Behavior by animal group

Parrot learning

Parrots (the order Psittaciformes) are among the animals most often described as "talking" birds, and this reputation makes them a popular but easily misunderstood case study in animal learning. They belong to a group of vocal learners — animals that can modify the sounds they produce based on what they hear — and some species reproduce human speech sounds and household noises with striking fidelity. That ability is real and well documented in particular species and individuals, but it is frequently overstated. This guide gives a broad, source-cautious overview of what learning in parrots involves and where the evidence stops.

Two cautions run through the whole page. The first is that reproducing speech sounds is not the same as using human language: a parrot copying words is engaged in vocal learning and mimicry, not in the open-ended grammar that defines human speech. The second is that parrots are a large and varied order, so a famous individual or a single well-studied species cannot stand in for them all. Much of what is most often quoted comes from captive birds, and captive findings do not automatically describe wild populations. This is a description of how learning is studied in parrots, not advice on keeping, training, taming, or breeding any bird.

A cautious group-level overview of how parrots learn vocally and socially, why their vocal learning is not human language, and why famous individuals and the diversity of the order resist sweeping claims.

Representative, not complete:

This page describes representative, well-studied examples of learning in parrots, not the whole order. Psittaciformes contains hundreds of species with very different vocal and social behaviour, and much of the most-quoted evidence comes from a handful of captive individuals or species. What is true of an African grey in a lab, or of one famous bird, is not automatically true of macaws, lorikeets, the kakapo, or wild populations of the same species.

Representative behavior themes

  • Vocal learning and the capacity to imitate soundsEvidence: Mixed evidence

    Parrots are open-ended vocal learners: many can modify the calls they produce throughout life rather than being fixed to an inherited repertoire. In captivity some species, such as the African grey parrot (Psittacus erithacus) and various Amazon parrots (Amazona), reproduce human words and other novel sounds, and certain individuals build sizeable repertoires. This is described cautiously as vocal imitation and learning, not as language, and it varies widely between species and individuals.

  • Social learning of contact calls and group "dialects" in the wildEvidence: Wild study

    In wild populations of several species, vocalisations are learned socially and can vary between regions and groups, producing what researchers describe cautiously as local call variants or vocal dialects. Studies of some wild parrots and parakeets report that individuals learn contact calls used to recognise group members or mates. This is socially learned variation, documented in particular species, and should not be read as a property of every parrot.

  • Imitation, problem-solving, and the limits of inferring itEvidence: Controlled study

    Some parrots have been observed in controlled studies solving novel physical tasks and, in a few experiments, matching aspects of a demonstrator's actions. Researchers are careful to separate true imitation (copying a specific action) from simpler routes such as trial and error or being drawn to the same object, and they build in controls. Reported abilities apply to the tested individuals and species, not to the order as a whole.

  • Famous individuals are not proof of species-wide abilityEvidence: Captive study

    Widely publicised cases of single parrots with large vocabularies or apparent label use are genuine observations, but a remarkable individual demonstrates what one bird did under particular conditions, often after years of intensive human interaction. It does not establish that the whole species, let alone all parrots, shares that ability. Careful writers attribute such results to the individual and study rather than generalising them.

  • Wide diversity across the order — "talking" is not universalEvidence: Broad-group pattern

    Psittaciformes includes hundreds of species across very different groups — from large macaws and cockatoos to lorikeets, lovebirds, parakeets, and the flightless, nocturnal kakapo (Strigops habroptilus). Vocal imitation of human speech is conspicuous in some species and minimal or absent in others, so "parrots talk" is an overgeneralisation. The group is better described as containing diverse vocal learners with very different tendencies.

Vocal learning is not human language

Parrots are vocal learners, meaning many of them can change the sounds they make based on what they hear, rather than being limited to a fixed, inherited set of calls. This places them, alongside songbirds, hummingbirds, and a few mammals, among the relatively small number of animals with this capacity. In captivity, some parrots reproduce human words, whistles, doorbells, and other novel sounds, and that imitation can be remarkably accurate. The ability itself is well documented and genuinely unusual among animals.

It is important, though, to describe this precisely. Copying speech sounds is vocal imitation and learning; it is not the same as using human language. Human language is defined by open-ended productivity and grammar — the capacity to combine a limited set of elements into an essentially unlimited number of new, meaningful utterances. Reproducing or recombining learned sounds, even flexibly, does not demonstrate that capacity. Some studies of particular individuals have explored whether trained parrots can use sounds as labels for objects or categories, and those results are interesting and debated, but they are reported as findings about specific birds under specific conditions, not as evidence that parrots in general possess language.

