Denisovans: Why a mysterious group of ancient humans doesn’t have a species name

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.

One of the things I try to do in Our Human Story is answer the most commonly asked questions about human evolution. Back in February 2021, I tried to explain something that bugs a lot of people: how Neanderthals and modern humans could interbreed if they were separate species. (Short answer: the boundaries between species are fuzzy).

This month we’re going to tackle another perennial source of confusion. Why don’t the Denisovans, an extinct human group that was once widespread in Asia, have a species name? And what should their name be, if they ever get one?

The question of what the Denisovans’ “official” name should be has been rumbling on ever since they were discovered in 2010. It came up again in June, when a major discovery was announced. A skull from Harbin in North China, dubbed the Dragon Man, had been identified as a Denisovan using molecular evidence. We had never had a Denisovan skull before, so this was the first time we had a good idea of what their faces were like.

When I went on New Scientist’s podcast The World, the Universe and Everything to talk about the find, host Rowan Hooper asked me why the Denisovans don’t have a species name. Why can’t we call them Homo denisovanensis or something, the way we call Neanderthals Homo neanderthalensis?

Time was short, so I gave what I hoped was a simple answer: “It comes down to the fact that we have never had enough information about the Denisovans to be able to describe them properly… Their DNA is as different from Neanderthals as Neanderthal DNA is from us. Just on that basis, they’re different enough to count as a new species. But that’s not enough, according to the official rules of scientific taxonomy. You can’t just say, ‘That’s a new species’. You actually have to be able to describe in detail what the species looked like, what its skeleton was like. And we’ve just never had that.”

While that’s true, there’s also a lot more to it. There are two entangled questions. First, which fossils are actually Denisovans (and which aren’t)? That’s a question about objective reality, and very tricky to resolve, because it involves considering dozens of fossils and decades of research. Second, which of the many names that have been assigned should actually take precedence according to our rules of taxonomy? That’s a legalistic question about human processes – and thus even trickier.

Who’s in and who’s out?

First, here’s a reminder about the Denisovans. They’re a mysterious group of humans, first described in 2010 on the basis of a sliver of finger bone found in Denisova cave in the Altai mountains of Siberia. DNA from the bone revealed it was neither a modern human nor a Neanderthal, but something different. Furthermore, many people today carry some Denisovan DNA, especially in South-East Asia and Melanesia – indicating that Denisovans and modern humans interbred.

This implied that Denisovans must have been fairly widespread in east Asia within the last few hundred thousand years. So where are all the Denisovan fossils?

Fast forward 15 years to the present, and a small number of Denisovan fossils have been positively identified. For instance, a lower jawbone was found in a cave on the Tibetan plateau, and was identified using both proteins from the fossil and DNA from sediments. Likewise, a jawbone was dredged from the Penghu channel off the coast of Taiwan: in April, preserved proteins confirmed it was Denisovan.

Still, we are a long way from having a complete skeleton. The identification of the Harbin skull as a Denisovan took us a step closer by giving us a face. But there’s still a whole lot of skeleton still to find.

Now, there are a great many hominin fossils from East Asia that could, in theory, be Denisovan. Many of the finds have proved hard to classify: they don’t seem to quite match modern humans, or Neanderthals or any of the other established species like Homo erectus. This is enticing: if enough of them prove to be Denisovan, we’ll have a much more complete picture and maybe we could formally describe the species.

But how do we decide which fossils are Denisovan? Ideally, we’d have molecular evidence – preserved DNA or proteins – we could compare to the original Denisovan remains. But most of the specimens either haven’t been analysed or haven’t yielded anything.

One of the most prominent attempts to solve this problem was a preliminary study posted in 2024, with revisions in March, by a group led by Xijun Ni at the Chinese Academy of Sciences in Beijing. The team compared 57 hominin fossils, looking at as many physical traits as possible. This enabled them to draw up a family tree of all the various fossils.

Ni’s team found Eurasian hominins clustered into three main groups: modern humans, Neanderthals and a third group. This third group included the original Denisovan fossils, the jawbone from the Tibetan cave, the Penghu jawbone and the Harbin skull. It seems like the third group is the people we’ve been calling the Denisovans.

