Crossing the senses

Noisy apricot blossoms

The Song dynasty poet Song Qi (宋祁, 998–1061) wrote a short poem to the tune “Yulouchun” about the first warm days of spring. Its last line became one of the most famous, and most argued over, in Chinese:

东城渐觉风光好，縠皱波纹迎客棹。
绿杨烟外晓寒轻，红杏枝头春意闹。

In plain English: the east of the city grows lovelier by the day; ripples like crinkled silk welcome the traveler’s oar; past the green willows and mist the dawn chill is slight; and on the red apricot branches, spring is 闹 (nào), “noisy,” “clamoring.” A silent picture of blossom is described in the vocabulary of a crowd.

The playwright and critic Li Yu (李渔, 1611–1680) did not like it. In his Kuci guanjian (窥词管见, “A narrow view of the ci poem”), entry 7, he wrote that 闹 is the word for the sound of quarrelling; peach and plum may compete in spring, but he had never seen red apricots make a din. And if 闹 may be used, he says, then 吵 (brawl), 斗 (fight), and 打 (hit) may be used too. He calls the word coarse and common. Li Yu’s test is reasonableness: the same entry praises another poet’s “clouds part, the moon comes, and the flowers play with their shadows” as something that could happen in the world.

Three centuries later the scholar Qian Zhongshu (钱锺书, 1910–1998) answered him in an essay called 通感 (Tōnggǎn, “feeling across,” usually translated “synesthesia”), first printed in the journal Wenxue pinglun (Literary Review) in 1962 and later gathered in his collections Jiu wen si pian (1979) and Qi zhui ji (1985). The word 闹, he said, takes the silent posture of a thing and speaks of it as if it were a ripple of sound: sight has acquired the feel of hearing. And Qian went on to say what such crossings are. In his words (在日常经验里，视觉、听觉、触觉、嗅觉、味觉往往可以彼此打通或交通), in everyday experience sight, hearing, touch, smell, and taste can open into one another; the realms of eye, ear, tongue, nose, and body need not keep their borders. Color seems to have temperature, sound shape, warmth and cold weight, and smell substance.

Qian’s own examples are wide. He cites Homer, whose old Trojan elders on the wall of Troy are compared to cicadas in a forest that “pour forth their lily-like voice” (Iliad 3.152, in A. T. Murray’s 1924 translation), and the English poet John Donne with his “loud perfume.” He also finds the same device in everyday Chinese. The common words 热闹 (rènao, bustling, literally hot and noisy) and 冷静 (lěngjìng, calm, literally cold and quiet), he points out, already pair heat with noise and cold with quiet. 热闹 contains the very character that Li Yu objected to, in a word every child learns. Qian was not an uncritical admirer of the device, either: he remarks that the Symbolist poets of the nineteenth century overused it.

How words travel

English does the same thing without noticing. A sweet voice borrows from taste, a bright sound from sight, a soft light from touch, a loud shirt from hearing. Linguists have looked for a pattern. In the 1950s the linguist Stephen Ullmann proposed that words tend to move along a ladder from the “lower” senses, touch and taste, toward the “higher” ones, hearing and sight; in 1976 Joseph Williams tested the idea against a large body of English adjectives in the journal Language. Later work, such as that of Francesca Strik Lievers and Bodo Winter, finds asymmetries and not exceptionless rules. The safe statement is a tendency: in English, words travel from touch and taste toward hearing and sight (a sweet voice, a soft light) more often than the reverse. Between sound and sight both directions occur: a loud shirt, a bright sound, and the Chinese 明亮的声音.

Footprints of nature

Qian drew on Bacon, and the lecture reports the passage with a twist, so it deserves a careful look. In The Advancement of Learning (1605), Book 2, Francis Bacon argues for a missing science that would collect the principles common to many arts. As an illustration he asks: “Is not the delight of the quavering upon a stop in music the same with the playing of light upon the water?” and quotes a line of Virgil, splendet tremulo sub lumine pontus, “the sea gleams under the quivering light” (Aeneid 7.9). These likenesses, he says, are not only similitudes, “as men of narrow observation may conceive them to be, but the same footsteps of nature, treading or printing upon several subjects or matters.”

The lecture renders this as “sound flows on the sea like light,” and calls it Bacon on synesthesia in the sixteenth century. It is not quite either. The book is from 1605, and the passage compares a musical effect to a visual one as an analogy between two domains, not a crossing of the senses within one perceiver. The word synesthesia itself does not appear in anything like its modern sense until 1892. But the underlying idea is Bacon’s and it is a good one: that the same patterns repeat in different materials, so a likeness found in one place may be a clue to a law in another.

