The first technical gestures: stone, bone, wood — The World, with the Evidence

Several views of the same orange, drop-shaped stone, knapped into small facets, next to a ruler. — A flint handaxe, knapped on both faces in the Acheulean period, a very long time ago: between about 500,000 and 190,000 years before the Common Era. The Portable Antiquities Scheme, Matthew Fittock, CC BY 2.0, source

On the face of it, all you would need to do is strike two stones together.

That way you get random flakes, and probably a crushed finger. Getting a clean cutting edge, in the place you want it, requires knowing several things. It is the oldest technical knowledge to have left any trace.

A stone that breaks the right way

Not every rock will do. You need one with a very fine grain, which splits in a regular way instead of crumbling. Flint is one, but there are many others. When they break, they leave a curved, smooth surface, rather like the inside of a shell — prehistorians call this a conchoidal fracture.¹

Next, you have to strike in the right place and at the right angle. The angle between the surface that is struck (the striking platform) and the face from which the flake is to come away matters a great deal: if it is badly chosen, nothing good comes off. Too flat, and nothing happens. Too close to the edge, and the stone shatters haphazardly.

The choice of hammer matters too. A hard stone does not give the same result as a soft stone or a piece of antler. You can strike directly, place an intermediate piece between the hammer and the stone, or press instead of striking (pressure).²

In the narrow zone where everything is right, the blow detaches a flake with an edge finer than a razor blade.

Knapping, then, is not bashing. It is choosing a material, taking aim, and knowing in advance what is going to happen.

Older than us

For a long time, it was thought that the first tools had been made by the first representatives of our genus, Homo. It was a fine story: the tool and the human appeared together.³

It even contained a second one, which is often forgotten: this technological development was supposedly brought about by climate change and the spread of savannah grasslands. Retreating forests, open spaces, ancestors forced to adapt — and the tool is born of constraint.⁴

In 2015, a team co-led by the archaeologist Sonia Harmand published the discovery of the site of Lomekwi 3, west of Lake Turkana, in Kenya. Stone tools were knapped there 3.3 million years ago.⁵

That is 700,000 years before the oldest tools known until then (Oldowan). Above all, it is before the appearance of the genus Homo.⁶

In other words, someone was knapping stone before us. Probably Australopithecus or Kenyanthropus.

The second part of the fine story falls too: the site lay in a wooded environment, not in an open savannah.⁷

This can be compared with our page on grasslands, which shows that forests and grasslands coexisted in a mosaic much earlier than was believed. Two different disciplines, the same story coming apart: that of our ancestors driven out of the trees by drought.

These tools — one hundred and forty-nine pieces — are also different. They are larger, more roughly knapped, and the traces show that the stone was held in both hands, or set on another stone serving as an anvil, before being struck.⁸

The authors are cautious in their own abstract: they write that these knappers had a developing understanding of stone’s fracture properties. Not mastery. Knowledge in the making.⁹

The fine story was false. The true one is more interesting: the tool does not wait for the human, it comes before it.

All that we no longer see

We call this period the “Stone Age”. The name is misleading.

These hominins did not live in a world of stone. They used wood, bark, plant fibres, hides, tendons, bone, shells. Doubtless far more than stone.

But organic matter rots. Only the stone remains.

So we name this period after the content of what has survived, rather as if archaeologists of the future, finding only our glass bins, decided that we lived in the Glass Age.

A very few wooden objects have escaped this fate. At Schöningen, in Germany, hunting spears of spruce and pine have been found, perfectly preserved in waterlogged sediment, next to the bones of wild horses.

Their age, moreover, has just been revised. They had been dated to about 300,000 years; a 2025 study, based on amino acid dating of the opercula of freshwater snails found in the same layers, puts them at around 200,000 years. They therefore date from the time of the early Neanderthals.¹⁰

They are enough to show what we have lost everywhere else.

How the gesture is recovered

Since objects do not speak, researchers have two methods.

The first is to do it again. You get hold of the same rock, sit down, and knap it again. By trying, you understand which gestures produce which traces. This is experimental archaeology.

The second is to look very closely. Under the microscope, an edge that has cut meat, wood or hide does not wear in the same way. This is use-wear analysis.¹¹

To read these traces, researchers make replicas, use them on different materials, and record the marks obtained in order to build up a reference collection.

Then, when analysing an ancient tool, they can determine the movement made — scraping, cutting, piercing — the material worked, and sometimes whether the tool was hafted.

What remains, what is missing

Material	Preservation	Level of knowledge
Knapped stone	Excellent (indefinite)	Very high
Bone and antler	Moderate to good depending on the soil	High
Wood	Very rare (anoxic/waterlogged settings)	Occasional but precise
Fibres, cordage, basketry	Exceptional cases only	Very low
Hides and leather	Exceptional cases only	Very low

Note 1

The term conchoidal is used to describe fracture with smooth, curved surfaces that resemble the interior of a seashell; it is commonly observed in quartz and glass.

Encyclopædia Britannica, “Conchoidal fracture”, mise à jour le 24 June 2026

See the full note ↓

Note 2

On peut détacher un éclat en frappant ou en pressant. […] Le nucléus est débité soit au percuteur dur, soit au percuteur tendre par percussion directe ou indirecte. Dans ce dernier cas, on utilise une pièce intermédiaire posée sur le bord du nucléus

A flake can be detached by striking or by pressing. […] The core is reduced either with a hard hammer or with a soft hammer, by direct or indirect percussion. In the latter case, an intermediate piece placed on the edge of the core is used

Tixier, Jacques, “Préhistoire – La taille expérimentale des roches dures”, Encyclopædia Universalis, section “Les grandes lignes de la taille des roches dures”

See the full note ↓

Note 3

Human evolutionary scholars have long supposed that the earliest stone tools were made by the genus Homo