麻豆传媒

THERE were no witnesses to the accident that almost killed the man known to
science as GY. 鈥淚 think I was just a kid who ran into the road, as
eight-year-olds do,鈥� he says, sandwich in one hand, strong coffee in the other.
鈥淚 just don鈥檛 recall it.鈥�

The lack of emotion in his voice is striking, but easily explained. First,
the accident happened a long time ago, 34 years to be precise. And then, truth
be told, GY鈥攖hink of him as 鈥淕raham鈥濃�攊s not entirely unhappy about
the near-fatal blow to the back of his neck that the car delivered that day.
Without it he suspects he鈥檇 鈥減robably be married off with ten kids somewhere鈥� by
now.

Instead, Graham is greatly in demand as the owner of a very special
brain鈥攁 brain that is a magnet to psychologists and the talk of
neuroscience conferences. A few weeks ago it was the turn of a team at the
University of Durham to put Graham鈥檚 brain through its paces. Next week, he is
off to Oxford, thence to Bangor in Wales, and following that, who knows: Munich,
Amsterdam, San Francisco, Prague? They all want him in their labs.

Such demand, in fact, that these days Graham usually spends no more than half
the working week pursuing his 鈥渞eal鈥� career as a psychiatric nurse. In one year
alone, he took part in 29 separate experiments, each lasting about three days.
Smart researchers know to book him several months in advance to avoid
disappointment.

For his services to science, Graham is paid expenses and compensated for loss
of earnings. But that is not the attraction. Graham does it because many of the
scientists he has worked with over the years have become friends, and because in
the end he is as interested as they are in discovering what makes his brain
special. He reads and keeps virtually every paper written about him. And that is
a lot of papers.

So, what is so special about Graham鈥檚 brain? Many relatives and friends
simply know that he has 鈥渇unny vision鈥�, Graham says. But psychologists and
neuroscientists call it blindsight. Courtesy of that blow to the head, Graham is
totally blind on the right side of his visual field: by all conventional
measures, he sees nothing to the right of his focal point with either eye. And
yet, when cued by experimenters, he can still use that blind field to accomplish
tasks you鈥檇 think of as requiring sight. Reaching out to grab an object,
discriminating between lines at different angles to each other, locating spots
of light on a screen, you name it, Graham can do it鈥攅ven if he says he
sees nothing at all. To him, it usually feels like guesswork. But clearly, even
if Graham himself is guessing, his nervous system 鈥渒nows鈥� what鈥檚 out there. In
other words, Graham sees in his blind field鈥� but unconsciously.

Actually, it鈥檚 more complicated than that. If a bright light is flashed
rapidly enough, Graham does become aware of something, usually 鈥渁 dark shadow鈥�.
And when objects or lights move fast enough, Graham experiences a strange
sensation he can only describe as something akin to pure movement鈥攎otion
stripped of form, colour or depth, although even this, he says, doesn鈥檛 really
come close to capturing it.

Not surprisingly, from neuroscientists armed with brain scanners to
philosophers and experts in artificial intelligence, many researchers treat
blindsight with great respect these days. Partly because it challenges everyday
notions about what 鈥渟eeing鈥� involves and partly because blindsight seems to
offer researchers a rare opportunity to investigate where and how in the brain
conscious perceptions are born.

Most of us think of sight as a sense that automatically involves and requires
consciousness, often of a rich and subjective kind. But blindsight researchers
know different. Having studied people like Graham for years, they know that not
all visual skills lead to or require consciousness. Quite the opposite. And more
and more of these skills are being uncovered all the time.

Graham is not the only blindsight subject courted by scientists, nor was he
the first. In fact, some of the most influential early research involved a
monkey called 鈥淗elen鈥� who lacked a segment of brain tissue thought to be vital
to normal vision but who could nevertheless find objects and reach out for food.
Among human subjects, though, Graham鈥檚 willingness to be experimented on and the
anatomical precision of the wound in his brain makes him more highly prized than
most (although, as blindsight researchers are quick to point out, nobody is
keeping any league tables).