A color for every letter

The word synesthesia in English has a narrower sense than Qian’s 通感. It names a condition in which a stimulus in one sense reliably causes an experience in another: a letter or a number appears in a color, a sound has a shape, a word has a taste. The lecture says that until recently the term appeared only in psychiatric case files. That is not right. Here is what is solidly known, and what is not.

Established. A Latin medical dissertation of 1812 by Georg Tobias Ludwig Sachs, a young man with albinism, describes colors for music, numbers, days, and letters in himself and his sister, and historians call it the first convincing medical account. In 1880 the English polymath Francis Galton published “Visualised numerals” in Nature, on people who see numbers laid out in space; his book Inquiries into Human Faculty (1883) contains the first color plate of such associations. In 1881 the Swiss psychiatrist Eugen Bleuler and a student, Karl Lehmann, argued that “colored hearing” runs in families and is not a disease. The modern word was fixed by Jules Millet in 1892. What defines the condition today is that it is involuntary and consistent over time: ask a person which color a letter has, and ask again a year later, and the answers match far better than any memory could explain. That test was applied in a 1987 study by Simon Baron-Cohen and colleagues, and a battery of such tests was published online by David Eagleman and colleagues in 2007. It runs in families, though no single gene explains it. A 2009 study of 43 families found linkage to a stretch of chromosome 2 and suggestive signs on others. In the brain, spoken words trigger the color-processing region in people who see colors when they hear (a 2002 study). And for letters and colors specifically, researchers found in 2013 that the pairings of eleven synesthetes could be traced to colored letters on toys they had played with as children.

How common is it? In 2006 a careful study led by Julia Simner, which did not rely on self-referral and used objective tests of genuineness, reported roughly 4 percent with some form, 88 times higher than the older estimate of 0.05 percent; the most common type was colored days of the week, and colored letters about 1 percent. Estimates vary with the method. So the safe statement is “a few in a hundred.”

Disputed. How the crossing happens is still argued. Vilayanur Ramachandran and Edward Hubbard proposed in 2001 that neighboring brain areas, one for color and one for numbers, are “cross-wired”; others favor a failure to inhibit feedback between areas, or learned associations. The idea that everyone is a “weak” synesthete has defenders, and a sharp critique, “Why we are not all synesthetes (not even weakly so)” (Deroy and Spence, 2013). The neurologist Richard Cytowic brought the condition to the public in The Man Who Tasted Shapes (1993), about a patient who said at a dinner, “There aren’t enough points on the chicken,” and proposed a deep, limbic basis for it. That idea is not the standard picture today.

Only reported. The most famous names are mostly self-reports. Vladimir Nabokov, in Speak, Memory, says that the long a of the English alphabet “has for me the tint of weathered wood,” and tells how, at about seven, he told his mother that the colors of his alphabet blocks were all wrong, and she, who had colors of her own, agreed. His wife, Véra, and his son, Dmitri, are said to have shared it, a pattern that fits what is known about families, but no one tested him in a laboratory. Wassily Kandinsky wrote in his memoir (1913) that at a performance of Wagner’s Lohengrin in Moscow, about 1896, he saw all his colors before his eyes and wild lines drawn in front of him. Whether he was a true synesthete or a theorist who borrowed the language is argued: a 2003 paper asked directly, “Was Kandinsky a synesthete?” and the answer remains disputed. The physicist Richard Feynman is reported to have said he saw equations in color. Arthur Rimbaud’s sonnet “Voyelles” (written about 1871, printed in 1883), with its black A, white E, red I, green U, and blue O, was probably a poetic construction: in a later work Rimbaud says he “invented” the colors of the vowels. Charles Baudelaire’s “Correspondances” (1857), “Les parfums, les couleurs et les sons se répondent,” states a doctrine of correspondence, not a case report.

Bouba and kiki

There is a middle case, which nearly everyone has. In 1929 the psychologist Wolfgang Köhler showed people two drawings, a rounded blob and a jagged star, and asked which was called takete and which baluba (in the 1947 edition, maluma). Most answered, he said, without hesitation: the jagged one was takete. In 2001 Ramachandran and Hubbard gave the demonstration new names, bouba and kiki, and the pairing has been tested ever since. Their own report was informal, and the figure usually quoted (95 percent, in one telling 98) comes from it. A better number comes from a meta-analysis of the classic studies in 2017: about 89 percent of answers matched the round word to the round shape, with the true value likely between 84 and 94 percent.

