Synaesthesia and savantism

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

“We only use 10% of our brain”. I don’t know where that idea originated but it certainly took off as a popular meme – taxi drivers seem particularly taken with it. It’s rubbish of course – you use more than that just to see. But it captures an idea that we humans have untapped intellectual potential – that in each of us individually, or at least in humans in general lies the potential for genius.

Part of what has fed into that idea is the existence of so-called “savants” – people who have some isolated area of special intellectual ability far beyond most other individuals. Common examples of savant abilities include prodigious mental calculations, calendar calculations and remarkable feats of memory. These can arise due to brain injuries, or be apparently congenital. In congenital cases, savant abilities are often encountered against a background of the general intellectual, social or communicative symptoms of autism. (The portrayal by Dustin Hoffman in Rain Man is a good example, based on the late, well known savant Kim Peek).

A new hypothesis proposes that savantism arises due to a combination of autism and another condition, synaesthesia. Synaesthesia is commonly thought of as a cross-sensory phenomenon, where, for example, different sounds will induce the experience of particular colours, or tastes will induce the tactile experience of a shape. But in most cases the stimuli that induce synaesthesia are not sensory, but conceptual categories of learned objects, such as letters, numbers, days of the week, months of the year. The most common types involve coloured letters or numbers and what are called mental “number forms”.

These go beyond the typical mental number line that most of us can visualise from early textbooks. They are detailed, stable and idiosyncratic forms in space around the person, where each number occupies a specific position. They may follow complicated trajectories through space, even wrapping around the individual’s body in some cases. These forms can be related to different reference points (body, head or gaze-oriented) and can sometimes be mentally manipulated by synaesthetes to examine them more closely at specific positions.

The suggestion in relation to savantism is that such forms enable arithmetical calculations to be carried out in some kind of spatial, intuitive way that is distinct from the normal operations of formal arithmetic – but only when the brain is wired in such a way to take advantage of these special reprepsentations of numbers, as apparently can arise due to autism.

It has been proposed that the intense and narrowly focused interests typical of autism can lead to prolonged practice of these skills, which thus emerge and improve over time. While certainly likely to be involved in the development of these skills, on its own this explanation seems insufficient. It seems more likely that these special abilities arise from more fundamental differences in the way the brains of autistic people process information, with a greater degree of processing of local detail, paralleled by greater local connectivity in neural circuits and reductions in long-range integration.

Local processing may normally be actively inhibited. This idea has been referred to as the tyranny of the frontal lobes (especially of the left hemisphere), which impart top-down expectations with such authority that they override lower areas, conscripting them into service for the greater good. The potential of the local elements to process detailed information is thus superseded in order to achieve optimal global performance. The idea that local processing is actively suppressed is supported by the fact that savant abilities can sometimes emerge after frontal lobe injuries or in cases of frontotemporal dementia. Increased skills in numerical estimation can also, apparently, be induced in healthy people by using transcranial magnetic stimulation to temporarily inactivate part of the left hemisphere.

This kind of focus on local details, combined with an exceptional memory, may explain many types of savant skills, including musical and artistic ones. As many as 10% of autistics show some savant ability. These “islands of genius” (including things like perfect pitch, for example) are typically remarkable only on the background of general impairment – they would be less remarkable in the general population. Really prodigious savants are much more rare – these are people who can do things outside the range of normal abilities, such as phenomenal mathematical calculations. In these cases, the increased local processing typical of autism may not be, by itself, sufficient to explain the supranormal ability.

The idea is that such prodigious calculations may also rely on the concrete visual representations of numbers found in some types of synaesthesia. This theory was originally proposed by Simon Baron-Cohen and colleagues and arose from case studies of individual savants, including Daniel Tammett, an extraordinary man who has both Asperger’s syndrome and synaesthesia.

I had the pleasure of speaking with Daniel recently about his particular talents on the FutureProof radio programme for Dublin’s Newstalk Radio. (The podcast, from Nov 27th, 2010, can be accessed, with some perseverance, here). Daniel is unique in many ways. He has the prodigious mental talents of many savants, for arithmetic calculations and memory, but also has the insight and communicative skills to describe what is going on in his head. It is these descriptions that have fueled the idea that the mental calculations he performs rely on his synaesthetic number forms.

Daniel experiences numbers very differently from most people. He sees numbers in his mind’s eye as occupying specific positions in space. They also have characteristic colours, textures, movement, sounds and, importantly, shapes. Sequences of numbers form “landscapes in his mind”. This is vividly portrayed in the excellent BBC documentary “The Boy With the Incredible Brain” and described by Daniel in his two books, “Born on a Blue Day” and “Embracing the Wide Sky”.

His synaesthetic experiences of numbers are an intrinsic part of his arithmetical abilities. (I say arithmetical, as opposed to mathematical, because his abilities seem to be limited to prodigious mental calculations, as opposed to a talent for advanced calculus or other areas of mathematics). Daniel describes doing these calculations by some kind of mental spatial manipulation of the shapes of numbers and their positions in space. When he is performing these calculations he often seems to be tracing shapes with his fingers. He is, however, hard pressed to define this process exactly – it seems more like his brain does the calculation and he reads off the answer, apparently deducing the value based at least partly on the shape of the resultant number.

