The claim is one of the most durable animal stories in circulation. Elephants hold funerals. They return to the bones of their dead, touch the skulls with their trunks, and stand for hours at the site with no purpose beyond being there. It travels through nature documentaries, wellness essays and social captions, usually as settled fact. Underneath it sits real research, and that research is more careful, and more interesting, than the version in wide circulation.
What follows comes from those papers. The literature is thin and mostly observational, built from a handful of studies and a larger pile of field anecdotes. No single study settles the matter, and there is no consensus that elephants grieve the way people do. What exists is a set of behaviours several teams have recorded consistently, and a set of interpretations that remain open.
What the field observations record
A 2020 review in the journal Primates by Shifra Goldenberg and George Wittemyer is the most useful synthesis, drawing together field observations of elephants at carcasses and adding their own from the Samburu National Reserve in northern Kenya. Three behaviours recur: approaching the body, touching and examining it, and investigating it with the trunk. They appear across every stage of decay, from a fresh carcass to bones bleached by the sun for years.
Their Samburu observations added detail. Elephants visited and revisited carcasses, stood near them, touched themselves and the remains, and in some cases showed temporal gland streaming, a secretion associated with states of arousal or stress. This interest was not confined to close relatives. Elephants attended to the dead of individuals they had not been strongly bonded with in life.
An earlier and much-cited case comes from a 2006 paper in Applied Animal Behaviour Science by Iain Douglas-Hamilton and colleagues, who happened to be tracking elephants in Samburu when a matriarch named Eleanor collapsed and died. A female from an unrelated family tried to lift her with her tusks. Over the following week, five families visited the body, and most of them, the authors noted, were not closely related to Eleanor.
The experiment that complicates the sentimental version
One study bears directly on the individual-mourning reading, and it is a controlled experiment. In a 2006 paper in Biology Letters, Karen McComb, Lucy Baker and Cynthia Moss offered Amboseli elephants a choice of objects: ivory, elephant skulls, the skulls of rhinos and buffalo, and pieces of wood. The elephants clearly preferred the elephant material, and ivory drew the most attention.
What rarely survives into the popular version is the second half of that experiment. When the researchers offered elephants the skull of their own former matriarch alongside the skulls of two unrelated matriarchs, the animals showed no statistically significant preference for their own. Sample size was small, which the authors stated plainly. Even so, the finding does not support the idea that an elephant returns to a skull because it recognises a particular lost individual. Interest in old bones seems to be about elephant remains in general, not one animal’s memory of another.
Where the words carry more than the evidence
Several of the words in the popular account claim more than the research does.
Funeral implies ritual, planning and shared meaning. What the studies measured was approach, contact, duration and revisiting. Silence for hours is selective, since the same literature that describes quiet, stationary investigation also records vocalising, gland streaming and attempts to rouse or lift the body. No apparent purpose other than to be there is a reading, not a measurement. Goldenberg and Wittemyer suggest a plainer possibility, that the intense sniffing and touching is the same information-gathering elephants do when they greet one another, now turned on a body.
None of this makes the behaviour trivial.
Few animals do anything comparable, and elephants do it repeatedly, at length, and toward individuals they are not related to. For now, though, the behaviour is easier to record than to interpret.
The burials, and why they are not yet settled
The newest entry in this literature shows the gap between observation and inference at its clearest. In February 2024, the Journal of Threatened Taxa published a paper by Parveen Kaswan and Akashdeep Roy describing five cases of Asian elephant calves found buried in irrigation drains on tea estates in northern Bengal, India. In each case the calf had been carried some distance and placed legs-upward in the drain, the surrounding ground was compacted, and the herds afterward avoided the paths where the bodies lay.
It is a striking record and worth taking seriously. By the authors’ own account, though, these are carcasses found after the event; no one observed a burial as it happened. The drains were part of the tea estates’ existing irrigation, so the calves ended up in channels that were already dug, not in graves the elephants had made. What supports the reading of intention is inference: calves positioned legs-upward, ground compacted around them by several animals, and injuries suggesting the bodies had been dragged some distance. That evidence is consistent with deliberate burial without settling it.
What the researchers themselves say
Caution tends to come from the scientists themselves; the confidence creeps in through the coverage. Wittemyer has described watching elephants with their dead as behaviour that clearly indicates advanced feeling, while adding that it is something the researchers cannot fully comprehend. He and Goldenberg close their review by calling for further work on elephant memory and on the possibility of grief and emotion, which is where people land when they suspect something real is there and know they have not pinned it down.
Elephants attend to their dead in ways that are unusual among animals, over and over and for long stretches at a time. Two questions stay open: what the investigation is actually for, and whether anyone will one day witness a Bengal burial as it happens.
