95% of people assign the exact same name to a spiky shape
When shown a spiky shape and a rounded shape and asked which is "Kiki" and which is "Bouba," almost everyone agrees. Across diverse languages and cultures, roughly 95% of people label the spiky shape Kiki and the smooth shape Bouba. This phenomenon, known as sound symbolism, reveals that human brains naturally map visual sharpness and roundness directly to specific speech sounds and physical tongue movements.
An Island Experiment and Nonsense Words
In 1929, the German psychologist Wolfgang Köhler conducted a simple visual experiment while working on the island of Tenerife. He drew two distinct shapes on a sheet of paper: one was composed of sharp, jagged angles with narrow spikes jutting in all directions, while the other was a continuous, soft, undulating blob. Köhler presented these drawings to participants and presented them with two made-up names, 'takete' and 'maluma.' When asked to assign one name to each drawing, the overwhelming majority paired 'takete' with the spiky figure and 'maluma' with the rounded one.
Köhler’s early experiment suggested that the human association between spoken sounds and visual forms was not purely random. At the time, conventional linguistic thought held that the relationship between the sound of a word and its meaning was completely arbitrary, an agreed-upon cultural convention rather than an intrinsic psychological link. Köhler’s findings offered an early, provocative challenge to this view, demonstrating that even when presented with entirely novel, meaningless pseudowords, human listeners shared consistent, intuitive expectations about what those words should look like.
The Modern Rebirth of Bouba and Kiki
More than seven decades later, neuroscientists Vilayanur S. Ramachandran and Edward Hubbard revisited Köhler’s paradigm to explore the neural foundations of metaphor and language. In their 2001 study, they adapted the test by substituting the nonsense words 'bouba' and 'kiki' while keeping the essential visual contrast between an angular, starburst-like shape and a smooth, curvy form. They tested both native English-speaking undergraduate students in the United States and native Tamil speakers in India, asking which shape was 'bouba' and which was 'kiki.'
The results were remarkably uniform across both groups: roughly 95% to 98% of participants designated the curved shape as 'bouba' and the jagged shape as 'kiki.' Because the effect manifested strongly across distinct languages and writing systems, it demonstrated that the phenomenon did not depend on familiarity with the Latin alphabet or specific cultural idioms. The test established what is now widely known as the bouba/kiki effect, bringing renewed scientific focus to sound symbolism—the idea that vocal sounds carry intrinsic perceptual qualities.
How the Brain Connects Sound and Shape
To understand why nearly everyone agrees on these labels, researchers look at the acoustics of speech and the mechanics of articulation. The word 'kiki' is composed of harsh, unvoiced stop consonants—specifically the velar consonant /k/—and a high front vowel. Producing these sounds requires abrupt, sharp movements of the tongue, and the acoustic waveform displays rapid, jagged transients with higher frequency energy. In contrast, 'bouba' features a voiced bilabial stop /b/ and rounded back vowels, which require rounding the lips into an open, relaxed posture and produce softer, continuous acoustic contours with lower frequency dominance.
Researchers propose that the brain performs a form of cross-modal sensory translation, mapping the physical and acoustic properties of speech onto the geometric properties of visual stimuli. When the auditory cortex processes the sharp acoustic spikes of 'kiki' or the motor cortex coordinates the abrupt tongue movement needed to speak it, the brain detects structural similarities with the sharp visual vertices of the jagged drawing. This cross-wiring between visual, auditory, and motor pathways allows abstract properties like 'sharpness' or 'softness' to translate effortlessly across different sensory domains.
Evidence Across Cultures, Age, and the Senses
Subsequent investigations have examined whether this cross-modal matching requires literacy, formal schooling, or long-term linguistic experience. Researchers tested remote populations, such as the Himba people of northern Namibia, who have traditional lifestyles and limited exposure to Western media or alphabetic writing. The Himba exhibited the same fundamental preference, confirming that the mapping does not rely on visual associations with letter shapes—such as the sharp angles in the letter 'K' or the rounded curves in 'B.'
Developmental studies reveal that this intuition emerges remarkably early in human life. Toddlers as young as two and a half years old, as well as pre-verbal infants tested using eye-tracking and preferential looking techniques, show consistent associations between rounded shapes and round-sounding pseudowords. Furthermore, researchers adapting the test for individuals who are blind found that participants who explored 3D plastic models through touch alone still mapped 'kiki' to the tactilely sharp object and 'bouba' to the smooth one, demonstrating that the effect spans haptic sensation as well as vision.
Rethinking the Foundations of Human Language
The bouba/kiki effect holds profound implications for theories of language evolution. For much of the twentieth century, linguistics was anchored in Ferdinand de Saussure’s principle of the 'arbitrariness of the sign'—the premise that there is no inherent reason why a dog should be called 'dog' rather than any other sequence of phonemes. While most mature vocabulary contains arbitrary historical developments, the robust nature of sound symbolism suggests that the earliest human proto-languages may have emerged from non-arbitrary, cross-modal mappings.
According to evolutionary models proposed by Ramachandran and other cognitive scientists, early hominin communication likely relied on synesthetic gestures, lip movements, and vocalizations that naturally mimicked physical properties of the environment. A sharp sound or a tight mouth posture would naturally represent a cutting stone, an edge, or a sudden movement, while softer, resonant sounds paired with rounded lips would communicate gentle, swelling, or rounded objects. Sound symbolism may have provided the evolutionary scaffolding that allowed early humans to establish mutually understood vocal labels before formal grammatical conventions existed.
Neurological Variations and Open Questions
While the bouba/kiki association is widespread, it is not entirely universal, and studying variations in how people respond has yielded valuable neurological insights. Ramachandran and Hubbard highlighted the angular gyrus—a region in the parietal lobe situated at the intersection of visual, auditory, and somatosensory processing areas—as a critical site for cross-modal integration. Individuals with localized damage to the angular gyrus often struggle with metaphor comprehension and perform poorly on the bouba/kiki task, failing to form the standard sound-to-shape pairings.
Researchers have also investigated neurodivergent populations, observing varied outcomes. Some studies have noted that individuals on the autism spectrum show a reduced or inconsistent bouba/kiki effect, suggesting differences in how multisensory information is bound together in the brain, though other experiments indicate that many autistic individuals display typical responses. Modern researchers continue to study how much of sound symbolism is innate, how phonological constraints in specific native languages subtly reshape these biases, and the precise degree to which sensory cross-activation shapes human cognition.
Key takeaways
•First discovered with 'takete' and 'maluma' in 1929 and modernized as 'bouba' and 'kiki' in 2001, the effect shows that 95% or more of people pair spiky shapes with sharp-sounding words.
•The phenomenon relies on cross-modal sensory mapping, where the brain links the sharp acoustic bursts and abrupt tongue movements of words like 'kiki' to visual angles.
•The preference appears in pre-verbal infants, remote non-literate cultures, and blind individuals evaluating tactile shapes, demonstrating that it does not depend on alphabet shapes or formal literacy.
•Sound symbolism challenges the assumption that language is entirely arbitrary, suggesting that natural multisensory mappings helped drive the initial evolution of human speech.