Two Million Years of Stone Tools

In the dry winter of 1959, Mary Leakey was brushing the side of a small gully in Bed I at Olduvai Gorge, in what is now northern Tanzania, when a flat-topped skull came clear of the matrix. She had been working the gorge with her husband, Louis Leakey, since 1931, returning season after season to a place that had so far yielded promise more than payoff. The find she made that July, lying among simple flaked pebbles, would change everything. The cranium belonged to a robust early hominin, and the crude worked stones scattered around it were among the first securely dated tools of their kind in Africa.

That July afternoon is often treated as the founding moment of African Stone Age archaeology, and not because the skull alone was decisive. What mattered was the association: a hominin and its handiwork in the same ancient layer. To understand why a pile of broken cobbles deserves this much attention, we need the whole arc of the record. It runs for more than two million years, and it is the single longest window we possess onto the workings of the human mind.

Slicing Deep Time Into Three Uneven Pieces

Archaeologists divide the long span of the Stone Age into three broad phases, each defined less by a calendar than by the dominant ways people made and used stone. The earliest, the Lower Palaeolithic, runs from roughly 2.6 million years ago to about 300,000 years ago, an almost unimaginable stretch covering the bulk of the human technological past. The Middle Palaeolithic follows, from about 300,000 to 50,000 years ago, a period during which Neanderthals and early modern humans overlapped across Africa and Eurasia. The Upper Palaeolithic is the most recent and the shortest, from roughly 50,000 to 10,000 years ago, ending as the last ice age released its grip.

These boundaries are coarse, and they should not be mistaken for sharp lines that nature drew. Real change happened unevenly across continents, and the same tool tradition could persist in one region long after it had been replaced elsewhere. Even so, the divisions are useful because they map onto something real, namely the rise and fall of distinct stone tool industries, each with its own characteristic methods and signature artifacts.

Five Steps Across Two and a Half Million Years

If you wanted to compress the whole story into a single image, you could line up five objects in chronological order and let their shapes tell it. First comes the Oldowan flake, a sharp sliver knocked off a cobble. Then the Acheulean hand axe, a symmetric teardrop worked on both faces. Then the Mousterian prepared core, from which a flake of predetermined shape was struck. Then the long, parallel-sided blade of the Upper Palaeolithic. And finally the microlith, a small, geometrically trimmed piece meant to be set into a handle alongside others.

This five-mode sequence covers more than two and a half million years, and laid out side by side the objects reveal an unmistakable trajectory. The progression moves from opportunistic to planned, from one-off to standardized, from a single working edge held in the hand to modular pieces assembled into composite tools. None of this means each step instantly and everywhere replaced the one before it, and the shorthand flattens a great deal of regional variation, but as a way of seeing the overall shape of the record, the five modes are hard to beat.

Pebbles, Flakes, and the First Stable Tradition

The oldest stable stone tool tradition is the Oldowan, named for Olduvai Gorge itself and attested from about 2.6 million years ago at sites such as Gona in Ethiopia and the lower beds at Olduvai. The technique is direct hard-hammer percussion, which is exactly what it sounds like: a hand-held cobble, the hammerstone, is struck against a target stone, the core, to drive off sharp-edged flakes. The crucial insight, easy to miss because the cores look more impressive, is that the flakes were usually the working tools. A freshly struck flake carries an edge sharp enough to slice through hide and disarticulate a carcass, and the discarded cores were often just the leftovers.

Simple as the Oldowan looks, it represents a genuine cognitive achievement. To strike a usable flake reliably, a knapper has to understand how stone fractures, choosing a core of the right material and delivering a blow at the right angle and force; get it wrong and the cobble simply crumbles or refuses to flake at all. The makers were early members of our own genus and possibly some of their robust contemporaries, and for well over a million years this modest repertoire of flakes and choppers was the cutting edge of technology on the planet.

An Even Older Beginning Than Anyone Expected

For a long time the Oldowan stood as the start of the whole story, and the start of stone tools was widely assumed to coincide more or less with the start of the genus Homo. Then, in 2015, Sonia Harmand and her team reported something that pushed the origin back dramatically. At Lomekwi 3, on the western shore of Lake Turkana in Kenya, they recovered flaked stone tools dated to 3.3 million years ago. These artifacts, grouped under the name Lomekwian, predate the oldest Oldowan tools by some 700,000 years.

The implication is striking. At 3.3 million years ago there were no members of the genus Homo yet, so the makers were almost certainly australopiths, the small-brained bipedal hominins best known from fossils like the celebrated Lucy. The Lomekwi tools are large and heavy, made by techniques that look distinct from later Oldowan knapping, which suggests not a single inventive moment but perhaps several independent experiments with stone over deep time. The find unsettled a tidy assumption, namely that toolmaking was the near-exclusive property of our own genus, and it reminds us that the record we read is only as old as the oldest site we have so far been lucky enough to find.

The Hand Axe and a Million and a Half Years of Symmetry

After the Oldowan came a tradition that would prove astonishingly durable. The Acheulean, named for the site of Saint-Acheul in the Somme valley of France where nineteenth-century scholars first defined it, appears in Africa from about 1.76 million years ago at sites such as Kokiselei and Lokalalei in West Turkana, and it persisted until roughly 200,000 years ago. Its signature artifact is the hand axe, a symmetric teardrop-shaped biface worked on both faces to a deliberate, often beautiful form, and by a wide margin it is the longest-lived tool form we know.