The effect has been found in toddlers as young as two and a half (Maurer and colleagues, 2006), and among the Himba of northern Namibia, a people with no written language (Bremner and colleagues, 2013), who also matched shapes to sounds though they differed from Westerners in matching shapes to tastes. A 2022 study across 25 languages and nine language families (917 participants) found the effect widely but not everywhere. Mandarin speakers, 49 in that test, were not clearly above chance, and the authors cautioned that two made-up words may not fit every language’s sounds. A 2016 study found that people in the United States and Taiwan both showed the effect but weighed the spikiness of the shape and the loudness of the sound differently. Earlier tests among two groups in Papua New Guinea and Nepal found it weak or absent. A fair summary: across dozens of languages, most people pair bouba with round shapes and kiki with spiky ones, from toddlers to a remote community, but it can weaken when the made-up words do not suit a language’s sounds.

That is a shared leaning of the senses, not synesthesia. Ramachandran and Hubbard suggested that the two are related, as the ordinary form of a gift some people have in extreme form; some researchers reject the link. Try it, and see where you land.

Crossing, and creating

The lecture ends with a larger claim. Great creators, it says, cross boundaries: they carry a feeling or an idea from one domain into another by analogy, not by logic, and that is what makes a work new. It applies the claim to business: a product that merely improves on an old one is an iteration, whereas one that produces “aesthetic surprise” in its user is a creation. The claim comes with biographical examples, and some of them do not hold.

The lecture gives the artist Yayoi Kusama as an example of the link between hallucination and art. Kusama was born on 22 March 1929 in Matsumoto, Japan, and reported hallucinations from about the age of ten: flashes of light, fields of dots, flowers that spoke. In 1977 she moved, by choice, into a psychiatric hospital in Tokyo, where she lived with a studio nearby for the rest of her life. In 2012 she illustrated an edition of Alice’s Adventures in Wonderland for Penguin. She died on 14 August 2026, aged 97. Nobody has documented her as a synesthete; her experiences were hallucinations, and this course presents them as such. The lecture’s other claims about her (hearing voices at seven, a Japanese plot to kill her, being the only Asian among the twenty greatest artists of the century) could not be verified and are left out.

It also invokes Lewis Carroll. Charles Dodgson (1832–1898) was a lecturer in mathematics at Christ Church, Oxford, from 1855 to 1881, and wrote on Euclid and symbolic logic; under the name Carroll he published Alice’s Adventures in Wonderland. The book was first printed in July 1865, but the illustrator, John Tenniel, objected to the quality of the printing and Carroll recalled it; the reprint (the first copy is dated 9 November 1865) went on sale at Christmas 1865 with the date 1866 on its title page. Through the Looking-Glass came out at Christmas 1871, dated 1872. It is a fine example of a logician writing nonsense on purpose. (The lecture adds C. S. Lewis as another Oxford mathematician; Lewis was an English scholar, so this course leaves him out.)

And it says the physicist Yang Zhenning (杨振宁, C. N. Yang) wrote that the idea of parity violation came to him from the perfect symmetry of snowflakes. We could not find such a statement. What the record shows is this. In 1956 Yang and Tsung-Dao Lee (李政道) published a paper in Physical Review, “Question of Parity Conservation in Weak Interactions,” prompted by a puzzle about two apparently identical particles that decayed differently. Chien-Shiung Wu (吴健雄) and a team at the National Bureau of Standards in Washington saw the asymmetry in cobalt-60 nuclei in December 1956, and the result was announced in January 1957; other teams confirmed it independently. The 1957 Nobel Prize in Physics went to Yang and Lee alone. Wu was not included, and in 1978 she received the first Wolf Prize in Physics. “Wu helped them win” is fair; “Wu shared the prize” is wrong.

What does the evidence say about creativity and the senses? It is modest. In 2008 Jamie Ward and colleagues studied 82 synesthetes and found that they spent more time on creative arts than comparison groups, but showed only a mixed result on standard tests of creativity (better on one, not on another); the authors wrote that there was no direct link. In 2010 Nicolas Rothen and Beat Meier found grapheme-color synesthesia in about 7 percent of 99 art students against about 2 percent of 96 other students: small samples, one type, but suggestive. Research on analogy gives more general support: Dedre Gentner’s work on structure-mapping, a 1983 theory, and observations of scientists in the lab in 2001 (Kevin Dunbar and Isabelle Blanchette) found that analogies play a large part in scientific reasoning, and a 2013 study (Roger Beaty and Paul Silvia) found that creative metaphors depend on fluid intelligence and the ability to retrieve ideas from memory. Arthur Koestler’s “bisociation,” the meeting of two unrelated frames of reference, from The Act of Creation (1964), is a theory, not a finding. And the lecture’s jump from 通感 to “cross-boundary innovation” to a user experience that is an aesthetic surprise rather than an iteration is its own argument; we found no study that supports it. The lecture also quotes “there is nothing new under the sun” as from the New Testament. It is Ecclesiastes 1:9, in the Old: “there is no new thing under the sun,” in the King James Version.