
objects are when it is obviously pointless. But Larry Weiskrantz from the
University of Oxford, who coined the term blindsight, has described how a few
can have a mysterious feeling of awareness under the right circumstances. “It’s
a sense that I haven’t got, if that makes sense,” was how one patient explained
it.
Blindsight is possible, the neurologists assume, because the visual pathway
splits into many parallel streams as it approaches the cortex and some streams
go on to different parts of the brain, bypassing the primary visual cortex.
Although these parts of the brain cannot create visual consciousness, they can
provide some unconscious information to guide behaviour. Blindsight can thus
give clues as to which parts of the brain are essential to generate
consciousness.
Damage even higher in the visual system or in the prefrontal
cortex—where the planning of behaviour takes place and links to motor
output are made—creates even more bizarre problems. Lesions may not so
much remove consciousness as strip away some of its attributes.
The conscious vision of patients with damage in the extrastriate cortex may
lack one or more qualities: the patients can “see” but they may not be able to
detect colour or movement. In philosophers’ jargon, they have lost one or more
of the “qualia” which populate the conscious sensory world.
Some damage within the extrastriate cortex may leave the patient able to
sense a full repertoire of qualia but destroy the ability to bind them together
to perceive a whole object. This is called aperceptive agnosia.
“Such patients,” says Petra Stoerig from the Institute of Medical Psychology
in Munich, “may have normal visual fields, normal acuity, normal brightness
discrimination, normal colour vision, normal motion processing, but still they
are unable to form an object out of these impressions.” If they are shown a
triangle, for example, they can see it but they cannot connect it with other
geometric objects such as a circle or a square. If they try to make drawings of
objects, they can produce only meaningless scribbles.
Defects elsewhere in the extrastriate cortex can rob consciousness of more of
its normal qualities. To recognise, as well as see, an object, you must be able
to create a web of associations around it by naming it and recalling things
about it. These processes are destroyed in patients suffering from associative
agnosia. They can see objects and make drawings of them perfectly well, but they
cannot recognise the objects, nor say what they do, explains Stoerig.
Even stranger is the world of people suffering from anosognosia. The
condition occasionally occurs after stroke damage to the right side of the brain
which leaves the patient paralysed on the left side of the body. Despite their
obvious paralysis, however, anosognosics claim that their useless limbs work
perfectly well.
“This has got to be the most peculiar thing I’ve ever seen in all of
neurology,” says Vilayanur Ramachandran, of the University of California, San
Diego who described his work with such patients to the conference. “Here is
somebody perfectly sane and rational, who watches her arm not performing and yet
claims she is not paralysed.”
When Ramachandran asked one patient to touch him on the nose, for example,
she insisted that she was doing so, even though her arm remained limp at her
side. When he asked her to clap, she beat the air with her good arm but said she
was clapping normally.
Another, after failing repeatedly to tie her shoe, insisted she had in fact
tied it “with both hands”—a point that normal individuals rarely bother to
mention. This is evidence that deep down, anosognosics may know the truth.
If this were a purely psychological delusion, it should apply equally to left
and right-side paralysis. But anosognosia shows up almost exclusively in people
with paralysis on the left side. This suggests that there must be specific
neurological damage to the right side of the brain, says Ramachandran.
Anosognosia is a problem of the mind’s belief system, not its perceptual
system, Ramachandran thinks.The mind needs a theory of the world in order to
organise and make sense of the constant stream of sensory inputs. But the
theory-making part of the brain must also be able to ignore inputs that don’t
fit with its world view, lest every mistaken perception shake us to our roots.
In Ramachandran’s hypothesis, this bull-headed theorist resides in the left half
of the brain.
The right half of the brain, he thinks, acts as devil’s advocate. When too
much conflicting data accumulates—for example, repeated awareness that the
left arm cannot move—the devil’s advocate overcomes the left brain’s
defence mechanisms and forces it to restructure its world view to fit the new
information.
He thinks that in people with anosognosia “that mechanism—your devil’s
advocate—is damaged, and the left brain is free to pursue a strategy of
denial and confabulation. There is no limit to the delusion.”
No longer need one spend time attempting to understand the far-fetched
speculations of physicists, nor endure the tedium of philosophers perpetually
disagreeing with each other. Consciousness is now largely a scientific
Dz.”
For those who think that neurobiology will provide the answers and the hard
problem is a philosopher’s delusion, this fighting talk from Nobel laureate
Francis Crick is just what is needed. His words, taken from an article published
two months earlier in Nature, were quoted approvingly by the
neurobiologists at Tucson. And although Crick was not at the conference, his
long-term collaborator, Christof Koch, of the California Institute of
Technology, was there to lay out the game plan.
Their first goal, hard problem or no hard problem, is to find a “neural
correlate of consciousness”—activities in the brain that correspond
specifically to the workings of conscious awareness.
The search begins by locating areas where changing neural activity can be
specifically linked to changing awareness of phenomena. To find these areas,
neurobiologists are making use of cunning experiments in which stimuli from the
external world hold constant while awareness changes—either spontaneously
or as a result of conscious activity.
