The Hadza people of northern Tanzania are Africa’s last surviving culture of hunter-gatherers, subsisting mostly on meat, honey, fruit and tubers. In 1985, a team of anthropologists embedded within a band of Hadza foragers noticed that the elderly women in the group were particularly industrious – they spent more time foraging for their families than younger women. The anthropologists concluded that the food the grandmothers gathered contributed to their grandchildren’s survival, thus strengthening the likelihood that their own genetic disposition towards long life would be passed on to future generations.
This observation led the anthropologists to develop the grandmother hypothesis, an explanation for human longevity that has since become one of the most controversial subjects in evolutionary anthropology. While the idea has won a large following, critics argue that it overlooks other factors that could have extended the human life span, including the contributions of other family members and the evolution of increasingly complex brains. Now a recently published paper by one of the founders of the theory is attempting to bolster the evidence with a computer simulation.
At the heart of the controversy is the fact that chimpanzees and other non-human apes live an average of about 45 years, while humans can live almost twice as long. The grandmother hypothesis proposes an explanation: because older women help care for the children of their adult female relatives, they ensure the survival of their own genes – including a tendency toward longevity – by enabling their daughters to have more children closer together without increasing child mortality. Humans wait an average of about three years between children, while the chimpanzee inter-birth interval is four to five years and orangutans eight years.
“You can’t really call women who are past their child-bearing years ‘post-reproductive’ because while they may not be fertile, there is a lot of evidence that they are doing important things for the reproduction of their genes,” said Kristen Hawkes, an anthropologist at the University of Utah who coauthored the grandmother hypothesis.
Primatologists have observed that female chimpanzees typically move to another colony when they reach sexual maturity. Though this is not always the case, adult female chimpanzees and their mothers often live apart, limiting the opportunity for older females to help with grandchildren. In the 15 years since Hawkes and her colleagues proposed the grandmother hypothesis, one of the most common criticisms of the theory has been that it lacks a mathematical proof of how infrequent grandmothering could have moved us from an ape-like lifespan to a human one.
Last October, Hawkes responded to her critics by publishing the results of a computer simulation supporting her hypothesis in the journal Proceedings of the Royal Society B: Biological Sciences. At the beginning of the simulation, grandmothers made up only one percent of female caregivers. But after thousands of simulated years the proportion increased to 43 percent, indicating that the initial presence of a few abnormally long-lived grandmothers can slowly drive up the lifespan over generations.
The simulations show, “That just a little bit of grandmothering is enough to account for the differences in life history between us and our closest living relative,” Hawkes said.
Her finding is consistent with the scientific literature, according to Rebecca Sear, an evolutionary demographist at the London School of Hygiene and Tropical Medicine. “The maternal grandmother consistently came out as positive in helping children, so it is quite possible that the human family has always involved input from post-reproductive women,” she said.
Sear thinks the new computer simulation is unlikely to end the controversy, however. “The problem with mathematical modeling is that the results you get out are only as good as the assumptions you put in, so it’s possible to create a model that will tell you anything you want to hear,” she said.
The grandmother effect kicked off almost two million years ago in the plio-pleistocene era, Hawkes believes, around the same time as the emergence of the genus homo. The dry environment at the time caused the African forests to recede, and those plants that held up well in the arid landscape – tubers, nuts, and hard-shelled seeds – were difficult for children to get at. This situation may have created a unique opportunity for older women to help provide for recently weaned children, permitting their daughters to reproduce sooner after giving birth. The theory is mirrored in the contemporary Hadza, where every nursing woman has a post-menopausal helper.
Grandmothers are unlikely to be the only ones providing support, however. Among the Hadza families Hawkes studied, only four of the eight cases she counted as “grandmothering” came from actual grandmothers. The rest were aunts and other female relatives – expanding the term “grandmother” to include any senior woman who helped raise the children of younger women.
Moreover, in some other hunter-gatherer societies, such as the Ache of Paraguay, grandmothers and other older female relatives are less influential. Most of the Ache diet comes from hunted game, typically produced by fathers. When their fathers die, Ache children are 2.6 times more likely to die themselves than children with living fathers.
“There is lots of evidence for why the grandmother hypothesis is at best incomplete,” said Michael Gurven, an evolutionary anthropologist at the University of California at Santa Barbara. “Men’s food production and contributions to the family are difficult to ignore.”
Gurven proposes that the human lifespan increased because older people were more skilled hunters, an approach known as the embodied capital theory. Since foraging ability doesn’t peak until about age 45, the time required to learn effective hunting strategies gradually prolonged the human lifespan as older, more prolific hunters passed on their genes, he said.
Other researchers also place more weight on alternative explanations for human longevity. Dr. David van Bodegom, a gerontologist at Leiden University Medical Center in the Netherlands, argues that the grandmother hypothesis does not give enough credit to the role of older men in extending the life span. He notes that men who live to be 80 years old have almost 20 percent more offspring than men who die earlier.
“I think one of the problems with the grandmother hypothesis is when it was first coined it was embraced because it was in line with everyday observation,” van Bodegom said. “Nowadays, everywhere we see grandmothers spoiling their grandchildren, so they almost seem programmed for it. It is a very comforting theory.”
Hawkes hopes that her computer simulation will answer some of the issues raised by the theory’s skeptics. Though the model cannot fully account for the complexity of the ancestral environment, she argues it is evidence of how large an influence grandmothers alone had on the human lifespan.
In order to demonstrate the potency of grandmothers, Hawkes and her colleagues purposefully weakened the grandmother effect in their model by programming all women over age 45 to assist with children, regardless of kinship. Had they defined “grandmother” more strictly by blood relation, the results may have been different.
In spite of the importance of supportive fathers in Ache society, London’s Sear notes in her review that fathers generally have little impact on child survival. Instead, other relatives may step in to provide support in the case of an absent father, making the loss less devastating. Maternal grandmothers are important, Sear believes, but she also credits older siblings and other kin.
That communal approach to child rearing is a shared trait from the Hadza to the Ache. Grandmothers, however, provide the most stable support for young mothers, particularly during the stressful time after the arrival of a second or third baby. As Hawkes puts it, “longer lived females can help more, and the females who get a greater benefit from help are longer lived,” and the cycle of grandmothering spins on.