Clink, clink, clink. The sound of stone-age hunter-gatherer toolmakers at work on the veld has faded away through time, but thanks to a group of archaeologists their every strike of stone on stone at their open-air “workshops” is being pieced together.

“We can, from A to Z, understand the complete process here,” says Dr Aron Mazel, the former assistant director of the KwaZulu-Natal Museum whose “rescue dig” of a site back in 1991 gathered the bits of stone that have led to a unique insight into the life of hunter-gatherers from over 100 000 years ago. But the credit for putting them back together, says Mazel, goes to his colleague from Germany, Gunther Möller, who analysed Mazel’s original finds for his Master’s thesis and has just been back to the province to extend the research. “He's a jigsaw magician,” says Mazel, for being able to recreate how a rough block of hornfels rock was shaped and prepared for use in making tools that might have ranged from scrapers to arrowheads. Given that Mazel’s haul consisted of 7529 pieces of stone, most of them less than one centimetre in size, and that Möller managed to arrange them by sight, without the use of AI or 3D scanning, the mantle of “magician” is no exaggeration.

Mazel and Möller were interviewed at the KZN Museum following their joint fieldtrip, before they flew back home to England and Germany.

The stone pieces were gathered about 13km from Nquthu in northern KZN, after a geologist, Greg Botha, alerted Mazel that he had discovered artefacts eroding out of the side of what is known as the Jojosi Donga, in the “erosional badlands” of the area. The artefacts were collected and measured at the time, but they were boxed and stored at the KZN Museum without being investigated until Möller, who is from the Eberhard-Karls-Universität Tübingen, decided to study them in 2023

It is not unusual to find many stone artefacts of hunter-gatherers out in the open, “but here was a situation which we had never heard about before. They were in situ, in their original context,” says Mazel, who accompanied Möller’s and his colleagues to further excavate the Jojosi Donga in September and October 2024. Altogether they have now researched seven of these stone-age workshops within about 250 metres of each other.

Most of the insights into the lives of Middle Stone Age people, in KwaZulu-Natal, have been obtained from shelters, and caves like Sibhudu Cave, which is inland from Ballito. Sibhudu was occupied from about 77 000 to 38 000 years ago, and provides evidence of early humans using plant bedding, as well as ochre, which suggests symbolic or ritual activities. However, says Mazel, “there are actually not many Middle Stone Age rock shelter sites known in KwaZulu-Natal”.

Fossils of Homo sapiens in South Africa date back to between 200 000 and 300 000 years ago, and the Middle Stone Age, during which tool-making was refined, spanned from about 300 000 to 30 000 years ago.

Now, the picture of their activities beyond their rock shelter homes has been expanded by the discovery of these stone workshops, or knapping floors. Knapping is what archaeologists call the process of removing flakes from the original pieces of stone (or cores) and then shaping the chosen pieces to make tools like scrapers and arrowheads. “What they would do,” says Mazel, “is take a big piece, and then they might either prepare a top, or hit it from the side with a hammerstone, and chip off flakes. Then they’d look at them and decide which to use or which not to use, for scraping or for cutting maybe. In this particular instance, the most common would probably be scrapers, to scrape fat off skin before tanning.”

“Imagine a hunter-gatherer society,” says Mazel. “There were small groups between 10 to 20 or 30 people who moved around in the landscape looking for raw materials. And at this specific place (Jojosi Donga) we can identify one of the tasks they were doing to live, namely gathering stone, working it, and extracting the pieces they needed.”

“This area is very rich in a type of stone called hornfels, which was sought after by hunter-gatherers to make stone tools from. And it was used over a long period of time.”

“It tells us the nature of the material left behind on a knapping floor. It's simply incredible. Every single bit of chipped stone that was part of the action that a hunter-gatherer in the Middle Stone Age performed there was left at the place where it happened.

“This gives us the opportunity to reconstruct the knapping actions. There are thousands of flakes and chips, basically the waste that is produced while working stone. And by putting back together the artefacts, how they were knapped - we call them refits - we can basically understand the action sequence completely, which gives us technological insights of which knapping strategies they actually used.”

