Could an Industrial Civilisation Have Existed on Earth Before Humans?

Background

In April 2018, NASA climate scientist Gavin Schmidt and University of Rochester astrophysicist Adam Frank published a peer-reviewed paper in the International Journal of Astrobiology titled "The Silurian Hypothesis: Would It Be Possible to Detect an Industrial Civilization in the Geological Record?" The name, borrowed with acknowledged tongue-in-cheek intent from the reptilian "Silurians" of the British science-fiction series Doctor Who, imagined as a civilisation that predated humans on Earth, signals the paper's actual purpose: not to argue that such a civilisation existed, but to work out, rigorously, what evidence it would leave if it had, and whether that evidence could plausibly survive tens of millions of years of geological churn.

The question has genuine scientific motivation beyond idle speculation. Schmidt and Frank's stated goal was partly to sharpen how astrobiologists think about detecting technological civilisations on other planets, where researchers cannot excavate and must infer activity from remote geochemical signatures alone, and partly to use humanity's own rapidly accumulating Anthropocene footprint, the carbon isotope shift from fossil fuel burning, plastic pollution, altered sediment patterns, as a concrete test case for what a future, far-distant observer would actually be able to detect of us.

Main Theories

What the paper actually proposes

Schmidt and Frank's central argument is that direct physical evidence, tools, bones, structures, would almost certainly not survive tens of millions of years: fossilisation is rare even for entire species, land makes up a small fraction of Earth's surface and is disproportionately subject to erosion and tectonic recycling, and human cities, roads, and buildings themselves cover only a tiny percentage of the planet's total surface area, all combining to make direct artefactual survival across deep time highly improbable. Instead, they argue any detectable trace would have to be indirect and geochemical: a disrupted carbon-isotope excursion resembling the one industrial fossil-fuel combustion is currently producing, unusual concentrations of synthetic or rare elements inconsistent with natural geological processes, or nitrogen and plastic-derived compounds persisting in sediment layers. They explicitly compare this hypothetical signature to the Palaeocene-Eocene Thermal Maximum (PETM), a real, well-documented episode roughly 56 million years ago when a large, rapid carbon release warmed the planet significantly, arguing a prior civilisation's industrial period could plausibly leave a similarly abrupt isotopic spike, distinguishable in principle from purely natural warming events if researchers knew specifically what combination of signatures to look for.

The critique: underestimating what survives

Geologist and palaeontologist Jan Zalasiewicz, in his 2025 book Discarded: How Technofossils Will Be Our Ultimate Legacy, called the Silurian hypothesis "a wonderful and deeply thought-provoking idea" while arguing its authors underestimated both how thoroughly humans have already transformed the planet and how effectively the fossilisation process actually preserves a legible record. Zalasiewicz's argument centres on the sheer chemical novelty of what industrial civilisation produces: smelted metals, alloys, and above all plastics are materials with no natural geological precedent, meaning their presence in strata would be immediately and distinctively anomalous rather than requiring the kind of subtle isotopic detective work Schmidt and Frank's more conservative, indirect-evidence-only scenario assumes. On this view, a genuinely industrial prior civilisation, one comparable in scale and material output to modern humanity, would likely leave a record harder to miss, not easier, than the original paper's cautious framework implies.

The Oklo reactor: a working test of the underlying logic

The paper's authors themselves point to a real geological case that demonstrates their central methodological point in miniature: the Oklo natural nuclear reactor, discovered in Gabon, West Africa, on 25 September 1972, when French nuclear-fuel processors noticed uranium ore from the site was measurably depleted in uranium-235, some samples as low as 0.29 percent against the normal 0.72 percent found everywhere else on Earth. The cause was not artificial: roughly two billion years ago, when natural uranium-235 concentration was higher, a uranium-rich rock formation at Oklo became critical and sustained sixteen separate self-moderating fission reactions, cycling on and off over hundreds of thousands of years as groundwater alternately moderated the reaction and boiled away to interrupt it. Physicist Paul Kuroda had theoretically predicted the precise conditions under which such a natural reactor could form back in 1956, sixteen years before Oklo confirmed it. Oklo demonstrates exactly the kind of subtle, easily overlooked geochemical anomaly, an isotope ratio slightly off from expectation, the Silurian hypothesis argues a genuine prior-civilisation signature might resemble; crucially, however, Oklo's own chemistry and geological context make clear it is entirely natural, underscoring how carefully any candidate signature would need to be distinguished from a naturally occurring process before being read as evidence of anything artificial. Oklo sustained fission roughly two billion years before the Manhattan Project built the first human-made reactor, a gap often cited to illustrate how unremarkable the underlying physics is once the right natural conditions happen to occur, with no engineer required at all.

Common Misconceptions

The Silurian hypothesis is frequently mistaken for a claim that a pre-human civilisation actually existed, sometimes cited alongside ancient-astronaut or lost-civilisation claims as though it lends them scientific credibility. It does the opposite: the paper is explicitly agnostic, states plainly that no evidence currently supports such a civilisation, and was written specifically as a methodological tool for astrobiology and Anthropocene research, not as a defence of any ancient-civilisation or ancient-astronaut claim, both of which Schmidt and Frank's own framework, if applied honestly, argues against by highlighting exactly how little unambiguous evidence any of those claims can actually point to.

It is also sometimes assumed the hypothesis proposes humans are not the first tool-using or intelligent species on Earth, a much broader and different claim than the one the paper makes. Other hominins and even some non-human animals are known to have used simple tools, a well-established finding unrelated to Schmidt and Frank's specific, much narrower question about a technologically industrial civilisation capable of producing a detectable planetary-scale geochemical signature.

Current Consensus

Astrobiologists and geologists treat the Silurian hypothesis as a legitimate, well-constructed thought experiment rather than a proposed historical event, useful precisely because it forces precision about what evidence for a technological civilisation, ours or a hypothetical prior one, would actually need to look like across deep time. There is no scientific dispute that no evidence for a pre-human industrial civilisation currently exists; the live methodological question, on which qualified researchers disagree, is how detectable such a civilisation's traces would actually be, with Zalasiewicz's 2025 critique arguing for greater confidence in fossilisation's power to preserve a legible record than Schmidt and Frank's original, more conservative framework assumed.

Why This Mystery Endures

The hypothesis endures partly because of the intellectual discipline that makes it unusual among the claims this site covers: unlike lost-civilisation or ancient-astronaut theories, which typically start from a specific artefact and reason toward a dramatic conclusion, Schmidt and Frank started from a rigorous evidentiary question and refused to answer it with anything beyond what the geochemistry actually supports, which is itself part of the paper's appeal. It offers readers drawn to out-of-place artifact claims a genuinely scientific version of the same underlying curiosity, using real, publishable astrobiological method rather than a single striking photograph and an unverified age estimate.

The idea also endures because it reframes a familiar anxiety, what would actually survive of us, into a specific, checkable question, which is precisely what has kept it circulating in renewed media coverage years after its original 2018 publication. Confronting how thoroughly geological time can erase a civilisation's material trace, and how carefully a genuine signature would need to be distinguished from something as easily misread as the entirely natural Oklo reactor, is a more sobering exercise than most speculative "lost civilisation" claims attempt, and considerably harder to dismiss, because it was never trying to prove one existed in the first place. This page joins abiogenesis, panspermia, and the dire wolf de-extinction dispute in this site's biology and origins of life hub.