Infinite sensation

New Scientist, 11 August 2001

EVERYONE knows there are five basic senses. But try separating them one from the other in your daily life and suddenly they don't feel so distinct.

Eat a banana, for instance, and try to taste it without smelling it and experiencing that banana-y texture on your tongue. Can you really just taste, or must you sometimes taste-smell-feel? Try talking to your lover. Listen to what is said without watching the mouth move or feeling the caress of a hand. Can you simply hear, or is there always an element of hear-see-touch? Even on the phone, can you hear a voice without imagining a face? Hard, isn't it?

The prevailing view of the brain still holds that there are five separate senses that feed into five distinct brain regions preordained to handle one and only one sense. The yellowness of the banana skin, the texture of its flesh, its smell and taste-each of these elements is parcelled up and analysed in isolation. Some theories of consciousness suggest that these dedicated brain areas somehow stamp each sense with a unique "feeling". Then, the theory goes, the brain pastes the fragments back together, calls on memory to give it a name and recall what it's for and, voila, a banana.

But perhaps it's time for a radical rethink of how the brain works. Tasks we've long assumed were handled by only one sense turn out to be the domain of two or three. And when we are deprived of a sense, the brain responds-in a matter of days or even hours-by reallocating unused capacity and turning the remaining senses to more imaginative use. All this begs the questions: are the senses really so segregated? Are they separate at all? Indeed, is it possible that our senses are continuously developing and merging so that each one of us has our own private view of the world?

It might be a big shift in thinking, but it began with a simple finding-the discovery of "multisensory" neurons. These are brain cells that react to many senses all at once instead of just to one. No one knows how many of these neurons there are-maybe they are just a rare, elite corps. But perhaps there are no true vision, hearing or touch areas dedicated to a single sense, after all. Perhaps all the neurons in our brains are multisensory-and we mistakenly label them "visual" or "auditory" simply because they prefer one sense over the others.

That's the view of Alvaro Pascual-Leone at Harvard University. He made a splash five years ago when he showed that people who were born blind use the visual cortex when they read Braille. He wondered if rather than lie idle, parts of the brain meant for seeing just started helping out with touching. His more recent work has convinced him that not only blind people but everyone has the capacity to swap senses if they have to. He thinks that the brain is much more versatile than most researchers would have us believe.

To test the idea, Pascual-Leone blindfolded healthy, sighted volunteers for five days running, taught them Braille and watched how their brains responded. He even fitted their blindfolds with photographic paper-just to be sure volunteers weren't tampering with them. Before, during and after the blindfolding, they had a series of brain scans while they were set different tactile and auditory tasks-feeling either Braille characters or brush strokes on their fingertips and listening to tones or word fragments. Before the blindfolding began, the "visual" areas were not switched on by the touching and hearing tasks. But as the week wore on the visual regions became more and more involved in routine touching and hearing.

If a person isn't seeing, Pascual-Leone found, parts of the "visual" cortex are roped in to help out in tasks involving other senses. In fact, the newly recruited regions soon become indispensable. When he tried temporarily disrupting the workings of the visual areas, using a technique called trans-cranial magnetic stimulation, or TMS, the blindfolded volunteers found it hard to read their Braille.

Taking the blindfolds off for just a day, though, was enough to undo the changes; suddenly touching and hearing tasks no longer triggered visual areas, even though volunteers were blindfolded again briefly for the scan. "Removing the blindfold and being exposed to the seeing world for 12 to 24 hours is sufficient to revert all changes induced by the five days of blindfolding," says Pascual-Leone.

What was astonishing was how quickly the brain seemed able to recruit new areas and equally effortlessly reverse that process. It was far too quick to be the result of new connections forming from scratch reasoned Pascual-Leone. "It must be assumed," he says, "that tactile and auditory input into the 'visual cortex' is present in all of us and can be unmasked if behaviourally desirable."

Pascual-Leone now feels the brain is not organised into "visual" and "auditory" and "tactile" regions at all. Instead he thinks it is split into units that have specific jobs to do or particular problems to solve-calculating distance, for example, or timing intervals. These problem-solving units simply use the best information available. Sometimes they may prefer certain senses to others, based on how suitable they are for the assigned computation, and sometimes they may use more than one, if that helps. Vision, for instance, might be the preferred way to judge distances. But if you can't see, hearing or touch can certainly fill in.

The preference of a particular problem-solving unit for a specific sense may explain the notion of sense-specific regions, he says. Just because an area tends to call on vision doesn't mean it can't process other senses, only that it may not bother if its first choice sense is on hand. This may have tricked neuroscientists into thinking that the brain is structured in parallel, segregated systems processing different types of sensory signals, says Pascual-Leone.

There is some good evidence that the brain can mix up the senses to solve particular problems. One of the main benefits of sensory integration may be better clarity and detection, says Barry Stein, at Wake Forest University in Winston-Salem, North Carolina, one of the first researchers to identify the brain's multisensory capabilities. Even weak signals should be taken seriously if they're picked up by more than one sense.

We are, for example, much more sensitive to a chemical when we combine smell and taste. Pamela Dalton, at the Monell Chemical Senses Center in Philadelphia, asked 10 people to smell benzaldehyde, a cherry-almond odour that has no taste, and to taste saccharin, a sweetener that has no smell. Before each testing session, she worked out the point where each volunteer could no longer detect each substance and prepared even weaker samples. Then she asked them to slosh the solution around in their mouths and sniff the odour at the same time. Combining taste and smell made both substances much more apparent, she found. "Ten minutes before, they hadn't been able to detect it," says Dalton.

