Originally published 2021-07-15.
The Silurian hypothesis is a thought experiment, not a discovery. In 2018 climate scientist Gavin Schmidt, then director of NASA's Goddard Institute for Space Studies, and astrophysicist Adam Frank of the University of Rochester asked a simple but strange question: if an industrial civilisation had existed on Earth millions of years before humans, would we be able to tell? Their answer was that most direct traces would be gone, but certain chemical and climate signals might survive in rock. They were clear that they do not think such a civilisation existed.
The idea is often shared online as "scientists say advanced beings lived on Earth millions of years ago". That is not what the paper says. What it actually offers is more interesting: a careful look at how long evidence lasts, and at what our own civilisation will leave behind.
Where the paper and the name come from
The study, titled "The Silurian Hypothesis: Would it be possible to detect an industrial civilization in the geological record?", was posted online in April 2018 and published in the International Journal of Astrobiology. Many news stories about it appeared in 2018, and versions of the story kept circulating for years afterwards.
The name is a nod to science fiction. In the BBC series Doctor Who, the Silurians are an intelligent reptile species that ruled Earth long before humans and then went into hibernation underground. They first appeared on screen in 1970. The authors borrowed the name for fun; it has nothing to do with the real Silurian period of around 440 million years ago, when life on land was only just beginning.
Why ancient evidence is so hard to find
The starting point of the paper is how little of the past survives. Humans in our modern form have existed for about 300,000 years, and industrial technology for only a few centuries. On geological timescales, that is a blink.
Now imagine trying to spot a civilisation of similar length from, say, 50 million years ago. Several problems pile up:
- Old surfaces are rare. Erosion, plate tectonics and sediment constantly recycle the Earth's crust. Very little of the land surface we can see today dates from before the Quaternary period, which began about 2.6 million years ago.
- Fossilisation is unusual. Only a tiny fraction of living things ever become fossils. Cities, roads and machines would be crushed, eroded or buried deep, and the chances of finding a recognisable object would be very small.
- Short events are easy to miss. A few centuries of industry would show up, at best, as a very thin layer in sediments that took thousands of years to build up.
For these reasons, Schmidt and Frank argued, looking for buildings or tools is not a promising strategy. Indirect evidence is a better bet.
What traces an industrial civilisation might leave
To work out what to look for, the authors started with the one industrial civilisation we know: our own. The Anthropocene, the name many scientists use for the age of major human impact, is already leaving a distinct signature in sediments. Among the signals they discussed:
- Changes in carbon isotopes. Burning coal, oil and gas releases carbon with a particular isotopic fingerprint, shifting the balance of carbon isotopes in the air, oceans and sediments. Frank summed up the irony that burning fossil fuels "may actually shut us down as a civilization."
- Climate and ocean shifts. Rapid warming, ocean acidification and sudden changes in which species thrive can be read in layers of rock and in microfossils.
- Fertiliser and erosion signals. Large-scale farming changes nitrogen levels and the amount of sediment rivers carry to the sea.
- Synthetic materials. Plastics and other artificial compounds that do not occur in nature could leave traces in ocean sediments.
- Unusual isotopes. Some radioactive isotopes, such as plutonium-244, are not produced naturally except in rare events like supernovae, so a layer rich in them would be a strong clue.
Natural events that look strangely familiar
The most thought-provoking part of the paper is its central finding: our own geological fingerprint, while clear, will not look very different from some natural events already in the record.
The best-known example is the Paleocene–Eocene Thermal Maximum, a sudden episode of global warming roughly 55 to 56 million years ago. Huge amounts of carbon entered the atmosphere, temperatures rose sharply, and the oceans became more acidic. Similar but smaller spikes, as well as ancient periods when large parts of the oceans lost their oxygen, show some of the same patterns.
Schmidt and Frank stress that these events have natural explanations, such as volcanic activity and the release of carbon stored in sediments. Their point is that a scientist millions of years from now, looking at our layer of rock, might struggle to tell it apart from such episodes. So they proposed tests that might separate an industrial cause from a natural one, for example looking for a combination of several unusual signals arriving at the same moment.
There is even a precedent for nature mimicking technology. At Oklo in Gabon, uranium deposits about two billion years ago reached conditions in which nuclear fission ran on its own, with no engineers involved. A future investigator finding depleted uranium could easily jump to the wrong conclusion.
Could a civilisation really have existed before us?
The authors are direct about this: they wrote that they "strongly doubt" any earlier industrial civilisation existed. There is no evidence for one, and the hypothesis was never meant as a claim that there was.
Some of the conditions would have been available, though. Most of the world's coal formed during the Carboniferous period, around 350 million years ago, so in principle a much earlier technological species could have found fossil fuels to burn. The question is not whether this happened, but whether we would even notice if it had. That is useful for astrobiology too: when we search other planets for signs of past civilisations, we need to know which traces would last.
What it says about our own future
The thought experiment ends up pointing back at us. Frank argued that the real challenge is building a large, thriving civilisation without pushing the planet into conditions that are dangerous for that civilisation itself. In his view, the Earth will carry on whatever we do. As he put it, "we're just creating niches for the next cycle of evolution." The open question is whether humans will still be part of the story.
The Anthropocene is still debated as a formal geological unit, and in 2024 geology's international governing bodies declined to add it as an official epoch. But nobody disputes that humans are changing the chemistry of the air, oceans and rock. Millions of years from now, that change may be the only sign that we were here.
Common misunderstandings
- "A study proved an ancient civilisation existed." It did not. It asked how such a civilisation could be detected, and the authors doubt one existed.
- "It is about the dinosaurs." No species is named as a candidate. Sentient dinosaurs belong to science fiction, which the paper mentions only for the name.
- "The research is from 2021." The paper dates from 2018; later headlines simply recycled it.
Sources: Schmidt and Frank: The Silurian Hypothesis (arXiv, 2018); Wikipedia: Silurian hypothesis; University of Rochester news release via EurekAlert (Adam Frank quotes); first reported to us via IFLScience.
COMMENTS