Why our common ancestor with chimpanzees keeps getting further away

West African chimpanzees

INTERFOTO/Alamy

This is an extract from Our Human Story, our newsletter about the revolution in archaeology. Sign up to receive it in your inbox every month.

Chimpanzees, along with bonobos, are our closest living relatives. Of all the millions of other animals on Earth, they are the ones that are most like us. And the more we learn about chimpanzee behaviour, the more obvious the similarities become.

For instance, wild chimpanzees peer at each other: one chimpanzee will sit close to another and closely watch what they’re doing. Nora Slania at the Max Planck Institute of Animal Behavior in Germany and her colleagues found that, from a young age, chimps peer at others as they perform a range of tasks. This included complex tasks like making tools, but also supposedly simple things like feeding and grooming.

The implication is that a great deal of chimpanzee behaviour is culturally learned by watching others. Obviously, chimps will instinctively try to eat food, but it’s not obvious what is good to eat, or how to peel a fruit or crack a nut. One way to learn is by watching how the adults do it.

Some chimpanzees have figured out how to eat army ants, which form huge colonies comprising millions of individuals. When the ants need to eat, they assemble into marching columns guarded by soldier ants with powerful jaws. While there is obviously plenty of nutrition in an army ant colony, one must be careful when dealing with them.

In Senegal, chimps catch Dorylus army ants living on the savannahs. Andreu Sánchez-Megías at the University of Barcelona in Spain and his colleagues argue that chimps vary their strategies depending on how aggressive the ants are. Army ants that hunt in leaf litter are not very aggressive, so chimps sometimes forage for them by hand. In contrast, army ants that hunt on the surface of the ground are aggressive, so the chimps catch them by lowering a stick, which the ants bite onto – allowing the chimp to eat them in relative safety. What’s more, the chimps prefer to catch these species while standing upright, or on a raised platform, like a bent tree branch. These are all savvy ways to minimise your chances of being bitten.

One more example: centuries ago, chimpanzees in Sierra Leone may have used tools to forage for oysters. They don’t do so today, because chimps no longer live on the Sierra Leone river estuary, so there’s nowhere for them to do it. But reports from 1678 and 1682 describe apes – whose description best matches chimps – using sticks to dislodge large oysters. If this is true, it adds to the evidence that chimps sometimes go, for want of a better word, fishing. Chimpanzees in the Nimba mountains of Guinea have similarly been seen fishing for crabs. Other populations have developed a host of methods for gathering algae.

These behaviours are creative and flexible, requiring careful observation and extrapolation – all traits we associate with ourselves. They are a reminder of how close chimpanzees are to us. So it is curious that, in parallel, genetic and fossil evidence has gradually pushed chimpanzees further away from humans.

The family tree

Replica of Sahelanthropus skull

Philippe Clement/imageBROKER/Shutterstock

I’m talking about the last common ancestor of humans and our two closest living relatives, chimpanzees and bonobos. We know this ancestral animal was an ape, and we know it lived millions of years ago, but we don’t know what it was like, or exactly where and when it lived.

In the timeline of human evolution I recently compiled, I gave a wide date range for the last common ancestor – and for many other key events in prehistory. That’s because many of these things can’t yet be precisely pinned down. Often, different lines of evidence – for instance, physical fossils or genetics – give different answers.

Hence, I put the last common human-chimpanzee ancestor sometime between 13 million and 5.5 million years ago. That is a range of 8.5 million years, which, in this context, is pretty wide. For comparison, the oldest putative fossil hominin is Sahelanthropus from about 7 million years ago – so the uncertainty I gave for the last common ancestor spans more time than the known hominin fossil record.

Somehow, an ancient population of apes became divided, with one group going on to become modern chimpanzees and bonobos, and the rest becoming humans and all other hominins. Here, I’d like to drill into exactly when and how this happened. I’m basing this on a study published in the journal Primates on 18 June, which synthesises all the estimated dates for the split between humans and apes. 

Lead author João d’Oliveira Coelho is a biological anthropologist at Gorongosa National Park in Mozambique. A few years ago, he was writing about his research and wanted to include a quick sentence stating when the hominin lineage began, so he looked around and found a smorgasbord of estimates – and some peculiar statements. Textbooks, even quite recent ones, often put the human-chimp split as recent as 4 million years ago, but would then mention the existence of significantly earlier hominins like Ardipithecus, Orrorin and Sahelanthropus. “How can you have both these statements in the same book? They are completely contradictory,” says d’Oliveira Coelho.