This is very neat if it’s true – but of course others disagree.

One contentious set of fossils comes from Hualongdong in South China. It’s a good collection: a nearly-complete skull with 14 teeth, an upper jaw, six isolated teeth and other bits. They’re all about 300,000 years old.

Ni’s team identified the Hualongdong fossils in the Denisovan group. However, a study in July led by Xiujie Wu, also at the Chinese Academy of Sciences, took a close look at the Hualongdong teeth. It found they didn’t match anything terribly well, and suggested they might represent yet another group. Of course, there’s another possible explanation: Denisovans were surely diverse, so maybe the Hualongdong Denisovans were a bit different from those elsewhere.

Meanwhile, there are many other mysterious Asian fossils, including the 260,000-year-old Dali skull and the also-260,000-year-old Jinniushan partial skeleton – both of which Ni’s team suggested were Denisovan.

At any rate, we have a growing list of Denisovan fossils, some more confidently identified than others. What are we going to call them?

Homo whatever

It so happens Ni was one of the researchers that described the Harbin skull in 2021. The team named it Homo longi. So maybe that’s what we ought to call the Denisovans?

But wait. A competing proposal was put forward last year, by Wu and Christopher Bae at the University of Hawai’i at Mānoa in Honolulu. In two papers, in Paleoanthropology and Nature Communications, they argued we should instead build a species around a set of fossils from Xujiayao in northern China. They proposed calling this new species Homo juluensis and including the original Denisovan fossils. So we should call the Denisovans Homo juluensis.

The selling point of this idea is the Xujiayao fossils do resemble the Denisovan fossils. In fact, Ni’s team also classed them as Denisovan. The difference is Bae and Wu wanted to treat the Xujiayao fossils as the “type specimen”, the one that the entire species gets named after.

This is simultaneously an argument about which fossils should be grouped together and about naming conventions. Let’s separate the two.

On the first front, the Homo juluensis proposal has a big problem. Bae and Wu explicitly said the Harbin skull isn’t a Homo juluensis/Denisovan, because it doesn’t look similar enough. However, the study from June clearly shows, using molecular evidence, the Harbin skull is a Denisovan. So as a description of the objective reality – which fossils are and are not Denisovan – Homo juluensis seems to have fallen flat.

What about taxonomy? The rules here are complicated. One key element is, essentially, first come first served: the first name to be applied is considered to have priority. On this basis, Homo longi has the advantage over Homo juluensis, because it was put forward three years earlier.

Are there any other possible names for the Denisovans?

The excavators of Denisova cave never formally described the Denisovans as a species. One member of that team, Anatoly Derevianko, referred to them as Homo sapiens altaiensis, which would make them a subspecies of modern human – but he didn’t do a formal description so it apparently doesn’t count.

This year, Derevianko has put out a series of papers proposing what Denisovans might have done in Mongolia, Uzbekistan, Tajikistan and Iran. He refers to them throughout as Homo sapiens denisovan. I haven’t been able to read the papers as only the abstracts are publicly available, so I don’t know if he has provided a formal description – but if he did, he did so four years after Homo longi was named.

If you really dig around, you can find a few more options. A 2015 paper uses Homo denisovensis and a 2018 study plumps for Homo denisensis. Neither has been widely accepted.

Finally, there’s the possibility of a really old name. Maybe someone named one of the Asian hominin fossils decades ago in an obscure paper: if that fossil turns out to be Denisovan, that name would have priority (assuming the description was done properly). However, Wu, Bae, Ni and others looked into this in a 2023 paper. They found key fossils were never properly named. There had been loose suggestions that, for example, the Dali skull could be called Homo daliensis, but these were throwaway remarks rather than formal descriptions.

At this point your head is probably spinning from all these fossil names and species names, so let’s sum up. The main point is we are fleshing out our understanding of the Denisovans – and that means we’re getting closer to being able to give them a taxonomic name.

For what it’s worth, based on my understanding of taxonomic rules, I think Homo longi has a good chance of becoming the official name. I’m not sure it would have been my choice, but it isn’t up to me. In any case, they’ll always be the Denisovans to me.