For this reason careful writers avoid phrasing like "parrots speak like humans." The accurate description is that some parrots are skilled vocal imitators that can learn and reproduce human speech sounds, while the deeper question of meaning and comprehension is studied narrowly, hedged heavily, and not settled by a bird's ability to produce the right noises.

Social learning, contact calls, and local variation in the wild

Away from the popular image of a caged bird saying words, much of the most informative research on parrot learning concerns how wild birds learn from one another. In several species, vocalisations such as contact calls — the sounds used to keep in touch within a flock — are learned socially rather than fully inherited. Studies of some wild parrots and parakeets have documented that individuals can converge on shared calls with group members or mates, and that calls vary between regions, producing patterns researchers cautiously describe as local variants or vocal dialects.

This kind of socially learned, locally varying behaviour is sometimes discussed under the careful, technical sense of "culture" used in animal biology: traditions passed between individuals by learning rather than inherited directly. That is a deliberately modest claim. It does not imply anything resembling human institutions, and it is documented for particular species and populations rather than for the order as a whole. Describing a flock's shared calls as learned variation is supported by observation; describing them in human social terms is not.

Because these findings come from specific wild populations, they are reported as such. A dialect documented in one species in one region tells us about that species there; it is not a general law of parrot behaviour, and other parrots may organise their vocal lives quite differently.

Why the diversity of parrots resists tidy claims

Psittaciformes is a large order spanning very different birds: macaws, cockatoos, Amazon parrots, African greys, lovebirds, lorikeets, budgerigars and other parakeets, and unusual outliers such as the flightless, nocturnal, ground-dwelling kakapo (Strigops habroptilus) of New Zealand. These birds differ in body, ecology, social structure, and vocal behaviour. Conspicuous imitation of human speech is associated with some species and is minimal or unreported in others, so a blanket statement that "parrots talk" misrepresents the group.

The same caution applies to learning more broadly. Some species have been the subject of controlled studies on problem-solving or imitation; many have barely been studied at all. When a striking ability is noted — a large learned vocabulary, success on a novel physical task, or apparent matching of a demonstrator's actions — it is most accurately attributed to the particular individuals and species tested, frequently in captivity after extensive human contact. It is not evidence of a uniform, order-wide intelligence, and this guide gives no ranking or score.

The honest summary is that parrots are a diverse group of capable vocal learners whose abilities are unevenly distributed and unevenly studied. Representative, well-documented examples can illustrate what some parrots do; they cannot define what all parrots are.

What this page does not claim

  • That parrots speak or understand human language — reproducing speech sounds is vocal mimicry and learning, not the open-ended grammar of human language.

  • That all parrots can imitate human speech, or that "talking" is a universal trait of the order; many species show little or no speech mimicry.

  • That a famous individual parrot's vocabulary or apparent comprehension proves a species-wide or group-wide ability.

  • That captive or laboratory findings describe how wild parrots learn or behave in their natural populations.

  • That this page offers any advice on keeping, taming, training, hand-rearing, feeding, or breeding parrots, or on teaching a bird to talk.

Related animal profiles & behavior pages

Species behavior profiles

Animal profiles

How these claims are studied

Group-level behaviour is easy to overstate, so these claims are kept cautious and labelled by evidence context. See communication vs language, captive bias in behavior research, and animal research sources for our methodology.

Frequently asked questions

Can parrots actually talk and understand what they say?
Some parrots are skilled vocal imitators that can learn and reproduce human words and other sounds, which is genuinely unusual among animals. But reproducing speech sounds is vocal mimicry and learning, not human language, which is defined by open-ended grammar. Whether particular trained individuals use sounds as labels for objects has been studied and debated, but those are findings about specific birds in specific conditions, not proof that parrots in general understand language.
Do all parrots learn to talk?
No. "Parrots talk" is an overgeneralisation. _Psittaciformes_ contains hundreds of species — macaws, cockatoos, lorikeets, parakeets, the kakapo, and many more — and conspicuous imitation of human speech is associated with some species and is minimal or absent in others. The order is better described as containing diverse vocal learners with very different tendencies rather than as a group of universal talkers.
Does a famous talking parrot prove the whole species is that capable?
No. Widely publicised cases of individual parrots with large vocabularies or apparent label use are genuine observations, but a remarkable individual shows what one bird did, usually after years of intensive human interaction in captivity. It does not establish that the whole species, let alone all parrots, shares that ability. Careful descriptions attribute such results to the individual and study rather than generalising them.
Why do studies stress that parrot findings come from captivity?
Much of the best-known research on parrot vocal and problem-solving abilities involves captive birds, often after extensive human contact. Behaviour in a home or laboratory can differ from behaviour in the wild, so captive findings are not automatically representative of wild populations. Researchers report results under the conditions in which they were observed and avoid assuming that captive abilities describe how the same species learns and communicates in nature.