2 is Orange but Two is Blue

by David F. Salisbury

In the world of synesthesia, numbers, letters and words have distinct colors

For nearly 300 years, people who claimed to hear colors, feel sounds or taste shapes risked being dismissed as having overactive imaginations. Last year, however, several scientific studies of this rare condition produced compelling evidence that it is a genuine perceptual phenomenon. Now, a Vanderbilt study of the most common form of synesthesia -- the perception that numbers, letters and words have distinct colors -- has confirmed these earlier findings.

Little is known about synesthesia's causes or its prevalence. Estimates range from one in 2,000 to one in 25,000 and there is also some evidence that the condition is more common in women than in men. Nevertheless, a number of famous people -- including the poets Baudelaire and Rimbaud, the painters Kandinsky and Klee, the composers Lizst and Scriabin -- have been linked to synesthesia. In an attempt to describe what synesthesia is like, novelist and synesthete Vladimir Nabokov wrote that he saw the letter "c" as light blue; associated "a" with the look of "weathered wood," and got a feeling like "a sooty rag being ripped" from the letter "r."

The subject of the Vanderbilt study consisted of administering a battery of perceptual tests to a middle-aged man known as "WO" who said that he had experienced synesthesia since childhood.

These findings support an idea that has been around for decades: that synesthesia is caused by a subtle cross-wiring in the brain. Specific regions in the brain process information about different aspects of the visual scene, such as color, shape and motion. Recent brain mapping studies have found that a primary color area is adjacent to an area that handles numbers and letters. Another color area lies next to a primary auditory area. If the neurons in these regions were more densely wired or strongly connected than normal, it could explain why some people see words, numbers and sounds in color.

"WO sees letters, numbers and individual words printed in black-and-white in vivid color. If the characters are printed in different colored ink, he can see that color as well," says Thomas J. Palmeri, assistant professor of psychology, who headed up the study.

The tests go far beyond the tests of individuals with synesthesia that have previously been reported. After determining that WO had normal eyesight and color vision, they drew up a list of 100 common words of one syllable. In two sessions, separated by more than a month, they asked him to name the color associated with the words in the list. They found that his color associations were remarkably consistent and associated the same color with 97 of the words. His only mismatches occurred in confusing beige, off-white and light brown. He also consistently matched the synesthetic colors of 12 words with colors from a Pantone palette containing more than 1,000 shades. "These associations are highly reliable. WO says that the colors have stayed the same all of his life and our observations lend credence to his claim," says Professor of Psychology Randolph Blake, who was another collaborator.

WO sees "2" as orange and "5" as green but he sees both "6" and "8" as nearly the same shade of dark blue. Using a special font so that the "2" and "5" are mirror images, the researchers made up two sets of computer displays. One set had single white "2s" hidden among varying numbers of white "5s." The other set had single white "8" hidden among varying numbers of white "6s."

The Stroop test is a classic example of how color can interfere with a simple identification task. When subjects are asked to name the color of ink in which color words are printed, they can do so faster when the ink color and color name are same ("red" printed in red ink) than they can when the two are different ("green" printed in red ink). The researchers adapted this test by printing words in colors that either agreed or conflicted with their synesthetic color and then recording how long it took WO to identify them. For example, WO sees the word "moose" as pink, so they printed it in pink letters and he sees the word "charge" in green, so they printed it in blue. They found that it took him longer to name words where the synesthetic color conflicted with the ink color than it did when they were the same.

"Although there are some subtle differences, these tests showed that synesthetic colors act in a way that is very like real colors," Palmeri summarizes.