Skip to content
NatureHQ

Behaviour and cognitionability

Social learning

Teaching has a three-part definition that says nothing about intending to teach. An ant meets all three parts.

Learning from others is everywhere; imitation is rarer than people assume, and teaching is rarer still. Most of what looks like copying is something simpler — an animal drawn to where another animal is, working the rest out itself. But a few species genuinely do teach, meerkats and tandem-running ants among them, and the ant is the uncomfortable one.

Watch a young animal acquire a skill after watching an adult and it is natural to describe it as copying. Usually it is not. Behavioural biologists separate several mechanisms that all look the same from a distance, and the differences matter because they demand quite different things of the animal. Local enhancement: one animal is simply drawn to where another one is, and discovers the food itself. Stimulus enhancement: attention is drawn to a kind of object rather than a place — a bird sees another bird pecking a milk-bottle top and afterwards finds bottle tops interesting. Emulation: the learner sees the result and works out its own way to get there. True imitation, copying the specific actions, is much less common than the word suggests and is genuinely hard to demonstrate. Then there is teaching, which reverses the direction: the knowledgeable animal changes its own behaviour, at a cost, for the benefit of the learner. The functional definition requires exactly three things — behaviour modified in the presence of a naive individual, a cost to the teacher, and faster learning by the pupil — and deliberately says nothing about intention. That is why an ant qualifies. A leader ant slows when the follower loses contact and speeds up when contact resumes, takes four times longer than it would alone, and the follower afterwards finds the destination far faster. All three conditions, in an insect. Either the definition is capturing something that does not need much of a brain, or it is capturing less than the English word implies, and both readings are worth sitting with.

Developed coverage · 77% complete · reviewed 2026-08-31

What this page covers

Learning from others is close to universal in animals with any social life. Teaching, in the technical sense, is documented in a handful of species — including an ant.

Often confused with: Imitation, which is only one of several mechanisms; Teaching

Quick facts

The teaching definition
Behaviour modified for a learner, at a cost, and the learner learns faster
Species that meet it
Meerkats and tandem-running ants, among a small number
What intention has to do with it
Nothing — the definition is deliberately silent on mental states
Most social learning
Is not teaching, and is not imitation either

Four things that look like copying

They demand very different things of the animal doing them.

MechanismWhat actually happensWhat it requires
Local enhancementOne animal goes where another is, and finds the food for itselfAttraction to another animal. Nothing else.
Stimulus enhancementAttention is drawn to a type of object rather than a placeNoticing what another animal is handling
EmulationThe learner sees the outcome and finds its own route to itUnderstanding that the goal is achievable
ImitationThe learner copies the specific actions, not just the resultMapping another body’s movements onto its own — rarer than assumed

The reason to keep these apart is that the simplest one explains most field observations, and it explains them without crediting the animal with anything. A behaviour spreading through a population is evidence of social learning of some kind. It is not evidence of imitation, and the two are routinely reported as though they were the same.

Do animals teach each other?

Yes — and the definition that lets them is stranger than the answer.

The short answer

Do animals teach their young?

A few genuinely do, by a definition that requires the teacher to change its behaviour for the learner at a cost to itself. Meerkats do it with scorpions. So do ants, which is the part that should give anyone pause.

Meerkat pups cannot handle a live scorpion. Adults hand the youngest ones dead prey, older ones prey with the sting removed, and eventually live prey — grading the difficulty as the pup develops. That alone could just mean adults recognise pup age. The experiment that settles it is a playback: play the call of an older pup to a group that only has young ones, and the adults start delivering live prey. The adjustment is driven by a signal from the learner, which is what makes it teaching rather than a by-product of feeding. Ants clear the same bar by a different route. In tandem running one ant leads another to a new nest site; the leader’s speed is controlled by how often the follower taps it with its antennae, so the follower sets the pace. The leader takes about four times longer than it would travelling alone, and the follower afterwards reaches the site far faster than an ant that was never led. Modified behaviour, a cost, faster learning: all three. Nobody claims the ant knows the other ant is ignorant. The definition was written to be measurable rather than mentalistic, and this is the price of that choice.

