Image
© South West News Service / Rex Features
Autistic musician Derek Paravicini performs his first professional concert at St Georges Hall, Bristol, UK
Clad in a dark suit and sunglasses, Derek Paravicini makes a beeline for the sound of my voice and links his arm into mine. "Hello, Celeste. Where have you come from today?" I reply and his response is immediate: "From Holborn?" He repeats the word several times, savouring each syllable. "Hol-born, Hol-born, Hooool-bbooorn. Where's Hoollll-booorn?" As our conversation continues, the substance of much of what I say doesn't seem to sink in, but the sounds themselves certainly do, with Paravicini lingering over and repeating particularly delightful syllables. "Meewww-zick. The pi-aan-o."

Such touching and immediate friendliness is not quite what I expected from my first meeting with the 29-year-old, blind musical savant, but his obsession with reproducing sounds certainly makes sense, given his talent. Paravicini can play just about any piece of music you request, entirely from memory, with formidable technical ability, despite having severe learning difficulties that mean he needs constant support in everyday life. And as I find out an hour later, he constantly improvises the pieces he has learned by ear, rather than simply copying as you might expect.

Paravicini is a prodigious savant - someone with a dazzling talent in one or two fields, normally music, maths, art or memory, but who also has some kind of a disability, such as autism. Psychologists have puzzled long and hard over savant skills, which confound their traditional understanding of intelligence. "What makes savants so interesting is this jarring juxtaposition of ability and disability in the same person," says Darold Treffert, a psychiatrist based in Fond Du Lac, Wisconsin who was also a consultant for the film Rain Man. "We are used to seeing skills that are consistent with each other."

But now researchers are beginning to unearth clues as to how savants' formidable brains work, and that in turn is changing our view of what it means to be a savant. In the past, savants were considered rare, solitary figures capable of mind-boggling skills that appeared as if by magic. "There have almost been suggestions that their skills appear like the birth of Venus in Botticelli - fully formed," says psychologist Richard Cowan, who studies savants at the Institute of Education, University of London.

A flurry of research published earlier this year in the journal Philosophical Transactions of the Royal Society B paints a very different picture. It turns out that these skills are far more common than previously thought. They may even arise from traits found in the general population, implying that savants are not fundamentally different from the rest of us. What's more, these skills may only blossom after years of obsessive practice, raising the question of whether many more people might cultivate similar skills, if only they had the motivation.

One of the biggest clues to the origins of savant talent lies in the fact that savants are far more common within the autistic population than among people with other mental difficulties. "When you talk about savants, you have to talk about autism," says Greg Wallace, who studies savants at the National Institute of Mental Health in Bethesda, Maryland. "In many ways they are inextricably linked."

Previously, about 1 in 10 people with autism were thought to have a special ability but in April, Patricia Howlin at the Institute of Psychiatry at King's College London found a much higher figure in the autistic adults she surveyed for savant skills or an exceptional cognitive ability. Those with an exceptional cognitive ability scored higher than the general population on at least one aspect of an intelligence test, which included arithmetic, spatial and motor skills and memory span. Savant skills included more fully developed talents, such as being able to name the elevation of both the sun and the moon at any time of day, on any specified date; being able to name the day of the week for any date in the distant past or future (a talent known as calendrical calculation) and perfect pitch. Importantly, the abilities and the skills had to be exceptional by the standards of the general population, but also well above the individual's overall level of ability. In total, roughly 30 per cent had some kind of special ability (Philosophical Transactions of the Royal Society B, vol 364, p 1359).

But while highlighting the link between autism and savantism, Howlin's results didn't offer any further clues as to why the two are so entwined. To investigate the nature of this link, Francesca Happé, also at the Institute of Psychiatry, and her colleagues decided to test which aspects of autism predispose people to talent, and whether these traits also occur in the general population, albeit less commonly.

Her team turned to a questionnaire called the Childhood Asperger Syndrome Test, which assesses social interaction, communication and whether someone has "repetitive and restrictive behaviours and interests" (or RRBIs) associated with autism. The team sent the CAST survey to the parents of more than 6000 autistic and non-autistic 8-year-olds along with a questionnaire assessing the child's special abilities in maths, musical ability, art and memory.

On analysing the results, one RRBI trait in particular emerged as the biggest indicator of talent. Children who "notice details that other people miss", or "remember details that other people miss" were twice as likely to have a special gift than those without these attributes (Journal of Child Psychology and Psychiatry, DOI: link).

Happé has concluded that this one aspect of autism rather than the whole "package", may turn out to be the "starting engine" for talent in savants. "An eye for detail is a cognitive style that runs through the normal population, and it is that quality that gives you the foothold into talent," she says.

But how could such an unremarkable trait give rise to a gift? For musical savants like Paravicini, Happé suggests that a bias towards small details might have led their developing brains to focus more on the exact notes than the overall melody, leading to perfect pitch and an exceptional musical memory. In art, a focus on small regions of a picture could lead to accurate perspective drawing.

Precocious artists

That theory fits with another study of non-autistic children with a talent for drawing that belies their age. Although these "precocious realists" don't qualify as savants because they don't have a disability, the ease, speed and accuracy of their drawing are similar to the distinctive qualities found in savant artists.