“It’s a 3D jigsaw puzzle,” says Möller. “There are certain physical attributes, angles and edges, that result from the knapping process, which allow me to determine if you use a stone hammer or a piece of bone, for example. But apart from that, if you look long enough at these artefacts and make up your mind how the raw block looked and how someone might have approached it, you can put the pieces back into the sequence and fit their surfaces together.” In total, he did 48 refits of the original stones from the 1991 haul.

The recent visits by Möller and colleagues has yielded a further five knapping sites, and many thousands more flakes and chips, and many more refits pending. However, relocating the original excavation location was itself a feat. “I could relocate the old excavation site,” says Möller, “with the amazing photographs Aron did, because there were a lot.” GPS technology was not available in 1991, and so the only way Möller could locate the site was from the angles in the photographs in relation to the hills in the background. On his recent fieldtrip, “we basically employed the same excavation strategies as Aron 30 years ago, but we have also plotted (using laser technology) the coordinates of every single piece greater than 2cm in size in x, y, and z axes.”

Mazel emphasises that the mere existence of these sites is extraordinary. “There would not have been much wind or rain that would have washed those artefacts away. Because most of them are very light and very flimsy, some of them, it would suggest that they were covered over quite quickly. A bit of wind would have blown in some dust. It'll begin to settle. And then there would be another layer, and they’ll keep building on each other. And then you've got this preserved layer.” It is this sedimentation that allowed Dr Svenja Riedesel of the University of Cologne in Germany to use Optically Stimulated Luminesce to date the Jojosi knapping floors and show that hunter-gatherers worked their stone artefacts there between 106 000 and 187 000 years ago.

Thanks to Mazel and Möller and their colleagues, it is possible to see that hunter-gatherers during this period would, during the course of their movements across the landscape for  thousands of years, repeatedly return to this hornfels quarry near Nquthu, choose a suitable rock to work, and settle down on the grassy plain to chip out some tools. “That doesn't mean that they’re constant every year,” says Mazel. “There would have been ebbs and flows, but we don’t have the data to know. But hypothetically the raw material was known about for over 300 000 years.”

With some questions answered, more arise.

Before this, “we didn't know that there were any knapping floors. Now we can really hone in and be more tightly focused,” says Mazel.

“The central question,” says Möller, “is how did the hunter-gatherers who inhabited that landscape, interact with it, and what tasks did they carry out there? Now that we have exposed artefact concentrations, which are limited and fade out towards the outer areas, what did they actually do except gather and work raw material there? We have some indications that some of these blanks (unmodified pieces) have been used as tools. But where are they? Hunter-gatherers were behaviourally extremely flexible. What's the trajectory of these pieces? Of course most hunter-gatherer interactions happened in the open landscape, but they most likely also occupied shelters in this period. So, where did they export the material to?”

“We know they specifically went there to access this raw material, which was abundantly available here. Okay. But we wouldn't know how that fitted into their movements or how they would have come across it in the first place. We wouldn't know any of that stuff.”

“We're also in the process of determining the use wear of artefacts. We have specialists to analyse what happens if you use a piece to scrape leather 100 times. What does it look like afterwards? You can't see this with the naked eye, but you can see microscopically.”

For example, chips in Mazel, “you can see if that tool was used for scraping skin. Or if that tool, which has a point like this, was used for cutting wood because you'll see through the microscope the polish, which would be reminiscent of wood. But we don't have much of these in there probably.”

Mazel adds that “we have to acknowledge that these hunter-gatherers were hugely knowledgeable of the environment. They would have known where the raw material was, and that would have passed down from generation to generation. They would have known the habits of different animals. But how did they track them? Did they spear them? How long did they knap for: one day, three days?”

For archaeologists these types of questions are difficult to answer, but another remarkable find such as that at Jojosi Donga may allow them to edge closer to answers.

For now, these questions hang in the wind, mingling with the Sunday chanting from nearby Ndindindi village, and the occasional lowing and bleating of the cows and goats that traverse the dongas.