Which humans first made tools or art – and how do we know?

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.

When writing headlines for stories about human evolution, the favourite superlatives are “oldest”, “earliest” and “first”. I have lost count of the number of articles I’ve written that used these.

And it’s not just about attracting more readers – even though it usually works that way. If a researcher can find evidence that a species or behaviour is older than previously thought, that is useful information. Figuring out the order in which things happened is crucial to understanding why they happened.

For instance, we used to think that all rock art was created within the last 40,000 years. That meant it had to have been created by our species (Homo sapiens), as other groups like the Neanderthals had died out by then. But it turns out that some prehistoric art is older than that, so it’s possible Neanderthals were artists too.

In the past month we’ve had a lot of “earliest” and “first” findings, and they got me thinking about how to interpret these things. When can we be confident that we have figured out how long something really existed for?

First!

Let’s start with a story I wrote, if only to get it out of the way. At a dig in southern Greece, archaeologists uncovered two wooden objects that appear to be tools: one seems to be a digging stick, the other’s specific use is difficult to identify. Both are about 430,000 years old, making them the oldest known wooden tools.

They aren’t that much older than the previous record-holders. The Clacton Spear found in the UK is thought to be 400,000 years old, though it was excavated many decades ago so the dating is necessarily uncertain. A set of wooden spears from Schöningen, Germany, were thought to be of a similar age, but their age has been revised downwards: some methods put them closer to 300,000 years old and a study from May 2025 indicated they were just 200,000 years old.

Bone tools also crop up in Europe around this period. At Boxgrove in the UK, researchers found a chunk of bone from an elephant-like animal, perhaps a steppe mammoth. It had been formed into a hammer for reshaping stone tools. The elephant bone is 480,000 years old, making it the earliest known use of elephant bone in Europe. However, elsewhere bone tools were in use much earlier. In east Africa, ancient humans were systematically making tools from bone, including elephant bone, 1.5 million years ago. The practice may date back even earlier, of course.

Let’s come forward in time a bit. At Xigou in central China, archaeologists have just documented a trove of 2601 stone artefacts, dating to between 160,000 and 72,000 years ago. The artefacts included hafted tools: that is, stone tools that were fitted onto something else, such as a wooden pole. This, the researchers say, is “the earliest evidence for composite tools in Eastern Asia, to our knowledge”.

And finally, in early January we learned that people in South Africa were hunting with poison arrows 60,000 years ago. Archaeologists found five arrowheads made of quartzite, which were coated with a sticky, poisonous liquid, probably from a plant.

Each of these finds has more to it than meets the eye.

Ever further back

The oldest known wooden tools are almost certainly not the actual oldest wooden tools. The issue here is preservation: wood rots, so our record of prehistoric wooden artefacts is pretty scanty.

When I spoke to Katerina Harvati, who led the excavation of the wooden tools, she was very clear that people were using wooden tools long before 400,000 years ago. It’s just that we haven’t been able to find them.

In fact, given that wood is easier to work than stone, and that chimpanzees sometimes make simple wooden tools, it may be that wooden tools are the oldest form of technology. If next week a paper arrives claiming to have found wooden tools from a million years ago, it will be a significant discovery, but – other than the fact of their preservation – it won’t be remotely surprising.

It follows that we should not hang any major narratives about human technological development on the age of the oldest wooden tools. We would need to systematically investigate sites where such tools could be preserved, at a range of ages, before we could be remotely confident about when people started using them.

Let’s now reconsider those poison arrows. What they are is the oldest known examples of arrowheads with poison on them. However, the story points out that arrowheads with a design consistent with modern-day poison arrows can be found tens of thousands of years earlier. Furthermore, like wood, poisons are prone to biodegrade.

This is something where I think we can be a bit more confident. Poison arrows are another form of composite technology – putting two or more items together – and that seems to only arise in the later stages of human evolution. It’s not something we see any evidence of early hominins like Ardipithecus or Australopithecus making – whereas I wouldn’t bet against them making simple tools out of wood or bone. We should not wed ourselves to 60,000 years ago as the origin of poison arrows, but our error margin is probably narrower.

And then there’s the question of the oldest art, which is an absolute horror show.

Prehistoric graffiti

The most famous ancient artworks are cave paintings, though there are also sculptures, engravings and much more. The trouble with a lot of this is that it’s really hard to date.

If you find a sculpture buried in sediments, it’s often possible to date the sediments. But cave paintings are much trickier. If they were made using charcoal, you may be able to use carbon dating, but only if they were made in the last 50,000 years: any older than that and carbon dating is useless. Most cave art has never been dated, and often can’t be dated given current technologies.

In the past couple of weeks, we learned that a hand stencil painted on a cave wall in Sulawesi in Indonesia was at least 67,800 years old. This makes it the oldest known rock art anywhere in the world, edging out a similar hand stencil at a cave in northern Spain, which has been attributed to Neanderthals.

Did you see the crucial caveat? It was the words “at least”. The way these artworks have been dated is by sampling thin layers of rock that have formed on top of them, due to water trickling over the rock surface and depositing minerals. These “flowstones” can be dated, but that just gives you a minimum age. The underlying artwork could be much older.

My point in going through all of this is not to say that we don’t know anything: on the contrary, we have a lot of information, much of which wasn’t available just 10 or 20 years ago. Instead, I want to think through how we might achieve a reliable timeline for human evolution and cultural development, and which bits are seemingly doomed to uncertainty.

In the fossil record, volume is helpful. Most palaeontologists don’t study large and charismatic animals like dinosaurs, but small things like marine molluscs. The reason is that these organisms are fossilised in huge numbers, which means it’s possible to trace evolutionary changes in a lot of detail. If a species is common in the fossil record and then suddenly disappears after 66 million years ago, that’s good evidence that the species really did go extinct at that point.

In the human fossil and archaeological record, what things do we have lots of, and what things are scarce?

There are many hominin species for which we only have a handful of specimens, especially the early ones. That means we have basically no information about how long they existed, or how widespread they were. We also can’t determine if one species evolved directly into another, or if some more complicated process occurred.

In contrast, our record of stone tools is pretty extensive. It has been inching further back in time: the current oldest known stone tools are the Lomekwian tools from Kenya, dating to 3.3 million years ago. I would not be surprised if older ones turned up. However, really early hominins like Orrorin (6 to 4.5 million years ago, maybe) and Ardipithecus (5.8 to 4.4 million years ago, ish) seem to have spent a lot of time in trees, so I actually would be surprised if they were making stone tools.

Wooden tools are another matter. Our sample size is small and spotty, and that’s down to limits to preservation. I don’t expect to have a reliable timeline of the development of wooden tools in my lifetime.

As for art, our main limit is technological. We have no shortage of preserved art: the problem is inventing ways to reliably date it. At the moment, we can’t draw a timeline of the development of art, and I am suspicious of any attempt to fit it into a wider narrative. But that could change as more art is dated and new techniques are devised. By the time I retire, I expect to have a much better grasp of how ancient humans’ artistic practices changed.

In some sense, all our stories about human evolution are provisional. That’s true of all palaeontology, of course, but some stories are more provisional than others. There isn’t much wiggle room about the end-Cretaceous extinction that took out the non-bird dinosaurs. But there is still a lot of wiggle room in the human story. Some of that can be closed by digging up more artefacts and creating better dating techniques, and some of it we might just have to live with.

Neanderthals, human origins and cave art: France

Embark on a captivating journey through time as you explore key Neanderthal and Upper Palaeolithic sites of southern France, from Bordeaux to Montpellier.