Just after the accident, doctors feared he would die or suffer massive brain
damage. Amazingly, however, the impact of the wound was confined to nerves in
the left half of a segment of tissue, called V1, at the back of the
head鈥攖he same segment that was missing in Helen. V1 is one of the main
reception sites in the brain for signals from the retina. It also plays a key
role in normal vision鈥攈ence Graham鈥檚 right-sided blindness.

Graham knew nothing of his blindsight until the late 1970s when researchers
at Imperial College in London, led by the late Keith Ruddock, began testing him.
Then as now, many scientists had a hard time accepting that someone could
respond, often with stunning accuracy, to visual stimuli they denied being able
to see. Nor were the scientists alone in their doubt. 鈥淔or several years,鈥�
Graham recalls, 鈥淚 thought I must be cheating somehow.鈥�

But equipment that tracked the direction of his gaze proved he wasn鈥檛
sneaking a look with his good field. And today, careful experimenters don鈥檛 even
let blindsight subjects say what they can and can鈥檛 see鈥攁 subjective
account which could be open to bias. Instead, the experimenters measure
perception directly by monitoring changes in the pupils of the subjects鈥� eyes,
which contract slightly in response to visual stimuli.

But even if the authenticity of blindsight is now unassailable, researchers
are still divided over how to interpret Graham鈥檚 funny vision. What, in the end,
is blindsight really telling us about the nature of visual perception and
consciousness?

Some researchers have suggested that blindsight is little more than a weak
version of normal sight, akin perhaps to peripheral vision. According to this
view, what the subjects鈥� brains have lost is not so much the ability to produce
visual consciousness as the ability to process basic visual information. Signals
from the retina follow the usual pathways through the brain: they just seldom
gather the requisite strength.

But this explanation is rejected by Larry Weiskrantz, an Oxford psychologist
who has probably done more than anyone to raise the scientific profile of
blindsight in the past 25 years. 鈥淏lindsight is not just having weak eyesight,鈥�
he insists. After all, in one visual field Graham is aware and the other he
isn鈥檛, yet in both fields he can achieve stunning levels of performance. On some
tasks, such as detecting a pattern of light and dark stripes, he sometimes does
even better in his blind field than in his normal one. And in a detailed
analysis of Graham鈥檚 blindsight skills, Paul Azzopardi and Alan Cowey, also at
Oxford, found it impossible to simulate his signal detection abilities using a
model based on degraded normal vision.

The distinction here might seem like hair-splitting鈥攂ut it isn鈥檛. If
Graham鈥檚 blindsight is just weakened normal vision, there is no need to argue
that what is specifically lacking in his brain is visual consciousness: he
simply lacks sight. End of story.

And if he and similar subjects just lack sight, it becomes less obviously
crucial for researchers to distinguish visual awareness from basic visual
perception in their theories of vision鈥攁nd more reasonable for them to
lump consciousness in with basic perception instead and say that they鈥檙e both
produced by the same brain mechanisms (which is what a handful of philosophers
committed to 鈥渆xplaining away鈥� the problem of consciousness would prefer to
do).

Instead, says Weiskrantz, blindsight subjects do not lack the ability to
detect things such as wavelengths, but rather visual consciousness
itself鈥攖he redness of red, and so on. If Weiskrantz is right, anyone who
thinks simulating such basic perceptual skills alone on a computer will
eventually produce a conscious machine is being a tad optimistic. Instead,
something else is required. No one knows what this second ingredient is, but
Weiskrantz and others believe that brain scans of subjects like Graham can at
least provide clues.

The idea is simple. Put someone like Graham in an fMRI brain scanner鈥�
designed to look for brain function rather than structure鈥攁nd get him to
perform a visual task in his blind field. In fact, get him to do it twice, first
in his unconscious seeing mode and then in his conscious seeing mode. Subtract
the brain scans and, bingo, the result should tell you whether鈥攁nd
how鈥攂rain activity differs between vision with and vision without
awareness.

Of course, there鈥檚 more to it than that, but experiments like this are now
under way in several labs, and so far the results seem to support the idea that
aware vision is not just a 鈥渕ore intense鈥� version of unaware vision. In Graham,
for example, 鈥渃onscious seeing鈥� seems to produce a different pattern of brain
activation compared with 鈥渦nconscious seeing鈥�. There is more activity at the
front of the cortex and less in the lower regions when Graham is aware of
something in his blindfield. There is activity deep down in a midbrain
structure, the superior colliculus.

