Finding ancient fire use in the Rising Star cave system

This week my colleague Lee Berger gave a Carnegie Science lecture on the future of exploration. He recounted some of the outcomes of our team's fieldwork in the Rising Star cave system this year, including his own perilous journey into the Dinaledi subsystem of the cave. There are some exciting finds from this fieldwork that the team is still working on. One thing that Lee was able to share is that we now have abundant evidence of ancient fires within deep chambers of the cave system where we also have evidence of Homo naledi bones.

The team identified charcoal, burned bone, or other evidence of fire in at least four chambers, widespread from each other in deep parts of the system. Some of this evidence is from excavations led by Dr. Keneiloe Molopyane in the Dragon's Back Chamber, including concentrations of well-preserved charcoal, ash, and discolored clay that appear to be small hearths. The excavations also turned up many fragments of animal bone. The Dragon's Back Chamber is immediately adjacent to the Dinaledi subsystem of the cave, and these excavations are into sediment at a similar depth as the situation of H. naledi fossil material in Dinaledi.

In addition to Dragon's Back, in two other deep chambers our exploration has encountered concentrations of charcoal. Both of these situations are areas where our team has not previously focused any investigations, and one of the chambers has burned animal bone on the cave floor with charcoal.

Besides the charcoal evidence, Lee identified surfaces with discoloration consistent with soot. Some of this evidence occurs within the Dinaledi Chamber above skeletal remains of Homo naledi.

People have put the question to us ever since 2013 when the team first discovered fossil remains from the system: Did Homo naledi use fire?

We have found naledi remains in several deep areas of the system, 30 meters beneath the surface and 100 meters or more from any current entrance. Three of these contexts we have reported: the Dinaledi Chamber, the Lesedi Chamber, and the U.W. 110 locality. Others we are working on now. The cave has small, narrow passageways, which can be difficult to navigate even with headlamps. Geological evidence shows that these areas were far from entrances and with similar narrow constraints at the time the naledi remains were deposited. It seems unlikely that any hominin species was using these areas without light.

Personally, it seemed weird to me that people really thought fire would be surprising from naledi. Sites with evidence of fire go back in Africa more than 1.5 million years. Swartkrans, South Africa, has a long record of burned non-hominin bone, many with cutmarks, within its Member 3 deposits dating to around a million years ago, and Swartkrans is only 800 meters from Rising Star. The Homo naledi fossils are at most 335,000 years old. Hominins were making fires in this exact part of South Africa more than a half million years before the Rising Star fossils.

So I've never been hesitant to tell people that I thought naledi had used fire. Being able to light some dark areas would account for their presence in chambers separated by more than 150 meters in dark areas of the cave.

At the same time, I've consistently told people not to expect physical evidence of fire in the cave. Fire has been hard for archaeologists to document in the deep past. When a small group of human foragers makes a fire, all that may remain are bits of charcoal and ash that are easily swept aside and unlikely to leave much of a mark. Only if a fire burns for a long time, or if people make fires in the same place again and again, and if these are quickly buried, may archaeologists eventually unearth a lens-shaped area of ash and burned debris.

I personally wasn't thinking that such long-sustained fires ever existed in the deep parts of the Rising Star system. We had found no other evidence of naledi living there, just their bodies. My assumption was that small fires used only for lighting would have left little trace. But it looks like I was wrong to assume. What we seem to be finding now is evidence of such small fires and possible associated bone material. We have not yet found any stone artifacts or debitage near these features.

What is ahead of the team now is further work to understand the context of fire use in the cave system. This is going to take some time. We have more excavation in Dragon's Back to do, and new documentation of other chambers that were not a focus of our research before this year.

So why talk about this now before completing these next stages of research?

I'm happy that our team has taken an open approach to public engagement from the first expedition into the Rising Star system in 2013. We shared the discovery of the rich hominin bone assemblage in the Dinaledi Chamber and streamed the first excavation work. Some of the early results of the work were easy for anyone to understand. We didn't have interpretations of the fossils, or dates, or traditional measures of peer-reviewed outcomes. We just knew that we hadn't seen anything like it and we thought it would be great for people to follow the science in progress.  

That's how I still feel. People ask me at every public lecture, “How did naledi see in the dark parts of the cave?” I think it's reasonable to say that we now have a lot of evidence of fire in the parts of the cave system where we have naledi bones. This is a spatial association and it seems ridiculous to hide it from people while we're working to understand the strength of any possible temporal association.

Withholding information like this would also lead other scientists astray. Peer reviewed journals have already published many commentaries by archaeologists challenging the idea that naledi used deep cave chambers. Some have suggested lost entrances, carnivores that left no marks on the bones, catastrophic cave collapse, or unnoticed erosion and redeposition of bones from a near-surface context. Just last week, the South African Journal of Science published two commentaries by archaeologists, one of them asking pointedly whether there is evidence of fire use in the cave.

All this is speculation, and it's peer reviewed. I'd rather that other folks used blogs and Twitter to air their opinions, because then it would be more like a conversation.

I can point to another example of a widespread misconception. Finding the age of the fire evidence is very important to understanding its context. Some archaeologists are saying that dating this material should be easy. They ask, why not just do a few radiocarbon dates and rule out that the charcoal may be recent? Their assumption is that a radiocarbon result would eliminate H. naledi and implicate humans.

Data that we reported in our 2017 paper helps show why finding a few recent dates doesn't make things easy at all. Two naledi bones yielded radiocarbon dates between 36,000 and 33,000 BP. We very much considered whether these naledi bones may actually be this young, and today I still would not categorically rule this out. However other methods including direct U-series-ESR ages on naledi teeth made the radiocarbon results stand out as inconsistent. To explain the inconsistency, the team hypothesized that the radiocarbon results reflect contamination by carbon from outside the bones.

The point is, finding a young radiocarbon date may not clarify the situation very much. Aside from the Dinaledi assemblage, the other naledi discoveries have no geological age estimate yet, including the Lesedi Chamber material. We cannot assume they are all the same age, and some of them may surprise us. I keep my mind open to the possibility that additional radiocarbon work may arrive at a young result.

The system is a dynamic underground landscape, and it will be critical to design a comparative geochronological framework that lets us test possible connections across chambers. When our team worked on the chronology of the Dinaledi Chamber, we benefited greatly from designing a project with multiple methods and double-blind sampling. Even after this extraordinary effort, there still remain many open questions. When I look around the world at other sites, there are many cave systems where it has taken many decades to work out the chronology. That science is continually advancing, and sometimes dates that seemed solid a decade ago are rewritten.

It's going to be an exciting ride for a while with Rising Star.

Note: Lee's lecture is available online for people to watch. It's a great summary of the work underway.