Check it for yourself

A few animals genuinely teach — they change what they do for a learner, at a cost to themselves, and the learner acquires a skill faster. Ants are among them.

Well supported

Good evidence backs this, though some details remain open.

Teaching, defined functionally as behaviour modified in the presence of a naive individual at a cost to the actor, which accelerates the naive individual’s learning, is documented in meerkats and in tandem-running ants. The definition makes no requirement about the teacher’s mental states.

Who this applies to
A small number of species meet the functional criteria. Most social learning in animals involves no teaching at all.Do not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Animalia
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

Both key cases include a manipulation rather than observation alone: playback of pup calls in meerkats, and measurement of leader speed against follower contact in ants. Both show the demonstrator adjusting to the learner.

How far it can be extended

Only a handful of species have been tested against all three criteria, and most social learning proceeds without any modification of the demonstrator’s behaviour.

Caveats

  • The functional definition deliberately excludes intention. An ant meeting it does not mean an ant intends to instruct.
  • That definition is contested precisely because it admits cases many people would not call teaching.
  • Most animal social learning — copying, being drawn to where others are feeding — involves no teaching whatever.

Still unanswered

  • Whether any non-human teaches with an understanding that the other does not know something.
  • Whether the functional definition should be tightened, and what would be lost if it were.

Last reviewed 2026-08-31

The evidence (2 studies)

  • Supports · primary

    Teaching in Wild Meerkats

    Thornton and McAuliffe, 2006 · Science

    Adults graded live, disabled and dead prey to pup age, and changed what they delivered when played the call of an older or younger pup.

  • Supports · primary

    Teaching in tandem-running ants

    Franks and Richardson, 2006 · Nature

    The leader’s pace is controlled by the follower’s antennal contact, the leader pays a fourfold time cost, and the follower subsequently finds the destination far faster.

How we know

Playing the call of an older pup, and watching the adults change the menu

When adult meerkats grade the prey they give pups, are they teaching, or simply reacting to pups they recognise?

Wild meerkat groups with pups of known age were observed while adults provisioned them, and the state of each scorpion recorded — dead, disabled with the sting removed, or intact and alive. Then calls were played back: the begging call of an older pup broadcast to a group that had only young pups, and the call of a younger pup to a group with older ones.

What happened

Adults delivered dead prey to the youngest pups, disabled prey to older ones, and live prey later. On playback of an older pup’s call to a group with young pups, adults began delivering more live prey; the reverse manipulation had the reverse effect.

What it shows

The adults are adjusting to a signal from the learner rather than to the learner’s identity, and the adjustment makes the pup’s task progressively harder as it becomes capable of more. That satisfies the functional definition of teaching: behaviour modified for a naive individual, at a cost, with faster learning as the result.

What it does not show

It shows nothing about what the adult understands. The functional definition is deliberately silent on mental states, and no meerkat is being credited with knowing that a pup does not know something. Whether pups would eventually reach the same competence without the grading is not directly measured, and this is one species with an unusually dangerous prey problem.

The controls — what makes this evidence rather than a story
  • Pup age known independently, so the natural gradation could be measured before any manipulation.
  • The playback substitutes the acoustic cue while leaving the actual pups unchanged, separating what adults hear from what is in front of them.
  • Both directions tested — older call to young group and younger call to old group — so the effect cannot be a general response to playback.

From Teaching in Wild Meerkats

Learning how to say something

Social learning shapes not only what animals do, but how they signal.

The waggle dance was the standard example of a wholly innate signal — an insect performing a complex encoding with no opportunity to have learned it. That turns out to be about three-quarters right. Bees raised in colonies with no experienced dancer to follow still dance, but with larger directional errors, and a disordered encoding of distance that they never entirely fix. Following an experienced dancer is part of learning to dance properly, which puts a bee in the same broad category as a songbird learning its song.