Image
© Unknown
To see if a detail-focused mindset might also explain these childrens' skills, psychologists Ellen Winner and Jennifer Drake of Boston College gave them the block design test, which assesses your ability to piece together a pattern from its component parts. Most people find this harder when they are shown an unsegmented version of the pattern versus a segmented one, but people with autism don't have this preference, demonstrating their skill at seeing a whole in terms of its parts even if there are no obvious dividing lines. "It shows they are able to do the segmentation in their minds," says Winner. The precocious realists did not have this preference either, indicating a talent for realistic drawing may arise from this isolated trait commonly found in autism (Philosophical Transactions of the Royal Society B, vol 364, p 1449).

Although these results help to pinpoint exactly what it is about autism that predisposes people to talent, it's still not clear why an eye for detail is more common in autistic people in the first place. Clues might lie in the work of Simon Baron-Cohen from the University of Cambridge, which suggests that people with autism are "hypersensitive" to sensory information. He now proposes that such acute sensitivity might predispose people with autism to pick out differences that would escape the rest of us, fostering an unusual focus on detail that leads to the development of savant skills (Philosophical Transactions of the Royal Society B, vol 364, p 1522).

But while a focus on smaller details may well be a necessary ingredient for talent, no one is yet claiming that it alone can explain the appearance of savant abilities. For instance, the ability to "toggle back and forth" between a detail-focused view and a broad view might be important to produce a good work of art. "To draw very well you have to be able to see the gestalt as well," says Winner.

Daniel Tammet, a prodigious savant who has memorised pi to 22,514 digits, believes his own talents have arisen from a special ability to connect different pieces of information together. "Savant abilities are linked to a highly associative type of thinking, an extreme form of a kind that everyone does - examples would include daydreaming, puns and the use of metaphors," he says.

There's also the question of how much practice contributes to the savant's specific talent. Savants practise tirelessly, but is that a big contributing factor, or are they born with major brain differences that predispose them to excel in their particular field?

The few studies of savant brains certainly suggest they are physically different from the average brain. For example, when Happé and Wallace studied the brain of a savant gifted at art, calendrical calculation and memory, they found his cortex was thicker in the areas associated with visuospatial processing and calculation and thinner in other regions associated with social cognition, compared with people who were neither savants nor autistic. But whether these differences were innate or grew with lifelong practice was still unclear.

The answer to that question may come from an unlikely source - a study of London taxi drivers who have acquired an encyclopedic memory of the streets of London known as "the Knowledge". Given that taxi drivers must remember the layout of 25,000 streets and the location of thousands of places of interest, and retrieve the information instantaneously, some researchers like Happé believe the Knowledge qualifies as a savant-like skill.

Eleanor Maguire and colleagues at the Institute of Neurology at University College London and colleagues found that drivers with the Knowledge have a bigger rear hippocampus than bus drivers and adults who do not drive taxis. In addition, the hippocampus appears to be larger the longer a taxi driver has been working, and shrinks once they retire (Philosophical Transactions of the Royal Society B, DOI: link). That suggests the drivers were born with typical brains which then adapt with training to accommodate the skill. If that's true of other talents, it is possible that although savants may have an unusual detail-focused cognition, their brains aren't fundamentally different at birth but become so with training.

Further preliminary evidence that many of us may have the basic brain equipment to develop such skills emerged from a calendrical savant known as GC. By scanning his brain, Cowan found that GC uses the same regions of his brain to process dates as non-savants use for mental arithmetic (Philosophical Transactions of the Royal Society B, vol 364, p 1417). In another experiment Cowan and his colleagues asked GC to learn a new calendrical system. With a limited amount of time to practise, he performed no better than a non-savant.

So if the traits that make a savant are found in the general population and practice is a big contributing factor, could we all develop a prodigious talent if we put in the hours? "I do not know any reason why not, though many would lack the motivation," says Cowan.

Motivation

But while the results of the brain scans show that savant brains may not start out with large physical differences, Happé's study - which considered both identical and fraternal twins - indicates that there is still a strong genetic component to the detail-focused mindset necessary for talent.

And it's doubtful that even those who do inherit the necessary eye for detail could unleash the phenomenal talents of someone like Paravicini through practice alone.

"I agree that many people could acquire the least creative types of savant skill, such as calendrical calculation," says Tammet. "I do not believe however that most people could learn to draw a city landscape in gorgeous detail like Stephen Wiltshire, or compose several albums of original jazz music like the American teenager Matt Savage, or create an original way of visualising numbers and language, like in my case."

In fact, it seems the remaining mystery is not so much how savants achieve their talents, but what drives them in the first place. "Motivation is a big unknown," says Wallace. "It's an enormous driving force in giftedness and in savants, but we don't know a lot about it."

One person who has something of an inside view on what contributes to savant ability is Paravicini's mentor, Adam Ockelford, a professor of music at Roehampton University in London who has watched Paravicini's talent blossom since the age of 4. When they first met, Paravicini was entirely self-taught and bashed at his plastic keyboard with his fists and elbows to reproduce the sounds he was hearing. It was only after years of practice that his technical skills developed.

But as researchers like Wallace have suggested, Paravicini seemed motivated way beyond the average music student. In fact, he seemed to be playing as if his life depended on it, and Ockelford thinks it's this that truly sets savants apart from their peers. "The survival instinct gets turned with extraordinary force into something else - in Derek's case music," says Ockelford. "When people see Derek, they think it is amazing, almost religious. But to me, it's mainly just hard work."

Gallery: Savant art: A window into their formidable brains