AI-assisted research; animal-behaviour specialist review pending. Wild studies are identified by population and publication date. The accompanying film shows historical zoo care, not field evidence from Sri Lanka. Detailed kinship genetics, calf development, olfactory and seismic communication remain research gaps. No claim is made that elephant calls can be translated into human sentences.
A group seen today is not the whole social network
Fission–fusion describes animals separating into smaller groupings and later joining others. A 2011 study by de Silva, Ranjeewa and Kryazhimskiy followed individually recognised female Asian elephants at Udawalawe. Although the groups seen together were usually small, repeated observations revealed enduring relationships within a much wider set of associates. The central finding is that changing company does not necessarily mean weak or random social relationships.
Long-term companions within a fluid society
Some associations strengthened during dry seasons, but individuals also retained long-term companions across seasonal changes. Population-level patterns did not simply reproduce the patterns found between pairs. This is why one photograph of a cluster cannot establish its permanent membership or identify a leader. The findings concern this studied population: they do not make Asian and African elephant societies interchangeable, or establish that every pair seen together consists of close relatives.
Beyond the trumpet: describing an acoustic repertoire
De Silva’s 2010 study of Sri Lankan Asian elephants classified at least fourteen vocal types using acoustic features: eight single calls, five combination calls and one possibly distinctive male call. Certain low-frequency calls differed among individuals, and call production varied with social context, age and sex. The categories describe patterns in recorded sound; they are not fourteen translated words. A useful distinction is between classifying a signal, identifying its caller and demonstrating what a listener understands. The paper’s observations support a varied communication system, not a universal phrasebook for interpreting any elephant recording.
How field recordings become research material
The University of Pennsylvania’s Linguistic Data Consortium catalogue describes 57.5 hours of Udawalawe elephant recordings, with 31.25 hours annotated. The collection mainly features adult females and juveniles, with field notes made by Shermin de Silva and Ashoka Ranjeewa. Annotation connects sound to observations rather than leaving researchers with isolated clips. The catalogue identifies possible regional dialects as a research interest, not a demonstrated conclusion of the collection. It also specifies licensing arrangements: an audio dataset being described online does not make its recordings freely reusable. This article links to the catalogue rather than republishing its audio.
Compare like with like
Across these studies, ‘social behaviour’ can mean several different measurements: repeated association between known individuals, the composition of a group, observed interactions, or the acoustic structure of calls. Those measurements answer related but different questions. Association is not itself proof of genetic kinship; a recording from one population is not a census of every call used throughout Asia. When comparing results, check the location, animals sampled, observation period and measurement method before treating an apparent difference as a contradiction. These distinctions help explain why long-term field research adds information that casual sightings cannot supply.
Watch a keeper’s account—with the setting made explicit
The accompanying Auckland Zoo film introduces Anjalee and Burma in a historical keeper-chat episode. Its article discusses play, reaching for food, dust bathing and relationships as elements of managed care. It offers visible examples for thinking about behaviour, but it is not footage from the Udawalawe or Wasgamuwa studies, an assessment of current housing, or an invitation to approach wild elephants. Publisher, title and thumbnail were checked through YouTube metadata. The narration has not been independently transcribed and fact-checked for this article.
Keeper Chat — Andrew introduces our Asian elephants!
Historical captive-care film in New Zealand. Not footage of the Sri Lankan field studies, current zoo arrangements or wild-elephant handling guidance.
Watch on YouTube ↗Follow the evidence
de Silva, Ranjeewa and Kryazhimskiy · BMC Ecology
LaDue et al. · Animals
Shermin de Silva · Behaviour / ResearchGate
Linguistic Data Consortium, University of Pennsylvania
Auckland Zoo
Sources consulted during this AI-assisted update: 8 September 2026. Source labels describe provenance, not independent fact-check approval. This article remains open to correction.