What Research Shows About Elephant Intelligence

Elephants show flexible problem-solving, extensive social knowledge, vocal learning, and sensitivity to the behavior of others. Those abilities make them important subjects in comparative cognition, but no single test produces an “intelligence score,” and human labels such as grief, empathy, or language need careful evidence.

The strongest case comes from converging field observations and controlled experiments. Brain size alone is not proof: a large elephant brain also supports a very large body and complex sensory and motor systems. The seven findings below explain what researchers actually tested and what can reasonably be concluded.

Elephants also gather information through a highly specialized trunk. TipsMake’s article on elephant trunk biomechanics looks at how this organ manipulates food and water.

1. African elephants distinguish information in human voices

A 2014 playback study tested whether wild African elephants could extract risk-relevant information from recorded human speech. The recordings varied the speaker’s ethnic group, sex, and age. The elephants responded more defensively to voices associated with the human group and demographic considered more threatening in that population’s experience.

The result does not mean elephants “understand languages” as humans do. It shows that they can learn fine acoustic distinctions in human voices and connect those cues with different levels of risk. The experiment and its statistical analysis are available in the PNAS study on human voice cues.

2. An Asian elephant used a movable object as a tool

In a controlled problem-solving study, a young Asian elephant named Kandula faced food suspended beyond trunk reach. After earlier unsuccessful approaches, he moved a cube beneath the food, stood on it, and reached the reward. He later moved the cube to other locations and used additional objects in related situations.

This mattered because the solution required treating the object as a movable platform. It also illustrated why an elephant may fail a primate-style tool test: the trunk is both a manipulator and a major sensory organ, so asking an elephant to hold a stick in its trunk can interfere with smelling and exploring the target. The original PLOS ONE problem-solving study describes the trials and limitations.

One individual’s success is evidence of a capacity, not proof that every elephant will solve the same task in the same way. Age, prior experience, motivation, test design, and access to objects all matter.

3. African elephants can follow a human pointing cue

Researchers tested 11 African elephants with an object-choice task: food was hidden in one of two containers and a person pointed toward the correct one. The elephants selected the indicated container more often than expected by chance from the start, despite not having been trained specifically to follow pointing.

Follow-up conditions varied how obvious the gesture was and where the person stood. The results support the conclusion that elephants can use a human’s directional gesture as useful information, not merely walk toward the nearest container. The Current Biology pointing study does not establish human-like language comprehension; it demonstrates flexible interpretation of a communicative cue.

4. Asian elephants respond affiliatively to distress

A systematic observational study of captive Asian elephants recorded what happened when one elephant displayed distress. Bystanders became more likely to make affiliative physical contact, including trunk touches, and to produce particular vocalizations. The researchers described the behavior as reassurance and discussed its similarity to consolation-like responses studied in some primates.

That is stronger evidence than a single touching anecdote, but it still requires restraint. Behavior can be documented directly; an animal’s private emotional experience cannot. The authors themselves used operational definitions and did not claim that an elephant experiences empathy in exactly the human sense. See the open PeerJ study of distress-related reassurance.

5. Elephants show unusual interest in dead elephants

Field observations have documented elephants approaching, touching, smelling, and remaining near dead members of their species. Experiments with skulls and ivory also found stronger interest in elephant remains than in some other large-animal remains. More recent collections of observed cases report repeated touching and investigation of dead calves.

These behaviors are consistent with close social bonds and sensitivity to death-related cues. They do not, by themselves, prove that elephants hold a human concept of death, conduct funerals, or always “mourn” in the same psychological sense as people. Claims that elephants routinely bury bodies or react only to their own species should therefore be treated cautiously. A Biology Letters experiment on elephant remains provides the controlled evidence behind this finding.

6. Koshik learned to imitate human speech sounds

Koshik, a male Asian elephant in South Korea, produced sounds that Korean listeners identified as five Korean words. He achieved the unusual acoustic match by placing his trunk tip inside his mouth and changing the shape of his vocal tract.

The researchers presented this as vocal production learning: Koshik altered his calls to resemble sounds from his human social environment. They did not show that he understood or used the five words with their ordinary meanings. Vocal imitation is cognitively notable, but it is not equivalent to speaking a human language. The recordings and listener tests appear in the Current Biology report on Koshik.

7. Elephant memory supports a large social network

Field playback work shows that female African elephants distinguish familiar from less familiar contact calls across a wide network. One study estimated that the tested females needed familiarity with calls from roughly 100 adult females across about 14 families to produce the observed pattern of responses. That is evidence for extensive social recognition, not a claim that elephants remember every event forever.

Related experiments found that groups led by older matriarchs responded more appropriately to calls from unfamiliar elephants. The likely advantage is accumulated social knowledge: a long-lived leader has more experience distinguishing routine contact from a possible threat. The Science study on matriarchal social knowledge links this experience to group decision-making.

Ecological memory is also important. Elephants move across large areas and must integrate changing information about water, forage, seasonal conditions, people, and other elephants. Specific stories about remembering a person or a water source can be plausible, but carefully designed studies are needed to separate memory from scent, repeated routes, social learning, or other cues.

What these studies do—and do not—show

• They show cognitive flexibility. Elephants can solve some novel physical tasks, learn meaningful risk cues, and adjust behavior to social information.
• They show specialized intelligence. Their abilities reflect elephant bodies, senses, ecology, and social lives rather than a human-style test hierarchy.
• They show individual and contextual variation. A result from one elephant, one population, or one captive group should not be generalized without further evidence.
• They do not justify legends as facts. Stories about disabling bells, saving particular animals, holding funerals, or displaying moral judgment need verifiable observations before being presented as evidence.
• They do not make conservation needs dependent on a ranking. Animals do not have to perform well on human-designed tasks to merit welfare and habitat protection.

Calling elephants “intelligent” is therefore reasonable when it summarizes a broad research record. The useful question is not whether they think exactly like humans, but how their memory, communication, problem-solving, and social learning help them navigate an elephant’s world.