Lomekwi 3: The Oldest Known Stone Tools, Found in Kenya

Lomekwi 3, a site in the badlands west of Lake Turkana in northern Kenya, has produced the oldest known stone tools in the world, made about 3.3 million years ago. Found in 2011 and published in 2015, the tools are some 700,000 years older than the classic Oldowan industry and older than any known fossil of our own genus, Homo. They suggest that the habit of shaping stone began with earlier, more ape-like human relatives.

Background: who made the first tools?

For much of the twentieth century, toolmaking was seen as a defining trait of the genus Homo. When Louis Leakey, Phillip Tobias and John Napier named Homo habilis, the “handy man”, in 1964, they linked the new species to the simple stone tools of Olduvai Gorge in Tanzania. The idea that tools and Homo arrived together became a textbook assumption.

That picture began to shift in 1997, when stone tools about 2.6 million years old were reported from Gona in Ethiopia, close to the age of the earliest known Homo. In 2010, researchers working at Dikika in Ethiopia described two animal bones, dated to about 3.39 million years ago, with marks they interpreted as cuts and percussion damage from stone tools. No tools were found there, and other scientists argued that the marks could have been made by trampling or by crocodiles, so the question stayed open.

The area west of Lake Turkana already had a long history of fossil hunting. In 1999 a team led by Meave Leakey found a 3.5-million-year-old skull at Lomekwi, which in 2001 became the type specimen of a new species, Kenyanthropus platyops. Sediments of the same general age are widely exposed in the Nachukui Formation along the western shore.

The discovery in 2011

On 9 July 2011, members of the West Turkana Archaeological Project, co-led by Sonia Harmand of Stony Brook University in New York and the French research agency CNRS, were heading for a different target when they took a wrong turn along a dry riverbed. Exploring the slopes on foot, they spotted stone flakes and cores lying on the surface. Some of the pieces were weathering out of sediment that appeared to be very old.

Excavations in 2011 and 2012 uncovered additional artefacts still embedded in the ground. The team reported 149 artefacts in total, including cores (the blocks from which flakes were removed), flakes, possible anvils and possible hammerstones. Most had been collected from the surface, while a smaller group, around twenty pieces, was excavated in place within the sediment layer. The project worked with the National Museums of Kenya and the Turkana Basin Institute.

Harmand presented the find at a scientific meeting in April 2015, and the full description appeared in the journal Nature on 20 May 2015.

How the tools were dated

The Turkana Basin is well suited to dating because volcanic eruptions in the region repeatedly spread ash across the landscape. These ash layers, called tuffs, contain crystals that can be dated with the argon-argon method, and many of them have been traced and dated across the whole basin.

At Lomekwi 3, the artefact-bearing layer lies between dated tuffs, including the Toroto Tuff of about 3.31 million years, which bracket its age. The researchers also measured the magnetic signature of the sediments. The layer falls within a short interval of reversed magnetic polarity known as the Mammoth subchron, about 3.33 to 3.21 million years ago, which supports an age of roughly 3.3 million years.

What the tools look like

The Lomekwi 3 tools are larger and heavier than most Oldowan artefacts. Many cores weigh several kilograms, and some of the stones interpreted as anvils are bulky blocks of volcanic rock. The flakes are relatively few and large compared with the sharp, small flakes typical of later industries.

The excavators concluded that the makers used two main methods. In one, a core was held on an anvil and struck from above, a technique called bipolar knapping. In the other, a core may have been swung or thrown against a stationary anvil, the “passive hammer” technique. Both are less controlled than the freehand knapping seen in the Oldowan, where a hammerstone held in one hand strikes a core held in the other.

Because of these differences, Harmand’s team proposed a new name, the Lomekwian, for the technology. They argued that it stood between the stone pounding seen in chimpanzees and other primates and the more skilled flaking of the Oldowan.

Who could have made them?

No hominin fossils were found with the tools, so their makers remain unknown. Two candidates are most often named. Kenyanthropus platyops is known from the same part of West Turkana and from sediments of similar age. Australopithecus afarensis, the species of the famous skeleton Lucy, lived in East Africa between about 3.85 and 2.95 million years ago and had hands capable of fairly strong, precise grips.

The earliest fossil currently assigned to Homo, a jawbone from Ledi-Geraru in Ethiopia, is about 2.8 million years old. Lomekwi 3 is therefore around half a million years older than the first evidence of our genus, which suggests that toolmaking began among australopith-like hominins.

Debate and doubts

The find was widely accepted as evidence of deliberate knapping, but it also drew criticism. In 2016, Manuel Domínguez-Rodrigo and Luis Alcalá questioned whether the excavated pieces could be firmly tied to the dated layer, arguing that the stratigraphy and the position of the surface finds needed clearer documentation. The Lomekwi team replied with further details of the geology and maintained their dating.

Another line of discussion concerns what counts as a “tool”. In 2016 researchers reported that wild bearded capuchin monkeys in Brazil, while pounding stones, sometimes break off sharp flakes that resemble simple human-made flakes, although the monkeys do not use them. This shows that flaked stone can be produced without any intent to make a cutting edge. Supporters of Lomekwi 3 point out that the patterns of flake removal there are repeated and systematic, which suggests deliberate production.

Later discoveries have also filled in the gap between Lomekwi and the Oldowan. At Nyayanga in western Kenya, Oldowan tools dated to between about 3.0 and 2.6 million years ago were published in 2023, found together with teeth of Paranthropus, another non-Homo hominin.

Timeline

• c. 3.5 million years ago: The Kenyanthropus platyops skull later found at Lomekwi is deposited in West Turkana sediments.
• c. 3.3 million years ago: Hominins knap stone at Lomekwi 3.
• c. 2.6 million years ago: The earliest widely accepted Oldowan tools are made at Gona, Ethiopia.
• 1999: Meave Leakey’s team finds the Kenyanthropus skull at Lomekwi; the species is named in 2001.
• 2010: Possible cut-marked bones about 3.39 million years old are reported from Dikika, Ethiopia.
• 9 July 2011: The West Turkana Archaeological Project discovers stone artefacts at Lomekwi 3.
• 2011 to 2012: Excavations recover artefacts in place in the dated sediments.
• 20 May 2015: Harmand and colleagues publish the find in Nature.

Why it matters

Lomekwi 3 changed the question of when human technology began. If its dating holds, stone toolmaking started before the genus Homo and was not the invention of a single “handy” species. The site places the roots of the long sequence that led to the Oldowan and later industries deep in the Pliocene, among hominins whose brains were still roughly the size of a chimpanzee’s.