Computer simulation fills in the blanks of Neanderthal extinction

A modern human and Neanderthal skull face off.

By hairymuseummatt (original photo), DrMikeBaxter (derivative work) [CC BY-SA 2.0 (http://creativecommons.org/licenses/by-sa/2.0)], via Wikimedia Commons ^[1]

For a long time, archaeologists have suggested that modern humans wiped out Neanderthals because we had greater technological and cultural development, which allowed us to find and exploit resources more readily than Neanderthals did. It's a plausible explanation, but it leaves us with pressing questions about the details of how this might have happened.

For a start, we know that Neanderthals had some culture, so exactly how much more would modern humans have needed to have in order to be more competitive? And modern humans entered Neanderthal territory in smaller numbers than the established Neanderthal population—could technology make up for what they lacked in numbers?

These questions highlight a major challenge with this model: there are other plausible explanations for the disappearance of Neanderthals. For example, they could have been wiped out by climate change or an epidemic.

To answer these questions, we need to understand how much of a difference technology would make to groups in direct competition with each other. We'd also need to know how much technology makes the difference. It's not clear whether there's a critical mass of tools that makes a difference, or whether a small group is likely to carry a large innovation toolkit.

These questions aren't easily answered by the fossil record or by studies with modern humans. But they can be partly answered by computer models that simulate different scenarios and see which ones end up matching reality. A group of researchers at Stanford University and Tokyo's Meiji University have built a model that simulates what happens when two groups with different levels of culture compete for the same resources.

Pure numbers

Competition between species is "ultimately a matter of numbers," the authors write. Where one species thrives, the other is reduced to zero. Modelling this interaction is a matter of boiling it all down to numbers: creating two mock "populations" in a computer model, giving them certain characteristics, and seeing what circumstances lead to one population wiping the other out.

There are already models widely used in ecology research that simulate competition between different species, so the researchers needed to add culture into that somehow. There's a useful bit of existing information here: it's often been found that there's a link between population size and culture. That is, larger societies are more likely to have a higher level of cultural and technological development.

This makes sense, because cultural innovation relies on people having ideas. The more people in a group, the more ideas they're likely to have. We can estimate how often each individual might be expected to contribute an innovation that spurs on cultural development. (In reality, one individual might contribute more and another less, but it evens out at a per-person rate over time.) Once you have a number for the size of a population, you can work out how many "culture points"—individual instances of innovation—that group can be expected to have.

There's also the fact that not all cultural innovations get picked up; sometimes humans overlook a good idea or don't learn something properly. Other times, it's chance, like when they migrate in a small group that lacks an expert axe-maker. In either case, the innovation gets lost. Because of this, you also need to include a number for the group's learning rate: how well they catch on to new ideas, and how many innovations get lost over time.

So, if you have two populations with different levels of culture and different learning rates competing over computer-modelled "resources," what happens? The researchers found that even if one of the groups was really tiny compared to the other one, it could out-compete the larger group if it had enough of an advantage in cultural development and learning rate. "If [humans] had already developed higher levels of some kind of culture, even communication ability, or tools, or better ways to utilize game that they hunted, then that would be sufficient for them to overcome the superior numbers in the [Neanderthal] residents," says Marcus Feldman, one of the Stanford researchers.

Possible worlds

Computer models always have a lot of information missing. That's part of why they're useful—they allow us to strip a complex process down to a few core essentials and see if it works like we expect it to. But it does mean that there are complexities that they can't cover. In the case of cultural evolution, there are so many processes at play that they're difficult to incorporate into a single model. For instance, this model didn't answer questions like whether the invading humans shared any of their technological development with the Neanderthals and just how much faster at learning humans really were.

Feldman is also quick to point out that computer models can only give us certain kinds of information. "We're proposing an idea that's a possibility," Feldman says. "We can't say anything about probability."

That is, the model can't tell us whether competition with modern humans is a more likely cause of Neanderthal extinction than climate change or an epidemic. What it can tell us is that there's every reason to think that if early humans had more technological innovation than Neanderthals, we could have ousted them just as many archaeologists think we could. To go beyond "could," we need a stronger understanding of just how different humans and Neanderthals were in terms of culture and learning.

PNAS, 2015. DOI: 10.1073/pnas.1524861113 ^[2] (About DOIs ^[3]).