That endurance is part of what makes the hand axe so puzzling. Producing a balanced, bifacially flaked teardrop requires the knapper to hold a mental template of the finished object and to work toward it through many careful blows, anticipating how each removal will shape the next. The symmetry is not strictly necessary for cutting, which raises the genuinely open question of why so much effort went into form. Whether the answer lies in handling, in display, in social signaling, or simply in shared norms of how a proper tool should look, the hand axe demonstrates cognitive and motor demands that go well beyond knocking a quick flake off a cobble.

Planning the Flake Before You Strike It

The next major shift is a subtle one, and it lives mostly in the mind of the maker rather than in the final object. The Mousterian industry, defined at the rock shelter of Le Moustier in the Dordogne valley of France and strongly associated with the Neanderthals, ran from roughly 300,000 to 30,000 years ago. Its key innovation is the Levallois prepared-core technique. Instead of striking flakes opportunistically and taking whatever comes off, the knapper first shapes the core itself, trimming its surface and edges in advance, so that when the decisive blow finally lands, a single flake of predetermined size, shape, and edge geometry comes away.

The Levallois method is a leap in foresight. It separates the work into two stages, a long preparation and a single payoff, demanding that the knapper plan the end product several steps ahead. This is planning in the literal sense, holding a future result in mind and working backward to achieve it, and it shows that the people of the Middle Palaeolithic were not improvising their tools so much as engineering them.

Blades, Needles, and Tools Made of Many Parts

From about 50,000 years ago, the pace and character of innovation change again. Across Eurasia, anatomically modern humans began producing long, slender, parallel-sided blades, struck off carefully prepared cores in series so that one core could yield many near-identical blanks. They also turned to new materials, working bone and antler into needles, points, and harpoons, which implies sewn clothing and more elaborate hunting and fishing gear. And crucially, they assembled composite tools, combining stone, wood, sinew, and pitch into single hafted implements such as a spear point bound to a shaft or a blade set into a handle.

In Europe this Upper Palaeolithic unfolds as a named sequence of cultures, the Aurignacian, Gravettian, Solutrean, and Magdalenian, each with its own toolkit and, in several cases, its own remarkable art. Analogous blade-based industries appear across Africa and Asia, so this is not a uniquely European development but a broad human one. The shift toward standardized blades and modular composite tools marks a recognizably modern way of making things, breaking a task into interchangeable parts and recombining them.

What a Used Edge Remembers

Knowing how a tool was made is only half the question; archaeologists also want to know what it was used for. Here the founding work is Lawrence Keeley's Experimental Determination of Stone Tool Uses, published in Chicago in 1980, which launched the modern field of lithic microwear analysis. The method is patient and empirical. Researchers make replica tools, use them to work known materials such as hide, wood, bone, and meat, and then study the microscopic polishes and striations that build up along the working edges. Different materials leave characteristically different traces, and by comparing those reference patterns to the wear on ancient artifacts, an analyst can now distinguish a hide-working edge from a wood-working edge from a butchering edge.

Microwear analysis matters because it puts behavior back into the stone. A flake is silent about its purpose until someone reads the polish along its edge, and that reading turns an inert object into evidence of a specific past act, scraping a hide, whittling a shaft, jointing a carcass. The approach is not infallible, since wear can be ambiguous and post-burial damage can mimic use, but it has transformed broken stones from a typological puzzle into a record of what people actually did.

More Than a Survival Kit

It is tempting to picture stone tools as crude survival gear, the bare minimum a vulnerable primate needed to scrape by, but that framing badly undersells them. The two-million-year record of stone toolmaking is the longest archaeological record of human cognition, planning, and social learning that we possess, far longer than the record of art or of anything else we might point to. Each industry is a technical accomplishment in its own right, embodying knowledge of materials and sequences of skilled action passed down across generations, which is itself evidence of teaching and learning.

The hand axe alone makes the point. Its insistent symmetry was not demanded by function, and the labor lavished on getting the form right suggests cognitive and aesthetic concerns reaching well beyond bare survival. Mary Leakey understood this discipline of careful reading better than most. Her 1971 monograph Olduvai Gorge: Excavations in Beds I and II, 1960-1963 is the founding document of African Stone Age archaeology, and her insistence on meticulous documentation, recording exactly what lay where and beside what, shaped how the archaeologists who followed her worked the African deposits.

Key Takeaways

The stone tool record runs from the Lomekwian at 3.3 million years ago, made by australopiths before the genus Homo existed, through the microliths of the Mesolithic, and it is the longest continuous record of human behavior we have. Conventionally the Stone Age splits into a Lower Palaeolithic (about 2.6 million to 300,000 years ago), a Middle Palaeolithic (300,000 to 50,000), and an Upper Palaeolithic (roughly 50,000 to 10,000), and four main industries chart the trajectory: the Oldowan of sharp percussion flakes, the Acheulean with its enduring symmetric hand axe, the Mousterian with the planned Levallois prepared core, and the Upper Palaeolithic with its serial blades, bone needles, and composite tools. A five-mode shorthand (flake, hand axe, prepared core, blade, microlith) captures the arc from opportunistic to planned, standardized, and modular. Lawrence Keeley's microwear analysis lets us read what those edges actually cut, and the named workers behind this record, Mary Leakey at Olduvai in 1959, Keeley in his lab in 1980, and Sonia Harmand at Lomekwi in 2015, remind us that the deep history of human thought is recovered one carefully documented site at a time.