A brain combining senses can also make better sense of ambiguous information. David Lewkowicz at the New York State Institute for Basic Research in Developmental Disabilities on Staten Island shows this nicely with a visual image of two balls moving from opposite sides of a screen, merging briefly in the centre, then continuing along their merry ways (see "Brain Games"). But when a beep sounds at the moment the two balls merge, what you see changes completely. Now, instead of passing through each other and continuing along the same trajectory, the two balls bounce off each other and return to the side they came from.

Combining hearing with vision can lead us to draw different conclusions about what we've seen too. A single flash of light, can appear to be two flashes when it coincides with two beeps, says Ladan Shams and her colleagues at Caltech in Pasadena. Even when we know there is just one flash, we can't help perceiving it as two. Apparently the brain won't let us draw contradictory conclusions from two different senses.

Increasingly, scientists are discovering that even everyday activities may actually make use of more than one sense. Consider the task of running your fingers over a pattern of raised ridges and deciding in what direction they are running. What sense do you call upon? Most of us would guess the obvious: touch. But a group at Emory University in Atlanta has demonstrated that in perfectly normal people parts of the "visual" brain are also essential for perceiving touch.

They started by scanning people's brains to see what regions were activated when they were trying to decide the orientation of some grating patterns on a touch pad. They found that a part of the brain that's involved in recognising objects by sight was active while people felt the gratings, even though they couldn't see them. "What excited us was what our subjects told us," says Krish Sathian, a lead member of the team. "When they were doing the tactile task, they were actually visualising in their mind's eye the orientation of the grating."

Did visual imagery just provide a convenient aid, or was it essential to the task? To find out, they used the TMS technique to disrupt the activity in the "visual" region the volunteers had been using. Suddenly, their volunteers could no longer tell the direction of the pattern.

The researchers concluded last year in the journal Nature (vol 401, p 587) that the "visual" cortex is closely involved in certain tactile tasks. They claimed it was the first time that visual processing was shown to be instrumental in ordinary tactile perception. But Sathian admits that the activated region may not really be visual at all. It could be a part of the brain that helps us visualise what's being touched. "We certainly can't rule out that what we're seeing is multimodal processing in an area previously thought to be just visual," he says.

Pascual-Leone's bold interpretation, that the brain is organised by task rather than by individual sense, is by no means the accepted one. Even most scientists who study multisensory processing consider it extreme. "At least some areas are exclusively unisensory," says Sathian. There's very clearly a primary visual cortex with strong inputs from the eye, he says, and a primary somatosensory cortex getting information from the body. But that's not to say that the map of the brain is static-far from it. New multisensory areas are being found all the time. "The boundaries are being pushed back," says Sathian, "just not pushed back all the way."

Those boundaries were seriously tested by an experiment that involved "rewiring" the brains of ferrets. The findings called into question the well-guarded notion that certain brain areas can only dedicate themselves to certain tasks. They suggest that, although the brain may tend to develop in a particular way, with vision processed at the back of the head and hearing on the sides, it doesn't have to be that way.

A group at MIT in Boston wanted to know how much they could override innate developmental pathways. "If we put the retina into the auditory cortex, will it see?" asks Sarah Pallas, a member of the team, now at Georgia State University in Atlanta. The researchers surgically rearranged one brain hemisphere in a handful of newborn ferrets, so that the nerves from the retina, which normally go to the visual thalamus and then on to the visual cortex, now connected to the auditory thalamus and eventually to the auditory cortex.

To their surprise, they found that the auditory cortex on the rewired side arranged itself like a visual cortex: the cells showed selectivity for orientation and motion, and they encoded a two-dimensional map of visual space. The rewired animals also seemed to behave perfectly normally. Using only the untouched hemisphere the researchers trained the animals to go to a food spout on one side of a test room if they heard a sound and one on the other if they saw a light. Amazingly, even after the visual cortex on the healthy side was completely destroyed, the animals found their way to the food.

"We were able to turn the auditory cortex into a visual cortex," says Pallas. "Maybe they couldn't recognise their grandmother with that, but they certainly could detect light." In fact, the young ferrets seemed so normal that the researchers had to mark them to tell them apart from their siblings.

The experiment revealed just how multimodal the brain may be. The amazing rewired auditory cortex was not only seeing-it was hearing at the same time, Pallas told a meeting of multisensory scientists in New York last autumn. Though the finding has not yet been published, she said that preliminary testing showed that the rewired auditory cortex was responding well to sound.

What's more, the study shows that what goes into the brain can have a lot of influence on how it's ultimately organised. Although some parts of the brain may be predisposed to become one thing or another, the rewiring shows they aren't predetermined. "Sensory inputs can influence the regional identity of the cortex," says Pallas.

But how far does this go? We can fairly assume that people deprived of sight early on will have their brains wired up differently from people who see. But what about someone who has been nearsighted since birth-could that person have a quite a different brain from someone who's experienced the world through sharper eyes? Is someone born into the high rises of Hong Kong wired up differently from a person growing up in the Gobi desert?

Pascual-Leone thinks that, both at the functional and the anatomical level, our brains are quite unique. "Blind people are not experiencing the world like a sighted person with eyes closed," he says, "but rather, they have a dramatically different world representation and hence consciousness." Indeed, maybe each of us has our own very personal take on the world, sensed by our own unique brain.

Alas, we only know how it feels to be ourselves, so it's impossible to know. And we can't ask those ferrets whether they were really seeing, or somehow hearing the light. It makes you wonder all over again about bananas-is the divine yellow fruit the very same to you as it is to me? Probably not.