Two days after her daughter died, Vuavua was still waving the flies away from the body.

She had been carrying it since the morning it happened, a small limb gripped in one hand as she walked, and she would keep carrying it for another seventeen days. Field researchers in the forest around Bossou, in south-eastern Guinea, filmed the lot.

What happened at Bossou

A respiratory illness swept the Bossou community in 2003 and killed five chimpanzees, two of them infants. Veve was two and a half years old. Jimato was just over one.

Writing in Current Biology, Dora Biro and colleagues logged what the two mothers did next. Vuavua kept Veve for 19 days. Jire kept Jimato for 68. Both bodies dried out completely and mummified into something closer to leather than to an infant, and the carrying went on regardless.

Grooming went on too. Both mothers groomed the bodies regularly and chased off the flies that circled them, twice using a tool to do it. Other chimpanzees sniffed the corpses, poked at them, and lifted their limbs to let them drop again. Juveniles were occasionally allowed to carry a body a short distance, in what looked like play.

Almost nobody in the group flinched, even once the smell turned overpowering.

Grips that living infants never get

An easy reading of all this is that the mother has simply not registered the change, and is minding a baby she expects to wake up. Susana Monsó and Antonio J. Osuna-Mascaró, both at the Messerli Research Institute in Vienna, argue in a paper published in Synthese that this asks too much of a mother’s obliviousness. A dead infant is cold, limp, unresponsive, silent and eventually rank, and a wild chimpanzee has handled plenty of live ones for comparison.

They lean on one detail from the Bossou footage in particular. Those mothers carried the bodies using holds they never use on a living infant, which suggests they had already reclassified what they were holding.

How common the behaviour actually is

Is Bossou a freak case? Partly. A 40-year record from Uganda’s Budongo Forest, published in the journal Primates, covers 191 births and 68 infant deaths. Researchers logged twelve carrying episodes among eight mothers. Nine ran one to three days, usually until the body began to decompose. Three ran far longer: 18, 56 and 89 days. Carrying that lasts weeks is rare even where the behaviour happens at all.

The Budongo team runs through the competing explanations, and there are plenty. One holds that carrying sharpens maternal skill, which would predict it mostly in first-time mothers. Another ties it to the strength of the bond. A third proposes that mothers allowed to keep the body show lower levels of stress hormones, an idea that has proved awkward to test in a forest.

The broadest survey to date pooled 409 cases across 50 primate species, drawn from 126 published papers. In Proceedings of the Royal Society B, Elisa Fernández-Fueyo and her co-authors at University College London found two patterns worth chewing over. Infants who died young were carried longest, with a sharp drop-off once they reached roughly half of weaning age. And carrying was more likely when death came slowly from illness than when it came suddenly, through an accident or infanticide.

That second pattern cuts against the tidy grief story. Fewer sensory cues meant more carrying, which is exactly what the death detection hypothesis predicts: that some of the behaviour depends on how detectable a death was in the first place.

Why nobody has settled it

Christophe Boesch, who has watched similar scenes at Taï Forest in Côte d’Ivoire, told Science that researchers have “certainly underestimated awareness of death in chimpanzees”. A review by the primatologist James Anderson lands somewhere cooler. He argues chimpanzees likely do register a death as final and as the end of function, but doubts they grasp that every creature is bound to die, that their own death is inevitable, or the deeper biological causes behind it.

Frans de Waal offered the least sentimental account of all, in coverage by National Geographic. A chimpanzee’s reproductive cycle stalls for about four years after a birth, and a female who wrote off her infant at the first sign of illness would be a poor bet evolutionarily. That kind of attachment does not switch off on command.

Every one of those readings fits the same footage, which is precisely the trouble. Sixty-eight days of it is compatible with grief, with hormones, with confusion, and with a mother who understands perfectly well what has happened and is not prepared to set it down.

Chimpanzees have presumably been doing this since long before anyone thought to write it up. What is new is the species standing at the edge of the clearing with a notebook, trying to work out what it is looking at.