If you show people a jagged shape and a curvy shape, and ask which is "bouba" and which is "kiki," nearly everyone makes the exact same choice. Over 95% of people label the round shape "bouba" and the jagged one "kiki." This phenomenon occurs across different languages and cultures, showing that our brains naturally map visual shapes to specific speech sounds in a universal, non-arbitrary way.
The Island Experiment That Started It All
In 1929, German psychologist Wolfgang Köhler conducted a simple experiment while working on the island of Tenerife. He drew two distinct shapes on a sheet of paper: one was a smooth, rounded cloud-like outline with soft curves, and the other was a jagged, spiky figure with sharp angles. He then presented participants with two invented, nonsense words—'maluma' and 'takete'—and asked them to match each name to one of the shapes. Despite having no cultural or linguistic definitions to rely on, the overwhelming majority of participants assigned 'maluma' to the soft, rounded shape and 'takete' to the angular, spiky one.
Köhler's experiment demonstrated a curious phenomenon: humans do not perceive abstract sounds and visual shapes in total isolation from one another. Instead, the mind appears predisposed to map specific auditory textures onto particular geometric contours. For decades, this finding remained a fascinating footnote in Gestalt psychology, demonstrating that sensory perception involves holistic patterns rather than purely mechanical data collection.
The effect received renewed scientific attention in 2001, when neuroscientists Vilayanur S. Ramachandran and Edward Hubbard adapted Köhler's test using the nonsense words 'bouba' and 'kiki.' Testing both English-speaking American undergraduates and native Tamil speakers in India, they found that roughly 95 percent of participants in both groups matched the jagged form with 'kiki' and the curvy form with 'bouba.' This striking consistency indicated that the mapping was not an artifact of a single language or alphabet, but pointed toward a fundamental property of human cognition.
To understand why 'bouba' feels rounded and 'kiki' feels sharp, researchers look to the physical mechanics of speech production and acoustic phonetics. When articulating the word 'bouba,' the vocal tract undergoes smooth, rounded physical transformations. The lips round outward to produce the vowel sound, and the transition between the voiced consonant 'b' and the open vowel is gradual and soft. The resulting sound wave is rich in lower frequencies and lacks abrupt, high-energy acoustic disruptions.
By contrast, pronouncing 'kiki' requires sharp, sudden movements of the tongue against the hard palate. The unvoiced stop consonant 'k' demands a rapid buildup and sudden release of air pressure, creating a sharp transient burst in the acoustic signal. Furthermore, the high-front vowel sound in 'kiki' requires the lips to stretch wide and tight, mirroring the tense, angular geometry of a spiky outline. The acoustic profile of 'kiki' contains high frequencies and sudden, jarring acoustic transients.
The human brain naturally aligns these physical domains. The visual system extracts properties like sharp vertex angles or continuous curvilinear contours, while the auditory and motor systems track frequency bursts, spectral shape, and the physical effort of articulation. When these sensory streams converge, the mind recognizes a shared structural quality between the visual sharpness of a point and the acoustic sharpness of a sudden, high-frequency sound.
Challenging the Arbitrariness of Language
For much of the twentieth century, modern linguistics operated under the foundational premise established by Ferdinand de Saussure: the principle of the arbitrariness of the sign. According to this doctrine, the phonetic sounds that make up a word have no intrinsic connection to the object or concept they describe. A dog is called 'dog' in English, 'perro' in Spanish, and 'chien' in French purely by arbitrary social convention, not because the animal naturally sounds like any of those syllable sequences.
The bouba/kiki effect presents a direct challenge to the idea that all linguistic connections are entirely arbitrary. While complex grammar and abstract vocabularies clearly rely on convention, the phenomenon demonstrates that sound symbolism—an inherent, non-arbitrary relationship between phonetic qualities and sensory meaning—is deeply embedded in human communication. Phonemes carry sensory baggage long before they are assigned formal dictionary definitions.
This connection offers compelling clues about the evolutionary origins of human speech. Some evolutionary linguists suggest that early human proto-language may have begun not with completely random vocalizations, but with cross-modal imitations of physical events, textures, and movements. A sound mimicking the sharp crunch of a snapping branch or the rounded motion of a rolling stone could have formed a shared, intuitive starting point for shared vocabulary among ancestral hominids.
Universality Across Ages and Cultures
Subsequent research has tested the boundaries of the effect to determine whether it depends on literacy, education, or cultural exposure. Developmental psychologists have found the bouba/kiki preference in toddlers and pre-literate infants, demonstrating that children make these cross-modal associations long before they learn the visual shapes of letters like the rounded 'B' or the angular 'K'. This effectively ruled out the early critique that participants were simply matching the shapes of written letters to the drawings.
Cross-cultural studies have expanded the testing to remote, non-industrialized populations with little to no exposure to global media or Western languages. Researchers visiting remote communities, such as the Himba people in northern Namibia, observed consistent matching patterns between sharp sounds and angular figures. While subtle variations in strength occur across different dialectical groups, the core tendency remains robust worldwide.
The effect extends beyond vision and hearing into other sensory modalities. Researchers investigating cross-modal correspondences have found that people regularly match 'bouba' with sweet, soft, or fatty tastes, creamy textures, and floral scents, while matching 'kiki' with sour, bitter, or carbonated sensations and sharp, pungent smells. This reveals that the effect is not merely an audio-visual quirk, but a general organizational principle across all human senses.
Neurological Pathways and Individual Differences
At the neurological level, the bouba/kiki effect depends on areas of the brain responsible for multisensory integration. Neuroscientists have highlighted regions such as the angular gyrus, located in the parietal lobe at the intersection of visual, auditory, and somatosensory processing areas. The angular gyrus helps translate sensory inputs across modalities, allowing an abstract visual feature to be translated into an auditory or tactile equivalent.
Further evidence for this neural mechanism comes from clinical studies examining individuals with localized brain damage or atypical sensory processing. Patients with damage to the angular gyrus often lose the ability to reliably match bouba and kiki figures, treating the task as a purely random guessing game. Similarly, some studies on autistic individuals have shown variations in the strength of cross-modal matching, suggesting that differences in sensory integration pathways can alter how sound symbolism is experienced.
Studies involving individuals with visual impairments provide further nuance. Congenitally blind individuals who explore tactile 3D models of jagged and rounded shapes often show cross-modal associations when tactile properties are mapped to phonetic sounds, although the strength of these mappings can differ from sight-based experiments. Such findings underline that while cross-modal mapping is a fundamental neural capacity, the specific sensory pathways through which we experience the physical world help shape how those associations manifest.
Key takeaways
•The bouba/kiki effect is a universal cross-modal correspondence where people consistently pair soft, rounded shapes with the word 'bouba' and sharp, jagged shapes with 'kiki.'
•The phenomenon is rooted in shared acoustic, motor, and visual features, such as the rounded lip movements and low frequencies of 'bouba' versus the high-frequency acoustic bursts of 'kiki.'
•Evidence from infants, remote cultures, and tactile studies challenges the linguistic assumption that all word sounds are completely arbitrary conventions.
•Brain structures involved in multisensory integration, particularly the angular gyrus, are central to translating abstract features between sight, sound, touch, and taste.