Pet dogs lingered on the round shape for 'bouba' and the spiky one for 'kiki'
Parma researchers found dogs looked more at a round shape on hearing 'bouba' and a spiky one on hearing 'kiki', in tests on 70 dogs. Dogs have co-evolved with people for thousands of years, so the result cannot yet say whether the link is old or learned.
The Scientist · Science desk
In the first experiment, 63 per cent of dogs concentrated their gaze on the spiky shape on hearing 'kiki' and 56 per cent lingered on the round shape on hearing 'bouba'. That experiment involved 51 dogs.
| Figure | Value | Note | Claim |
|---|---|---|---|
| First test: gazed at spiky shape on hearing 'kiki' | 63% of dogs | 9 | |
| Lingered on round for 'bouba' | 56% of dogs | 9 | |
| Dogs in the first test | 51 dogs | 3 |
What happened
- Over roughly 15 minutes, a human voice said 'bouba' or 'kiki' from speakers on either side while the round and spiky shapes swapped between the two screens.
- New Scientist reported that in the first experiment 56 per cent of dogs lingered on the round shape for 'bouba' and 63 per cent concentrated their gaze on the spiky shape for 'kiki'.
- A second experiment with 19 other dogs had owners wear headphones playing rock music so they could not hear the words.
Why it matters
- What it limits constraint The result supports a claim that the link is not unique to humans, but a claim that it is evolutionarily old has to wait for puppies and wild canids.
- What doesn't add up contradiction ScienceAlert reports a spiky-shape bias in the first experiment that vanished with headphones, while New Scientist says the outcome was the same. The two fit together if only the baseline moved. Then the word-to-shape pairing is the comparison to trust.
- Who has to decide decision Cats are the comparison Biffi names for the dog-history question: a positive result would favour general perceptual mechanisms and count against adaptations tied to dogs' cooperation with humans.
Next story · Science · 2d ago
Researchers say the Wildlife Photographer of the Year elephant has picked up other females' dead calvesConfirmed2 publishers
The second experiment was built to remove any owner cues the dogs might have been following [11]. Owner influence is the obvious worry when a person sits with the dog [3], and the headphones address it directly. The authors add that cueing is an unlikely explanation for the spiky-shape bias in the first experiment, since humans typically prefer rounded or curved shapes [19]. For that bias they offer another cause: the angular faces of ancient dogs and wolves, set against round infant human faces, could plausibly push dogs the opposite way [20].
The headphone group is small. Nineteen of the 70 dogs is 27 per cent of the sample [6][18]. Gaze was scored by an AI that tracked head direction as a stand-in for where the dogs were looking [5], and the authors report that the differences were significant [7]. The coverage does not give test statistics, effect sizes or the denominator behind New Scientist's 56 per cent for 'bouba'. That figure is not a whole number of the 51 dogs: 28 dogs would be 55 per cent and 29 would be 57 [23], so it may be a share of something else, such as trials.
On the larger claim, a narrow reading fits the evidence. Since 2001, many studies have found that people across cultures tend to name a curved shape 'bouba' and a spiky one 'kiki', and the effect has recently been shown in chicks and possibly macaques [13]. Biffi, who led the work published in Proceedings of the Royal Society B [1], said that if animals without human-like language also associate sounds with shapes spontaneously, at least some of the mechanisms behind sound symbolism may not be specific to language. Instead, Biffi said, they "may reflect more general principles of how the brain integrates information from different senses" [14]. We think the dog data support that statement. ScienceAlert's write-up is more guarded: it says a lot more research is needed before scientists can definitively say dogs link 'Bouba' with round shapes and 'Kiki' with sharp ones [12].
The thing this doesn't tell you is whether a dog that had never heard people talk would pick the same shapes. New Scientist reports it is unclear to what extent the effect depends on dogs' experience of human speech or on their having co-evolved with humans for thousands of years [15]. The team is collecting data on puppies and wild canids, including foxes [21]. We'd expect puppies to speak to exposure and wild canids to co-evolution.
Attila Andics at Eötvös Loránd University in Hungary said the study suggests that matching speech sounds to shapes relies on general mechanisms [16]. He said an exciting next question is how far the links in dogs "arise through statistical learning or more hardwired expectations" [17].
What to watch
- Data from puppies and wild canids such as foxes, which would show whether the effect appears without years of exposure to human speech or co-evolution with people.
- Whether cats, which Biffi's team also wants to test, show the bouba-kiki effect.
