The Bouba-Kiki Effect: Why the Human Brain Refuses to Believe Words Are Arbitrary

The foundational contract of modern linguistics, drafted by Swiss scholar Ferdinand de Saussure in 1916, rests on a single binding constraint: the arbitrariness of the sign. For human language to function, Saussure argued, the link between a spoken sound and its underlying concept must be entirely random. A tree is only a tree because English speakers agree it is; the French call it an arbre, the Irish a crann. "The bond between the signifier and the signified is arbitrary," Saussure declared, establishing the rule that the physical universe dictates nothing about our vocabulary. If that constraint holds, any random collection of syllables has an equal chance of describing any physical object. But nearly a century of cognitive science has proven that the human brain routinely violates Saussure's contract.

The most famous breach of this linguistic law began on the Spanish island of Tenerife in 1929. Psychologist Wolfgang Köhler presented participants with two abstract line drawings—one a jagged, star-like explosion of sharp angles, the other a smooth, undulating cloud. He then offered them two nonsense words, "baluba" and "takete," and asked which word belonged to which shape. The universe may not care what we call a shape, but Köhler's subjects did. They overwhelmingly assigned "takete" to the sharp angles and "baluba" to the curves.[3]

Seventy years later, neuroscientists V.S. Ramachandran and Edward Hubbard revived the experiment with a slight phonetic tweak, asking college students to categorize the nonsense words "bouba" and "kiki." The results were staggering. Across both American undergraduates and Tamil speakers in India, 95% of participants independently decided that the rounded shape was a "bouba" and the spiky shape was a "kiki." The association was so automatic that it appeared to bypass cultural conditioning entirely, revealing a hardwired synesthesia lurking in the general population.[3]

The mechanics of this phenomenon—now universally known as the Bouba-Kiki effect—root themselves in the physical architecture of the human mouth. Pronouncing the word "bouba" requires the lips to form a relaxed, circular O-shape, while the vocal cords produce a low-frequency, continuous resonance. Conversely, articulating "kiki" forces the mouth into a taut, narrow grin, while the tongue snaps against the palate to produce sharp, high-frequency acoustic bursts. The brain simply maps the physical geometry of the vocalization onto the visual geometry of the object.[4]

This crossmodal correspondence is not a learned behavior. Developmental psychologists have tracked the Bouba-Kiki preference in infants as young as four months old, long before they acquire language. It also transcends literacy; researchers have documented the exact same sound-shape mapping in populations with no written language system. The human mind, it seems, comes pre-installed with a baseline assumption that sharp sounds belong to sharp things.[3]

Yet the effect is not an absolute biological mandate. In 2021, a massive global study led by linguist Aleksandra Ćwiek tested the Bouba-Kiki effect across 917 participants speaking 25 different languages and using 10 distinct writing systems. While the study found strong overall evidence for the phenomenon, the success rate was not uniform. Exactly 17 of the 25 languages validated the effect, proving that while the bias is universal, local linguistic environments can amplify or suppress it.[1][3]

The exceptions in Ćwiek's data reveal the limits of our hardwired sound symbolism. Speakers of Mandarin Chinese, Turkish, Romanian, and Albanian showed match rates that hovered below 50%, essentially performing at random. Researchers suspect that the specific phonetic structures of these languages, or deeply ingrained cultural associations with certain syllables, override the brain's default geometric mapping. The biological impulse exists, but cultural software can patch over it.[3]

Neurological wiring also dictates the strength of the association. While the global average match rate sits around 88%, the number collapses in neurodivergent populations. Individuals diagnosed with autism spectrum disorder map the shapes to the sounds at a rate of just 56%—barely above a coin flip. Because autism often involves atypical sensory processing and reduced cross-network connectivity in the brain, the failure to automatically link a sound's frequency to a visual shape provides a measurable window into how neurodivergent minds compartmentalize sensory input.

The implications of these findings stretch far beyond abstract shapes. A July 2025 study published by the American Psychological Association demonstrated that sound symbolism actively shapes how we perceive other humans. When participants were asked to evaluate novel social groups, the phonetic structure of the group's name dictated their initial impressions. Groups with "round" names were perceived as more cooperative and easygoing, while those with "sharp" names were judged as more precise, aggressive, or competitive.[2]

This inherent phonetic bias offers a compelling theory for how human language began. If the earliest hominids possessed a natural instinct to pair sharp sounds with sharp tools and soft sounds with soft materials, the first vocabularies were not arbitrary at all. They were acoustic pantomimes. Saussure's rule remains the functional bedrock of modern linguistics, but the Bouba-Kiki effect proves that before our ancestors agreed on arbitrary rules, they built their first words out of the shapes they felt in their mouths.[1][4]