Communication
Body language, vocal learning, symbols and buttons, and the honest limits of communication technology.
Comparative cognition is not a leaderboard. It asks a narrower and far more interesting question: which operations can this species, or this individual, perform under this experimental design — and how do we know? Every page in this branch labels its evidence so you can see where demonstrated science ends and where enthusiasm begins.
A parrot correctly labelling quantities is evidence about numerical cognition. It is not evidence that the bird “thinks exactly like a four-year-old.” A parrot using a mirror to locate a hidden object demonstrates mirror-mediated information use. It does not automatically demonstrate self-recognition. Holding those apart is not pedantry — it is the difference between comparative psychology and marketing copy.
So this branch reports what a study actually measured, how many animals it measured, and what the result does not establish. Where a famous claim has weakened since it was published, you will find that here rather than a tidier version of the story.
These labels appear beside individual claims, not whole pages. One page routinely carries several grades, because that is how the literature actually looks.
Directly supported by controlled peer-reviewed experiments, converging primary evidence, or an authoritative record source. It does not mean every member of the species shows the ability.
Scientifically interesting, but resting on small samples, one or a few individuals, limited replication, conference work, or an inference that has not yet generalised.
Viral, anthropomorphic, commercially repeated, or simply stronger than the available data justify. Often built on a real finding that has been stretched.
Guinness World Records currently publishes the record. A record is a documented attempt under one organisation's rules — not a scientific census.
A historical or widely repeated claim whose primary-source chain we could not close. Shown with its provenance problem stated, never as settled fact.
Owner-demonstrated or viral material. Genuinely enjoyable, and kept visibly separate from experimental evidence rather than mixed in beside it.
Each room takes one research area, states what has been demonstrated, and names the leap that is usually made just past it.
Tool manufacture in Goffin's cockatoos, inference by exclusion, object permanence, self-control, kea statistical reasoning, and voluntary helping between greys — with the sample sizes kept visible.
Product-relevantWhat the mirror mark test actually tests, why no parrot has passed it convincingly, and the 2024 result showing parrots engage far more with live video than with recordings of the same birds.
ProvenLearned signature calls in green-rumped parrotlets, regional dialects in yellow-naped amazons, and adult budgies reshaping their calls around a changing social network.
One bird, tested wellSnowball's beat synchronisation survived an experimental tempo manipulation. Whether your parrot has a favourite genre is a much weaker claim than the internet suggests.
ProvenHearing that peaks around their own call frequencies rather than in ultrasound, ultraviolet colour vision an ordinary screen cannot reproduce, and flicker sensitivity worth designing around.
GWR verifiedPuck, Oskar, Apollo, Cookie, and Spoodles — plus the 130,000-bird lifespan dataset that dismantles “large parrots live a hundred years,” and why there is no smartest bird.
Not peer reviewedDisco, the several birds called Einstein, and viral clips generally. Worth watching, worth enjoying, and deliberately not filed beside Alex as equivalent evidence.
ProvenWhy a parrot can sound human at all: the syrinx, tongue articulation, the core-and-shell brain organisation, and the 2025 map of a speech-motor-like system in the budgerigar forebrain.
Bird Smart is the evidence layer. The practical layers live elsewhere in the field guide, and they are meant to be read together.
Body language, vocal learning, symbols and buttons, and the honest limits of communication technology.
LIMA-first behaviour science, target training, and reading a cue cluster instead of a symbol dictionary.
Species-by-species profiles, where realistic expectations replace “which parrot is smartest” shopping.
How 150 million years of avian evolution produced a brain that packs primate-scale neuron counts into a small skull.