Long-established work on illusions provides the most fertile hunting ground
for such effects. When we look at the famous “vase” illusion, after a little
while, we alternately see the vase and two faces. The stimulus does not change
but what we see in our mind’s eye does.
The most complete experiments of this sort came from Nikos Logothetis of
Baylor College of Medicine at Houston. “Like Crick and Koch, we wondered if
there were any neurons that are specifically related to the act of perceiving,”
said Logothetis. His experiments on monkeys used visual illusions that can be
generated by a phenomenon called binocular rivalry.
In Logothetis’s first experiments, a set of stripes slanting one way was
shown to one eye and an identical set slanting in the opposite direction was
shown to the other eye. After a short while, the two alternate irregularly
—just like the vase and the two faces—as the incompatible inputs
from the two eyes battle with each other: the stimuli do not change but what the
monkey “sees” does.
Logothetis trained his monkeys to press a bar according to which way the
stripes appeared to be oriented. At the same time, he made electrical recordings
from numerous places at different levels along the visual pathways.
The firing pattern of many cells remained constant whatever the monkey
reported seeing—the neurons were locked into the unchanging stimuli
presented to each eye. But the behaviour of some neurons correlated very closely
with the monkey’s awareness. Their firing pattern changed dramatically just
before the monkey switched bars to report that the lines were changing from one
orientation to another.
Many of these neurons were found in an area known as V4, at the top of the
hierarchy of the visual cortex. This location fits well with the view of
consciousness put forward by Crick and Koch. They believe that you cannot be
consciously aware of the information processing that goes on in the lower parts
of the visual system, from the retina up to the primary visual cortex called V1.
Consciousness is related to the high-level, “explicit”, representations
generated at the top of the visual cortex.
An explicit representation is seen in cells that respond only to a complex
property of an object, rather than to dots or lines or patches of brightness.
The best known of the “explicit” neurons are those found in the mid-temporal
cortex that respond only to faces seen from a particular angle. Damage to this
area causes prosopagnosia, an inability to recognise familiar faces.
To generate a full conscious experience, Crick and Koch postulate that cells
which code explicitly for a face, for example, must somehow link up to many
other neurons that relate to them—perhaps to the name of the person,
memories involving the person and so on. They must also link to the motor cortex
so that the experience can generate a response.
How all this happens is anyone’s guess right now, but we would expect
consciousness to arise in neurons linking the highest parts of the visual system
with the prefrontal cortex which contains the language centres and areas
involved in planning action.
As Koch explains: “Naively put, neurons in the visual part of the brain
project forwards to the prefrontal [cortex], and the prefrontal looks back at
the high-level visual output. That interaction is where we believe the neural
correlate of consciousness arises.”
At Tucson, Logothetis reported new binocular rivalry experiments with monkeys
pitting pictures of faces against pictures of objects while recording from face
recognition neurons. Once again, he found some cells that started firing just
before the monkey pulled a lever to say it was now seeing a face.
But when he was asked if he thought he had found possible candidates for the
neural correlate of consciousness, Logothetis was cautious: “When there is a
pattern that is consistently happening somewhere, you still don’t know if it is
a cause or an effect.”
Finding neurons that appear to track visual awareness is still just the first
step to pinning down the neural correlate of consciousness. The key step now is
to find out what these neurons are connected to and how they respond.
At least the hunt has begun and more results can be expected as researchers
turn to humans. Excitement was high at the possibilities offered by functional
magnetic resonance imaging —the most sophisticated of all of the new brain
scanners.
Roger Tootell of the Massachusetts General Hospital’s Nuclear Magnetic
Resonance Center is one of the pioneers with his work using the waterfall
illusion: if you stare at a waterfall or anything moving continuously and then
look away, stationary objects seem to stream by in the opposite direction. Using
a brain scanner to study subjects while they experienced the illusion, Tootell
was able to map the parts of the brain that changed as the perception faded
away. Once again, parts of the mid-temporal cortex appeared critical.
The real fun will come when these imaging techniques are applied in cases
where visual awareness has been split off from information processing in the
brain. Weiskrantz has already experimented with one of his blindsight subjects
who develops some degree of “awareness” of the stimulus if it has enough
contrast. His experiments were designed to see what changes in the brain as the
subject shifts between the “aware” state and the “unaware” state. Results are
expected any day.
“As far as consciousness is concerned, one is in a powerful position if one
can compare aware and unaware modes,” says Weiskrantz. “We may be able to sneak
up on the process of the neural basis of awareness.”
There’s just one small snag, however: blindsight patients are hard to find.
Instead, researchers may be able to turn to “induced blindsight”. Late last
year, Christopher Kolb and Jochen Braun from Koch’s group at Caltech reported an
experiment like that described in the figure below. Such displays, ” too
terrible to behold” as Braun described them, leaves people able to use vision
while removing conscious awareness of what they are “seeing”.