Daniel is also the European record holder for rembering the digits of the number pi – to over 20,000 decimal places. This feat also takes advantage of the way that he visualises numbers – he describes moving along a landscape of the digits of pi, which he sees in his mind’s eye and which enables him to recall each digit in sequence. The possible generality of this single case study is bolstered by reports of other savants, who similarly utilise visuospatial forms in their calculations and who report that they simply “see” the correct answer (see review by Murray).

Additional evidence to support the idea comes from studies testing whether the concrete and multimodal representations of numbers or units of time are associated with enhanced cognitive abilities in synaesthetes who are not autistic. Several recent studies suggest this is indeed the case.

Many synaesthetes say that having particular colours or spatial positions for letters and numbers helps them remember names, phone numbers, dates, etc. Ward and colleagues have tested whether these anecdotal reports would translate into better performance on memory tasks and found that they do. Synaesthetes did show better than average memory, but importantly, only for those items which were part of their synaesthetic experience. Their general memory was no better than non-synaesthete controls. Similarly, Simner and colleagues have found that synaesthetes with spatial forms for time units perform better on visuospatial tasks such as mental rotation of 3D objects.

Synaesthesia and autism are believed to occur independently and, as each only occurs in a small percentage of people, the joint occurrence is very rare. Of course, it remains possible that, even though most people with synaesthesia do not have autism and vice versa, their co-occurrence in some cases may reflect a single cause. Further research will be required to determine definitively the possible relationship between these conditions. For now, the research described above, especially the first-person accounts of Daniel Tammett and others, gives a unique insight into the rich variety of human experience, including fundamental differences in perception and cognitive style.

Murray, A. (2010). Can the existence of highly accessible concrete representations explain savant skills? Some insights from synaesthesia Medical Hypotheses, 74 (6), 1006-1012 DOI: 10.1016/j.mehy.2010.01.014

Bor, D., Billington, J., & Baron-Cohen, S. (2008). Savant Memory for Digits in a Case of Synaesthesia and Asperger Syndrome is Related to Hyperactivity in the Lateral Prefrontal Cortex Neurocase, 13 (5), 311-319 DOI: 10.1080/13554790701844945

Simner, J., Mayo, N., & Spiller, M. (2009). A foundation for savantism? Visuo-spatial synaesthetes present with cognitive benefits Cortex, 45 (10), 1246-1260 DOI: 10.1016/j.cortex.2009.07.007

Yaro, C., & Ward, J. (2007). Searching for Shereshevskii: What is superior about the memory of synaesthetes? The Quarterly Journal of Experimental Psychology, 60 (5), 681-695 DOI: 10.1080/17470210600785208


  1. [...] Synaesthesia and savantism and Where do morals come from?. The second is a review of Braintrust: What Neuroscience Tells Us about Morality by Kevin Mitchell [...]

  2. [...] Synaesthesia and savantism and Where do morals come from?. The second is a review of Braintrust: What Neuroscience Tells Us about Morality by Kevin Mitchell [...]

  3. I find it strange that we don’t all have savant-like mental calculation abilities. After all, these are simple operations performed by computers easily. Compared to what else the brain can do, real number arithmetic is not very impressive. Why should one need major brain abnormalities to be able to perform such things efficiently? Is suppression of local processing sufficient to explain the generally awful arithmetical ability of humanity?

  4. That’s an interesting question, though it is always dangerous to try and say why something is the way it is. One answer may be simply that our brains are not wired like computers – we are very good at doing some types of things that computers are awful at and vice versa. Our brains are optimised to perform the functions that ensure the survival of the organism and transmission of their genes. It presumably is not particularly adaptive to be able to multiply 36,452 times itself very quickly or to be able to figure out that October 10th, 1970 was a Tuesday.

    What I find amazing is that anyone can do those things and that a bang on the head might release those latent abilities! Of course, it might be possible that only some people have these latent abilities, since bangs on the head do NOT usually lead to savantism. You might, for example, be able to liberally bash me about the head without any noticeable improvement in my mathematical abilities!

  5. Jimi Hendrix was a synaesthete, he associated colors with musical notes, it helped to make him a musical genius. A great many of his songs reference to colors, even as a child he was entranced with a relationship between colors, music, and emotions that the rest of don’t perceive.

  6. I’ve read Daniel Tammett book, “Born on a Blue Day” and it was a fascinating read.

    I also saw a TV documentary of him doing mental arithmetic, and noticed that his index finger was moving back and forth, up and down, as if he were moving something in rows. A year or more later I saw a documentary on Japanese kids who had been trained on the abacus for years, and then the top ones were selected for a special academy that teaches a method called Soroban of doing advanced calculations using very large abacuses, with many rows. The kids train for hours a day, and like playing piano, when they have to do a calculation they have learned to move so many beads back and forth and read off the rows – so much so that they eventually can do the calculations without the physical abacus being present. They do still move their fingers rapidly, as if manipulating the beads though. I though bingo – that’s what Daniel Tammett is doing too!

