The consciousness of sight

We have known since before the first world war that area 17, at the occipital pole of the brain, contains a map of visual space.¹ Stimulation of the brain at a given spot in the cortical map reliably excites a visual sensation at a corresponding point in space; damage will reliably extinguish vision there. The American neuroscientists David Hubel and Torsten Wiesel won the Nobel prize for physiology in 1981 for showing that some of the neurones in this map detect the edges and corners which are the elements of visual forms.

The simple picture of the striate cortex as a visual map that analyses shapes has been greatly embellished since then. Area 17 performs not one visual task but several: blobs of the visual cortex rich in the enzyme cytochrome oxidase are specialised to receive information about colour, while the areas between the blobs separate at least two further streams of information, describing motion, form, and depth. The explanation for so much parallel processing in the visual cortex was supplied by another major discovery.

Surrounding the striate cortex, and receiving the streams of visual information which flow from it, are upwards of 30 further maps of the visual world (fig 1). This comes as something of a shock, given that our everyday visual experience strikes us as being unified and orderly. Why are they there?

Summary points

• The visual world is mapped in as many as 30 visual areas beyond area V1, the primary visual cortex

• There is a division of labour among these areas: area V4 specialises in perceiving colour, V5 in perceiving motion

• Neuroscience is uncovering correlations between visual experience and cortical activity

• Some visual processes, such as blindsight, occur below the threshold of consciousness

Figure 1.

Major visual areas of monkey brain. Sulci are arrowed. 11 (middle temporal) corresponds to area V5. Adapted from Douglas et al²

Two general principles may help to explain them. Firstly, the maps subspecialise. To take the best studied example, area V4 is crucial for seeing colours, V5 for perceiving visual motion. Secondly, as visual information passes down a chain of visual maps it is gradually transformed, the activity of neurones in the later areas describing comparatively global features of a visual scene, such as the presence of a complex shape, rather than edges and corners.³ These discoveries begin to make it possible to glimpse the means by which the brain gains knowledge through vision.

Truth in illusion

The evidence coming in from the science of sight supports an ambitious claim—that every distinction drawn in our experience and behaviour will be reflected in a distinctive pattern of activity in neurones. If this is true we could predict that even the experience of illusions should be mirrored in the brain.

Multistable perceptions switch to and fro before our eyes between alternative readings, like the fish and the geese in Escher’s playful design (fig 2). There are now data to suggest that while we are looking at patterns like these the brain entertains both readings simultaneously, a critical minority of neurones switching at intervals between fish and goose to determine which we perceive.⁴

Figure 2.

Sky and Water I by M C Escher, 1938 (woodcut)

Semir Zeki and colleagues have studied the brain’s response to a picture, Enigma, by the French artist Isia Leviant, that evokes a compelling illusion of motion.⁵ Using positron emission tomography, the technique which makes it possible to see the areas of the brain activated by a given task, Zeki et al have shown that Enigma excites visual area 5—a region specialised for the perception of motion.

Subtracting consciousness

This work shows correlations between activity in the brain and the contents of visual experience. But plenty of neural activity seems to proceed without exciting any experience at all—for example, the processes which adjust the size of pupils in response to light or emotion and the mechanism which causes eyes to flick to and fro as a person looks at the view from a train. So what distinguishes conscious from unconscious activity in the brain? Can we home in more precisely on the neurology of experience?

Almost 25 years ago Lawrence Weiskrantz and colleagues began to study a patient, DB, whose right occipital cortex had been excised in the course of surgery to deal with an arteriovenous malformation.⁶ Weiskrantz had been puzzled by a discrepancy between work with monkeys, which recover reasonable vision after loss of striate cortex, and clinical experience with patients with damage to the striate cortex, who seem to remain blind. DB denied seeing anything on the left. What would happen if he were urged to guess?

To his astonishment, DB was able to point quite accurately to stimuli in his blind field, to indicate their orientation and even their shape, despite his insistence that he could see nothing. Weiskrantz christened this uncanny ability blindsight. For those pursuing the neural basis of consciousness, the interest of blindsight was clear: by subtracting the processes which subserve blindsight from those concerned with ordinary vision one might be able to isolate the activity which is crucial for visual awareness.

Blindsight shows a dramatic dissociation between awareness and performance. Since it was first described, a range of other implicit abilities akin to blindsight have been recognised. Another patient, DF, has agnosia for form after carbon monoxide poisoning: she cannot make even the simplest visual judgments about shape.⁷ Yet when asked to post a letter through a slot, she adjusts the movements of her arm and hand to the task’s demands just as you and I would. DF both can and cannot see.

What exactly has robbed patients like DB and DF of visual awareness while sparing their abilities, against all expectation, to perform these visual tasks? An evasive answer is the best available now: the loss of certain kinds of activity in certain crucial areas of the visual cortex impairs or extinguishes visual awareness, without necessarily abolishing visually guided behaviour.

In future work it would be a great help if experimenters could deliberately but temporarily reproduce phenomena such as blindsight. Recent work has begun to explore this avenue. If visual stimuli are presented briefly, for say 30 ms, they may not be consciously perceived but can none the less be shown to make an impact in the brain. A recent experiment using positron emission tomography, for example, showed that the briefly presented image of an angry face can excite the amygdala, a nucleus known to play an important part in generating fear, even if the subject is unable to report seeing all the image.⁸ Work like this, in the next few years, should help to define the neural conditions for consciousness.

Alan Cowey has provided a fascinating postscript to the story of blindsight.⁹ Remember that monkeys seem to recover quite well from the damage to striate cortex which renders humans blind. Could it be that these monkeys are in fact making excellent use of blindsight? Cowey has tried to answer this question with monkeys that had lost their striate cortex on one side. He had already shown that they could respond to visual stimuli in their blind hemifield. To this extent they could see. But in an ingenious experiment he gave them an opportunity to indicate whether a stimulus in the blind hemifield bore a closer resemblance to a light presented in their normal visual hemifield or to a blank trial in the normal hemifield during which no stimulus appeared. His monkeys unanimously reported that a stimulus in the blind field was more like a blank than a light, suggesting that, like human subjects with blindsight, they have developed a capacity to respond to visual targets which they can no longer consciously see.

Night at the end of the tunnel?

The neurobiology of sight is thriving. It even looks as if a genuine science of consciousness may be getting off the ground. Nevertheless, some sceptics reading this article will think that this science of consciousness misses the real point—how does the clockwork generate the experience of vision, the leaping light of the fireside, the shimmering of the snow? This question, the really hard one, is arguably untouched by all the work so far. It is because this question seems so terribly wrong, and yet so naggingly right, that the problem of consciousness is so difficult.¹⁰,¹¹ But as the sensible Scots philosopher David Hume was fond of observing, we sometimes just have to let our troublesome arguments be.

Figure.

Neurones are fundamental