The patterns appear to interfere with visual processing in the primary visual
cortex in much the same way as lesions in this region interfere with processing
in people with true blindsight. These experiments will make it possible to use
MRI to study neural changes as the brain shifts between consciously “aware” and
“unaware” states.
Even “hard man” Chalmers was impressed by the new possibilities. Induced
blindsight work is most promising since it deals with normal subjects. That’s
going to explode in the next few years,” he predicts.
Koch’s own vision of the future of neurobiology was even more euphoric. If,
as he thinks, there are very specific neurons that have to be activated to
encode the specific content of visual awareness, then they must have something
unique about them.
“If that is true, then by definition there has to be a set of genes that
codes for them and that means that at some point you’ll be able to get an
antibody or set of antibodies for the neurons that are your correlate of visual
awareness. This will then let you use the power of molecular biology to make
incredible progress. You can then stain these neurons, you can maybe transiently
inactivate them and see what happens.”
It might even be possible to create a “zombie”, a creature that has
everything but awareness, thereby showing by default just what consciousness
gives us
Not everyone wants to see the “hard problem” solved. Unlike the so-called
“easy problems”—explaining how the brain carries out its various
information processing tasks from analysing the colour of objects to processing
a stream of words—solving the hard problem means understanding phenomenal
consciousness.
Susan Greenfield of the University of Oxford described a theory by which the
fleeting recruitment of populations of neurons could be linked to the level of
consciousness. But the theory would not necessarily grasp how it feels to be a
specific individual with a unique, private view of the world. All anyone would
be able to point to, she said, was an increasingly refined correlation between
the behaviour of a group of neurons and some measure of consiousness. That would
not solve the hard problem.
Greenfield was happy with that. “If we really and deeply knew how groups of
neurons generated consciousness, then we couldn’t exclude the possiblity that we
could hack into each other’s consciousness,” she said.”If we did that, then we’d
annihilate the individual, and I for one would not want to see that day.”
Two big ideas emerged at the first Tucson conference in 1994 that have caused
a stir in the consciousness community ever since: one was the “hard problem”, as
championed by Chalmers, which is very much alive.
The other is the brainchild of Roger Penrose, a mathematician from Oxford
University, and anaesthesiologist Stuart Hameroff of the University of Arizona.
Consciousness, they claim, arises from quantum-mechanical processes taking place
within tubes of protein inside nerve cells (see “Quantum states of mind”, New
Scientist, 20 August 1994, p35).
On the face of it, quantum mechanics is tremendously seductive. Quantum
processes operate without cause and effect, a very appealing notion since it
leaves room for free will and spontaneity.
Penrose also argues that human minds do things that networks of nerve cells
and the computers modelled on them can never do: “Understanding is a quality, I
claim, that cannot be captured in any form of computation whatever.” The
unpredictability of quantum events provides a noncomputable way for
understanding to arise in the brain, he argues.
Despite their appeal, however, quantum processes take place at atomic or
subatomic scales and in the merest whisper of a microsecond—far too small
and fast, seemingly, to affect nerve cells.
But at the first Tucson conference, Penrose and Hameroff proposed that
microtubules—cylindrical tubes of protein molecules called tubulin, which
form the internal skeleton of cells—might provide a safe haven in which
quantum events could multiply until they became powerful enough to make a
difference.
What are these events? Quantum theory maintains that an electron, for
example, has no location at any particular time until some later event requires
it to have one. Until then, the electron could be anywhere, and its position is
described by a probability function.
Similarly, the outcome of any event at the quantum level is not determined
until some later event demands it. In a new twist, Penrose and Hameroff
suggested that vast armies of tubulin molecules may suddenly and spontaneously
resolve their quantum uncertainties. Each time this happens, we have an
experience, they said.
The audience was somewhat less than overwhelmed, however, and a show of hands
indicated that most of them remained sceptical, in large part, because no one
has found any experimental evidence that anything like this is actually going
on.
“If we’re going to see the quantum approach flower, what we need is not just
a matching of equations. What we need is some good experimental evidence,” says
Churchland, who is one of the tartest critics of what she calls “Penrose’s
ٴǾԲ”.
Moreover, says Churchland, “even if the theory is true, how does that explain
the phenomenon at issue? I haven’t seen the slightest explanation of what all
that might have to do with consciousness.”
Of course, that’s pretty much what Chalmers says about the
neuroscientists …
It’s easy to see in each of these two images that there is a small square
(here towards the upper left corner) which stands out because it contains short
lines running at right angles to those in the surroundings. But if the two
images are shown simultaneously to the right and left eye for a very short time
the square vanishes. The images appear to fuse in the brain: adding them
together will obviously superimpose a line slanting to the left on every line
slanting to the right so that no overall area of discrepancy remains. The square
cannot now be seen but if it is moved from one corner of the figure to another,
subjects can still reliably report where it is. They can thus “see” the square
but cannot consciously perceive it—the phenomenon of induced
blindsight.