    IMO, the calendrical calculation however is based on everyone’s ability to measure elapsed time – only taken to an extreme – and is different in nature than calculating square roots and such.

  7. kjmtchl, I agree that our brains would not have been evolutionarily optimized to perform such calculations; that must be part of the answer. However, that only pushes the question back to what our brains were optimized for that makes them so bad at arithmetic. Were they optimized for dealing with the ‘high-level’ concepts of everyday language? Perhaps trying to force arithmetic into this architecture is extremely inefficient, and those savants are able to use an efficient low-level architecture that is blocked off for the rest of us. This could explain why arithmetic ability correlates with general intelligence differently in normal people and in savants, due to good arithmetic ability in normal people signifying fast and reliable processing of the high-level concepts which make up most of human thought which is orthogonal to savant ability.

  8. pconroy, thanks for those comments. I have seen the abacus calculators, who are also quite amazing. I do not think what DT is doing is the same thing, though – at least he doesn’t describe it in that way. Though he does seem to be manipulating objects in space when he moves his fingers while performing a calculation, he also describes how the correct answer “appears” to him and he just needs to trace the outline of its shape to figure out what it is. At least, that is how it sounds to me – I may not be doing his description justice. He did say that the shapes of different numbers “fit together” somehow in ways that let him figure out the right answer.

    As for calendar calculations, I do not see how just measuring elapsed time very accurately could accomplish that, especially for very distant dates. That may be more a failure of my powers of imagination, however, as interestingly, an ability to estimate time extremely accurately without any chronometer is a recognised area of “genius” in some savants. (I do not know if those are the calendar calculators but that might be known).

  9. Lemniscate, that’s an interesting idea – that the formalities of language imposed on numbers may override a more effective way to think about and manipulate them. It is interesting to ask what is the normal function of the circuits that CAN perform these calculations but usually do not. If they were not selected for the ability to do complex calculations, what were they selected for?

  10. Synaesthesia is interesting because it’s a qualitative difference between individuals, not just a quantitative one.

    Vladimir Nabokov was a lifelong synaesthete. He was also a math prodigy at age 5 — he could calculate logarithms in his head — but then he got a severe fever and that skill vanished.

    Nabokov is a good example of the genius as a well-rounded, successful, happy man. But he had one lack: he couldn’t hear music at all. I wonder if his synaesthesia and amusicality were related?

  11. Thanks Steve. Nabokov’s biography “Speak, Memory” is a fascinating read and includes detailed descriptions of his synaesthesia, giving a very personal (and beautifully realised) account of his unique perceptual experiences and how they, literally, coloured his thinking. I hadn’t realised that he was tone deaf, or amusic. I have lately written a review contrasting synaesthesia and several forms of associative agnosias, including tone deafness, face blindness, dyslexia and others. Synaesthesia seems to result from hyperconnectivity, at least locally, while these other conditions seem to arise from hypoconnectivity, especially a failure to connect anatomically distributed nodes in a network. Autism also is thought to be characterised by increased local hyperconnectivity at the expense of long-range connectivity. Perhaps there is a closer relationship between these various “symptoms” in some individuals than the theory espoused above (which assumes their independence) would suggest.

    Curiouser and curiouser: genetic disorders of cortical specialization.
    Curr Opin Genet Dev. 2011 Jun;21(3):271-7. Epub 2011 Feb 4.

  12. From what I’ve seen in history writing the 10% myth came out of pioneering 19th Century brain mapping work. A researcher made a silly claim to a newspaper reporter or what he said was misunderstood. But the meme took hold and nothing it seems will ever kill it. It drives me up the wall that every single science fiction show on television eventually trots it out as a plot device. All too often it signals the shark has been jumped.

  13. Fascinating, as my mother has ASD symptoms and Dyslexia, while 2 female siblings have Dyslexia, and my brother has ASD symptoms and is tone deaf and face blind, and hyperlexic. He was in hospital as a 7 yo for a few weeks and read 27 books, including Dr Jekyl and Mr Hyde. I have some ASD symptoms myself and was somewhat hyperlexic as a child – I taught myself to read at 4 yo, and was reading an autobiography of Gandhi at 7 yo, and was told I had the reading age of a 12 yo. My son has a diagnosis of ASD and seems to be hyperlexic also, at 2.5 yo he has the vocab of a 5 yo – according to his speech therapist. He also, like my brother, sometimes exhibits Echolalia. He will use phrases out of context, like he said to be the other day, “Dad, I’ve got bad news or you!”, and when I asked what news, he said, “I don’t know”. My daughter also had some ASD symptoms, but at 8 yo, these seem to have abated, she is also a precocious reader and may be hyperlexic.

    Another thing my brother would do is remember loads of dialogue from movies, but sometime fail to completely understand the plot, or characters actions or MO in the movie.

Leave a Reply