Estimates of the timing of divergence between hominins and the bonobo-chimpanzee clade, published over a 56-year period. There is an early stage with much uncertainty (1960–1980s) due to the large disparity of values and scarcity of studies. For this period, the model suggests a divergence date around 6 million years ago. Following the discovery of the earliest hominins around 2000, plus studies of the chimpanzee genome, the trend towards older divergence estimates accelerated significantly, with a split suggested around 8.4 million years ago.

João d'Oliveira Coelho et al. (2026)

He and his colleagues began systematically compiling estimates of the human-chimp split, aided by a handy online database called TimeTree which compiles estimates of all such “species divergences”. Eventually they had 202 estimates, published from the late 1960s to the 2020s. Nobody else has systematically analysed the estimates in this way, says d’Oliveira Coelho (and I can’t recall seeing anything comparable).

At first glance, the uncertainty range hasn’t narrowed, or at least not much. If you just look at estimates prior to 1980, they range from 4 to 14 million years ago. Estimates published from 2015 onwards vary from 4 to 12 million years ago. Superficially, not much has changed.

However, don’t let the outliers fool you – on any question that attracts a lot of attention, there will always be some wild results. The important thing is the average.

D’Oliveira Coelho found a clear trend: the estimated timing of the human-chimp split has been steadily moving further back in time. Early studies had an average age of about 6 million years; more recent ones have pushed it back to around 8.5 million years. In other words, we may have been evolving separately from chimpanzees for 2.5 million years longer than we thought.

Distant cousins

Why has the date moved so far back?

Some of it is simply a matter of better data. We got the first draft of the chimpanzee genome in 2005. Everything before that was based on snippets of the genome, not the whole thing. Twenty years later, we got a much more complete version of the chimp genome, which filled in a lot of missing pieces and corrected many errors. Nowadays we also have a great many human genomes, meaning that we have a better sense of how much the DNA sequence varies.

But it is also about the underlying methods. You can estimate how long two species have been evolving separately by comparing their genomes. If the DNA sequences of both species are very similar, they probably split only recently, and if there are lots of differences, they split a long time ago. In essence, it’s simple.

In practice, it’s fiendish. Different parts of the genome evolve at different speeds. Genes involved in immunity change at a furious rate to keep up with new diseases, while genes involved in fundamental processes like making proteins barely change at all. Some species’ genomes change faster than others. Changes in population size matter. Changes in generation time – that is, the average age at which a species reproduces – also matter, and this can vary between males and females.

In response, geneticists have built complex models that include all these factors. The trouble is, figuring out just one of these things – say, the generation time – is a research project in itself. Inevitably, one has to make assumptions: 100,000 years ago, did Homo sapiens still have the same generation time as today? What about the ancestors of chimpanzees 3 million years ago?

“There’s all these things you need to model, and they are not in straight lines,” says d’Oliveira Coelho.

However, one of the biggest factors in pushing the date of the split back is the gradual acceptance of key fossil finds. The oldest putative hominins have all been discovered in the 21st century. Orrorin and the oldest Ardipithecus were first described in 2001, with Sahelanthropus following in 2002. This pushed the fossil evidence of humanity way back, from around 4 million years ago to 7 million years ago.

In theory, this should have immediately pushed back genetic estimates. Geneticists can use well-dated fossils as “anchors” in their models: instead of guessing how long a particular branch of the family tree was, they can input a hard data point. But for a long time they didn’t.

“Even after the three oldest hominins, Ardipithecus, Orrorin and Sahelanthropus, were published, it took years and years and years for geneticists to include that,” says d’Oliveira Coelho. Sahelanthropus is about 7 million years old, but it wasn’t until around 2015 that the average estimate of the human-chimp split became older than this. D’Oliveira Coelho blames this on a lack of communication between physical anthropologists and geneticists, and on some geneticists being overconfident about their data and methods.

His team ran its own meta-analysis, which placed the split between 8.69 million and 7.28 million years ago. However, d’Oliveira Coelho emphasises that that won’t be the last word. The only thing he is confident about is that claims of a really recent split can be safely ruled out, based on the fossil evidence. There’s still a window of millions of years when it could have happened.

To sum up, 50-odd years of research has slowly pushed the date of the last common ancestor further back in time. This means that chimpanzees and bonobos are more distantly related to us than we thought: many more generations separate humans today from the ancestor we shared with them.

Yet the more we learn about chimpanzees, the more they look like us. A study published in March showed that, when chimpanzees play together, they can quickly mimic the expressions their playmates make when they laugh. This may make for smoother social interactions and more enjoyable play. A June study, by a different group of researchers, found that chimpanzees and other great apes laugh in ways that are remarkably similar to us. This was especially true of laughter elicited by tickling. The researchers argued that “great apes have been laughing in a recognizable way to modern humans for at least 15 million years”.

I rather like that idea.