Saturday, Jan. 25, 2003

  More columns by Michael Smith

If that question makes no sense to you -- how can a number have colour? -- you're part of the boring majority. If it does make sense, you have synaesthesia, a puzzling mental phenomenon that affects perhaps five people in every 10,000.

The author Vladimir Nabokov, as a toddler, used to complain that the colours of the numbers and letters on his blocks were "wrong" -- they didn't match the ones his brain assigned them. His mother, also a synaesthete, understood and sympathized.

That anecdote illustrates several aspects of synaesthesia -- it's automatic, starts very young, and tends to run in families. (Character-colour links, by the way, are the most common form of synaesthesia, but other senses can also mingle; one expert told me about a man who had tactile sensations on different parts of his body when he heard different musical instruments.)

While synaesthesia has been known for at least 300 years, researchers still don't know much about it. That's why a report in the current edition of Nature, by University of Waterloo psychologist Mike Dixon and colleagues, is so fascinating.

Dixon studied "C.," a 21-year-old woman with the standard character-colour synaesthesia. (In psychology jargon, the stimulus -- a number or letter in this case -- is called the "inducer"; the resulting sensation, the colour, is called the "concurrent.")

The researchers showed two things first -- that C.'s character-colour links were always the same and that they were automatic. The next question was: what inducer is needed to produce a sensation?

And it turns out that C. didn't need to see a particular character in order to feel the concurrent sensation -- just the idea of the character was enough.

This, in itself, isn't new -- synaesthetes have reported for years that they can get concurrent sensations just by thinking about an inducer. But for the first time, Dixon and his colleagues have shown experimentally this is true.

Here's how: They showed C. a simple math problem: a number, followed by an operator (a plus sign, for example), and then another number. For each number, of course, C. perceived an accompanying colour. But the kicker was this -- she was asked to calculate the answer and then name a patch of colour that appeared on the screen.

Naming a colour is an easy task, of course, unless you happen to have another colour in your mind. There's a well-known test in which people are shown a word and asked to name the colour of the ink it's printed in. Reaction times slow down sharply when the word is a colour name -- 'RED' -- but the ink is, say, green, apparently because the meaning of 'RED' interferes with the perception and naming of green.

So it was with C. For her, the figure 7 is associated with the colour yellow. If she was shown 5+2 and a patch of yellow, she was significantly faster at naming it than if the patch was some other colour.

But C. would never have seen the figure 7; it only arose as the result of some mental arithmetic. There was no physical inducer, proving experimentally what synaesthetes have been saying: the concept alone is enough to trigger the sensation.

But the result may also help to locate the phenomenon, according to Daniel Smilek, a graduate student who was part of Dixon's team. The experiment, Smilek said, shows that meaning plays a key role in synaesthesia and "meaning is usually processed in the cortex."

But if synaesthesia was just some sort of sensory interference -- a kind of leakage between systems -- it's unlikely meaning would play any role at all. So the case of C. may lead to better understanding of where exactly in the brain synaesthesia happens.

Still, it pays to be cautious. Richard Cytowic, a pioneer in the modern study of synaesthesia who literally wrote the book on the topic, said he'd like to see more experiments on more people before he draws any conclusions.

"A single-case study is very shaky ground," said Cytowic, author of "The Man Who Tasted Shapes" and "Synaesthesia: A Union of the Senses."

On the other hand, cognitive neuroscientist Peter Grossenbacher of the National Institute of Mental Health in Bethesda, Md., said he thought the Dixon paper was a "significant advance in the published knowledge of synaesthesia" because it's the first experimental confirmation that synaesthetes are right about how concepts can drive their sensations.

Grossenbacher has been studying the topic for years and has research links with more than 200 synaesthetes.

Exactly who are synaesthetes is hard to know. They are very often women, usually well-adjusted and often quite bright. There are some high-profile examples: Nabokov; the musician Alexander Scriabin; the artist David Hockney.

But, although one survey in England put the incidence at 5 in 10,000, most synaesthetes don't make a lot of noise about their unusual sensations because they learn very early that other people don't have the same responses.

In C.'s case, Smilek said, she learned as a young girl that others do not associate colours with numbers: "She was trying to teach her younger sister how to do math in colour," he said. Unfortunately, the colourful lessons didn't work; her sister's numbers stayed relentlessly black and white