Newly hatched chicks show bouba-kiki sound–shape mapping

Newly hatched chickens appear to share a perceptual quirk long treated as a human linguistic curiosity: the “bouba/kiki” effect, where certain sounds feel naturally matched to rounded versus spiky shapes.

According to Ars Technica, Italian researchers Maria Loconsole, Silvia Benavides‑Varela, and Lucia Regolin tested chicks just one to three days after hatching—animals young enough to be largely untrained by experience, yet mobile enough to make measurable choices. The protocol was simple: present two shapes (rounded and angular) and play audio recordings of a human voice saying “bouba” or “kiki.” Control conditions included silence and classical music.

The headline result is awkward for a century of confident claims about the arbitrariness of symbols. When “bouba” was played, roughly 80% of chicks moved first toward the rounded shape; with “kiki,” that figure fell to about 25%, with preference shifting toward the spiky object. The effect strengthened slightly in older (three‑day) chicks, but was present even at one day old.

The researchers interpret the finding as evidence for “crossmodal correspondence”: low-level couplings between sensory modalities in which an auditory cue biases visual categorization. Some correspondences are physically intuitive—high pitches map to smaller objects, low pitches to larger ones—because of how sound is generated. Others are less obvious, such as links between pitch and brightness. Ars notes that crossmodal correspondences have been observed across disparate species (dogs, tortoises, chimps), suggesting they are not a boutique feature of human language.

That matters because earlier failures to find bouba/kiki-like behavior in nonhuman primates helped fuel a tidy evolutionary story: perhaps humans possess a special pre-linguistic mapping mechanism that bootstraps sophisticated language learning. The chicken data undercuts that narrative. If a day-old chick—hardly a candidate for proto‑syntax—shows the same mapping, then bouba/kiki looks less like a “language module” and more like a general-purpose perceptual heuristic.

The implication is not just zoological: if symbol–meaning pairings may be biased by biology, then debates about culture, education, and “social construction” operate under constraints that cannot be voted away. That does not mean language is determined, or that words are fate.

The study also offers a mundane explanation for why adult primate experiments may have failed: mature animals have competing motivations (stress, dominance, boredom, training history) that can swamp subtle biases. The absence of evidence may have been a methodological artifact—an evergreen theme in behavioral science.

If this holds up, the bouba/kiki effect is not a meme. Some “arbitrary” conventions may be riding on ancient, cheap neural shortcuts—well before any committee drafted a curriculum on semiotics.