- The paper's test statistics and effect sizes, and what the 56 per cent figure is a share of.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+10
- Incentives
- Not scored
- Confidence60
Perspective Coverage
3 publishers- Builder
- Builder 47%
- Operator
- Operator 30%
- Investor
- Investor 23%
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
The study, led by Eleonora Biffi at the University of Parma in Italy, was published in Proceedings of the Royal Society B: Biological Sciences.
- [2]
The team recruited 70 domestic dogs to take part in two almost identical experiments.
- [3]
The first experiment involved 51 dogs and their owners; each pair entered a room with a screen and speaker on the left and another screen and speaker on the right, and the owner sat with the dog.
- [4]
Over roughly 15 minutes, the screens and speakers alternated through combinations of a human saying 'Bouba' or 'Kiki' over the speakers and the two shapes swapping between the left and right screens.
- [5]
A GoPro recorded each dog, and the researchers fed the videos into an AI designed to track where the dogs' heads were pointing, a stand-in for where they were looking.
- [6]
The second experiment involved 19 other dogs and was identical in design except that owners wore headphones playing rock music so they could not hear the 'Bouba' and 'Kiki' sounds.
- [7]
Biffi and team report that dogs looked significantly more at the rounded shape while 'Bouba' was played, and looked at and explored the spiky shape when 'Kiki' was played.
- [8]
Dogs in the first experiment showed a general bias toward looking at the spiky shapes; that bias disappeared in the second experiment when owners wore headphones.
- [9]
In the first experiment, 56 per cent of the dogs lingered on the round shape when they heard 'bouba', and 63 per cent concentrated their gaze on the spiky shape when they heard 'kiki'.
- [10]
When owners wore headphones playing music, the outcome was the same, with a slightly stronger bias toward round shapes for the word 'bouba'.
- [11]
The purpose of the headphones in the second experiment was to remove any owner cues the dogs might have been following.
- [12]
ScienceAlert says there is still a lot more research to be done before scientists can definitively say dogs associate 'Bouba' with round shapes and 'Kiki' with sharp ones.
- [13]
The association between certain sounds and spiky or round shapes is common to humans from cultures all over the world; since 2001, many studies have reported that people are much more likely to give the made-up name 'bouba' to a curved shape and 'kiki' to a spiky one, and the effect has recently been shown in chicks and possibly macaques.
- [14]
Biffi said that if animals without human-like language also spontaneously associate sounds with shapes, at least some of the mechanisms underlying sound symbolism may not be specific to language; instead, they 'may reflect more general principles of how the brain integrates information from different senses'.
- [15]
It is unclear to what extent the effect depends on dogs having experience of listening to human speech, or on the fact that dogs have co-evolved with humans for thousands of years.
- [16]
Attila Andics at Eötvös Loránd University in Hungary said the study suggests that matching speech sounds to shapes relies on general mechanisms and not on complex linguistic representations.
- [17]
Andics said an exciting next question is to what extent the links in dogs 'arise through statistical learning or more hardwired expectations'.
- [18]
The 19 dogs in the headphone experiment are about 27 per cent of the 70 dogs in the study.
- [19]
The authors wrote that owner cues are unlikely to explain the spiky-shape bias, because humans typically show a stronger preference for rounded or curved shapes than for spiky ones.
- [20]
The authors suggested that the angular facial configuration of ancient dogs and wolves, in contrast to the round facial morphology of human infants, could plausibly contribute to an opposite bias.
- [21]
Biffi says the team is now collecting data on puppies and wild canids, including foxes, and also wants to see if the effect exists in cats.
- [22]
Biffi said that if cats also show the bouba-kiki effect, it would provide further evidence that sound symbolism may reflect general perceptual mechanisms, as opposed to adaptations specifically related to dogs' long history of cooperation and communication with humans.
- [23]
A share of 56 per cent is not a whole number of 51 dogs: 28 of 51 dogs is 54.9 per cent (about 55) and 29 of 51 is 56.9 per cent (about 57).
Sources
3 independent publishers whose own reporting we read for this story.
- newscientist.comDogs are susceptible to the bouba-kiki effect
1 article · October 13, 2026
- phys.orgYour dog may link sounds with shapes, just like you do
1 article · October 14, 2026
- sciencealert.comDogs Seem to Agree With Humans That 'Bouba' Is Round And 'Kiki' Is Spiky
1 article · October 13, 2026
Topics and entities
Topics
Entities
- University of Parma
- Proceedings of the Royal Society B