Archaeologists named the Stone Age for its most durable material. The wood disagrees

Two pieces of shaped wood, one alder, one willow or poplar, sat in waterlogged sediment thirty meters below the surface of a dry Greek valley for approximately 430,000 years. When researchers from the University of Tübingen, the University of Reading, and the Senckenberg Nature Research Society extracted and analyzed them, they found something the record had not prepared anyone for: microscopic cutting and carving marks, left by a human hand, on the oldest wooden handheld tools ever recovered. The site was Marathousa 1, a fossil lakeshore in Greece’s central Peloponnese. The publication was in the Proceedings of the National Academy of Sciences. The lead researchers, Professor Katerina Harvati and Dr. Annemieke Milks, had pushed the wooden tool record back by at least 40,000 years.

The headline — “oldest wooden tools ever found” — misses the argument. The interesting question is not whether these pieces are old. It is why they surprised anyone. The answer is damaging to the way archaeology has narrated human prehistory. These tools survived because the lakeshore at Marathousa denied oxygen to the microbes that decompose organic matter. Every other lakeshore, every forest floor, every campsite where early humans shaped and used wooden tools across the hundreds of thousands of years before this deposit formed: those tools rotted. Their absence from the record is not evidence they did not exist. It is evidence that wood does not last. The “Stone Age” — the term applied to 3.4 million years of human prehistory — is, in a real and underacknowledged sense, a record problem. We named an era for its most durable material, then built a theory of human cognitive evolution on what that material happened to preserve.

This is not an argument against the importance of stone tools. It is an argument about the danger of building a comprehensive theory from an incomplete sample.

The Marathousa find arrives in a sequence of wooden discoveries accumulating force for thirty years. In 1995, researchers at Schöningen, Germany, found eight wooden throwing spears, spruce and pine, dated to roughly 300,000 years ago — balanced with their center of gravity one third from the tip, the same distribution as modern javelins. The Kalambo Falls find, published in Nature in 2023, pushed the argument further. A team led by Larry Barham of the University of Liverpool documented a wooden structure in Zambia — two interlocking logs joined by a deliberately cut notch — dated to at least 476,000 years ago, predating Homo sapiens entirely. As Barham argued in The Conversation, we had been “wrong to underestimate our ancient relatives” — and the underestimation was at least partly methodological.

What the Marathousa tools add is not just an older date. They push the evidence deeper into what Harvati described to SciTechDaily as “a critical phase in human evolution, during which more complex behaviors developed.” Discover Magazine noted the site’s waterlogged conditions “created exceptional preservation,” revealing “early human technology included more than stone, even if most of that evidence has since decayed.” The World of Paleoanthropology has described the wooden tool problem as “the missing half of the Paleolithic toolkit.”

The case for stone

The counter-argument carries real weight. Stone tools required genuine cognitive investment. The Levallois technique requires planning a sequence of preparatory flakes before striking the final form — abstract spatial reasoning, not improvisation. Acheulean hand axes, produced from 1.75 million years ago, carry bilateral symmetry that implies a mental template before the first strike. These behaviors are directly readable in the material. Wood and stone served different functions: stone for cutting, wood for digging, throwing, construction. Stone was not dominant by default but by design.

But the problem is not whether stone was useful. It was. The problem is the story built from stone alone. The baseline on which the cognitive ascent was measured was always incomplete. We were tracing one thread of a cable and calling it the cable. The Schöningen spears were extraordinary in the record. They were not extraordinary for the humans who made them.

What we know / What is disputed

Established: Marathousa 1 tools are the oldest handheld wooden tools ever recovered (430,000 years, PNAS 2026). Kalambo Falls wooden structure: at least 476,000 years old, Nature 2023, predates Homo sapiens. Schöningen spears: approximately 300,000 years, aerodynamically precise. No findings disputed in core dating or material description.

Disputed: Which hominin species made the Marathousa tools (likely Homo heidelbergensis or pre-Neanderthal, unconfirmed). Whether wooden tools were as widespread as stone implies. Whether cognitive timelines require revision. Whether the “Stone Age” label distorts research investment.

What is not in dispute: the era we have called the Stone Age has been read through a filter we did not choose. The tools that lasted told us the story. The tools that did not last are only now beginning to answer.