Bees raised with no experienced dancer to follow still dance — but they get the direction wrong more often, and never fully fix their distance signalling.

Emerging evidence

Real findings exist, but too few or too recent to be settled.

Honey bees reared in colonies containing no older, experienced dancers perform waggle dances with significantly larger directional divergence errors and disordered distance encoding. Directional error improves with practice; the distance encoding deficit persists.

Who this applies to
One species, in colonies experimentally composed of same-aged bees, which does not occur naturally.
Studied in
Apis mellifera
Why we rate it this way, and what the caveats are
Emerging evidenceModerate confidence

A well-designed manipulation with a clear result, from one laboratory. It overturns a long-standing assumption, which is a reason for interest and also a reason to want replication.

How far it can be extended

A single species under a manipulated colony structure; other Apis species have not been tested this way.

Caveats

  • The dance remains substantially innate — bees with no model still danced, and still encoded direction and distance.
  • This does not make the dance a language; it moves where the line between instinct and tradition sits.
  • Which aspect of following a dancer is doing the teaching is unidentified.

Still unanswered

  • Whether the same holds in other honey bee species.
  • Whether colonies with persistent distance errors forage measurably worse.

Last reviewed 2026-08-31

The evidence (2 studies)

What this builds into

Follow it further

  • Animal culture

    Social learning that persists long enough to become a tradition

  • Honey bee

    The dance, the brake on the dance, and the part that has to be learned

  • Crow

    Learning which humans are dangerous from birds that were never caught

  • Animal intelligence

    Where social learning sits among the other abilities

Claims about this, checked

Things people have heard, and what the evidence actually supports.

The research behind this page

9 studies, newest first. Each one has a page explaining what it found and what it could not show.

2023Science

Social signal learning of the waggle dance in honey bees

Bees with no opportunity to follow experienced dancers still danced, but with persistently larger directional errors and disordered encoding of distance.

2014Nature

Experimentally induced innovations lead to persistent culture via conformity in wild birds

The trained solution spread through each sub-population and remained the dominant method, persisting across years and through substantial population turnover.

2013Science

Network-Based Diffusion Analysis Reveals Cultural Transmission of Lobtail Feeding in Humpback Whales

A new technique — striking the surface with the tail before diving to feed — spread through the population in a pattern that followed the social network closely, appearing first in animals that associated with whales already using it.

2006Science

Teaching in Wild Meerkats

Adults gave dead prey to the youngest pups, disabled prey to older ones, and finally live prey.

2006Nature

Teaching in tandem-running ants

Leaders slowed when the follower fell behind and sped up when tapped, adjusting their pace to the follower.

1999Nature

Cultures in chimpanzees

Thirty-nine behaviours, many of them involving tools, varied between sites in ways that ecology did not explain.

1978Science

Cultural transmission of enemy recognition: one function of mobbing

Observer birds began mobbing the harmless object, and passed that response on to further birds through a chain of several individuals, with the acquired response persisting through the transmission chain.

1978Zeitschrift für Tierpsychologie

The adaptive significance of avian mobbing

Mobbing serves several non-exclusive functions, and the balance differs by species and context.

1967Harvard University Press

The Dance Language and Orientation of Bees

Returning foragers perform a repeated figure-of-eight "waggle" run on the comb.

This page is a stop on a longer route

A guided journey reads several subjects in a deliberate order, with an argument for why one follows another. You can join in the middle.

Where to go from here

Each of these follows from something on this page — a relationship in the evidence, a claim people ask about, or the next mechanism along.

How complete this page is, and what it is still missing

NatureHQ publishes its own gaps. This page is at 77% completeness against what we would call a finished subject, and was last reviewed on 2026-08-31. It carries 4 claims and answers 2 mapped search questions.

  • The imitation literature — what counts, and how rare it really is — deserves fuller treatment than one table row.
  • Social learning in fish and invertebrates beyond ants is barely represented.
  • Conformity biases are described for great tits and not examined elsewhere.