And that second finding helps to confirm the answer to a different question
of how blindsight happens in the brain. If Graham鈥檚 eyes cannot get signals to
the V1 area that is the main reception site for right field vision, how does all
that visual information guiding his blindsight 鈥済uessing鈥� get into the cortex?
The answer is along other, secondary routes that can bypass V1. And one of these
goes via the superior colliculus.

Like other researchers, Weiskrantz believes this secondary pathway is to some
extent operating in all of us, although its activity seems to be more fully
developed in people with cortical blindness. In other words, we may not realise
it, but we probably all have the pathways used in blindsight. 鈥淚t would be a
waste of effort for the brain to spend time making events conscious that don鈥檛
really require it,鈥� Weiskrantz points out. 鈥淭here are lots of times in life
where we carry out visual discriminations without any awareness at all. It鈥檚
when we鈥檙e going on automatic pilot.鈥�

Some aspects of blindsight suggest a role for it as a defence mechanism or
early-warning system. Sudden flashes and fast moving objects trigger strange
awarenesses in Graham鈥檚 blindfield, and in his unconscious seeing mode he
responds better to red, a colour that is often linked to danger. Nevertheless,
the notion that blindsight is nothing more than a primitive form of visual
perception or evolutionary throwback is beginning to look more suspect.

At the Medical Research Council鈥檚 Cognition and Brain Sciences Unit in
Cambridge, Anthony Marcel discovered that blindsight subjects not only perceive
aspects of shape, curvature and form rather better than was previously thought
in their blindfields鈥攖hey can even register whole words.

In one cunning experiment involving Graham and another subject, Marcel
flashed a word with various meanings (say, 鈥渂ank鈥�) into the good field, just
after a word relating to one of the meanings (such as 鈥渕oney鈥� or 鈥渞iver鈥�) had
been flashed in the blindfield. Even though the subjects reported no awareness
of the word in their blindfields, it biased their interpretation of the word
they could see. For the first time, comments Weiskrantz, we have an observation
linking blindsight with higher-level cognition. 鈥淭his is a major departure.鈥�

Marcel, however, believes researchers will have to build up an even fuller
picture of what blindsight subjects can鈥攁nd can鈥檛鈥攑erceive before
they can tackle the most important question of all: how the nonconscious visual
perceptions seen in blindsight relate to the conscious visual perceptions of
normal sight. Even though the known repertoire of visual discrimination that
subjects seem able to achieve without consciousness continues to expand, there
may be some visual skills鈥攕uch as the selective focusing of attention on
one of several objects鈥� that really cannot be achieved without
consciousness because they are part of the very mechanism that creates visual
consciousness.

Other limitations of blindsight are already clear. It cannot be used to
identify targets for purposive action, Marcel notes. No matter how thirsty
Graham becomes, he can鈥檛 use blindsight to grasp a glass of water unless someone
else tells him the glass is there. Nor can he create thoughts and imaginings out
of what he perceives in that field. 鈥淚 can鈥檛 replicate the sensation of an event
in my imagination,鈥� Graham says. 鈥淚 know that little about it. I can鈥檛 actually
go home and think that鈥檚 what it looked like to me.鈥�

Even so, Graham鈥檚 blindsight skills have been improving of late. In the past
few years, the levels of brightness and speed he requires in a visual stimulus
before he can consciously sense it have dropped quite sharply, possibly because
of all the experiments he does. Graham jokes that one day he would like to use
his blind field to read a novel, but he knows that is unlikely. Indeed, right
now he鈥檇 just like a straight answer to a straight question. 鈥淚f I鈥檓 so good at
discriminating wavelengths, orientations, shapes and movement in my blind
field,鈥� he asks, 鈥渨hy can鈥檛 I see in it? For 20 years, I鈥檝e been doing these
experiments, but still no one can tell me that.鈥� Perhaps one day someone will.
But first they鈥檒l have to answer another question. What is the nature of
consciousness?