Around almost the entire Australian coast, Aboriginal communities have told stories of a time when the sea came inland and swallowed country that people had walked on, hunted across and named. Two researchers have argued that some of these stories may be accurate memories of real events, the drowning of the old coastline as the seas rose at the end of the last Ice Age, between roughly 7,000 and 13,000 years ago. If they are right, these would be among the oldest reliably dated oral traditions anywhere on Earth.
It is a striking claim. It is also one that careful scholars dispute, and the disagreement is worth setting out plainly rather than resolving in favour of the more romantic reading.
What the stories say, and how they were dated
The geographer Patrick Nunn, of the University of the Sunshine Coast, and the linguist Nicholas Reid, of the University of New England, gathered accounts from 21 locations around the coast and published them in the journal Australian Geographer in 2016. The stories vary, but the common thread is the sea breaking in over land that is now seabed.
The Kulin peoples around Port Phillip Bay, in what is now Victoria, tell of the bay flooding over what had been a hunting ground. Gunai accounts from Gippsland describe ancestors living on land that is now underwater. Stories from the Spencer Gulf region describe a time when the gulf was dry or marshy before the water arrived. A Ngarrindjeri account describes the creation of Kangaroo Island when the sea cut it off from the mainland.
The dating method is ingenious and is also where the argument turns. For each site, Nunn and Reid worked out the minimum depth of water that would have to be removed for the story’s details to be literally true, then read off, from the established curve of post-glacial sea-level rise around Australia, the most recent time the land could have looked that way. By that calculation the stories correspond to inundations between about 7,250 and 13,070 years ago. Around Spencer Gulf, depending on which part of the gulf the story describes, the implied age is somewhere between 9,000 and 12,000 years.
How a story could last that long
The obvious objection is that no story could survive intact for 300 or 400 generations. Reid’s answer is that some Aboriginal knowledge systems were built to prevent exactly the drift that erodes most oral traditions. In his account, stories tied to country carried kin-based obligations to tell them correctly, with the telling cross-checked across relatives in a way that worked like error correction, a scaffolding that could hold a narrative stable across very long spans.
The setting makes this less far-fetched than it first sounds. Aboriginal Australians have occupied the continent for tens of thousands of years, with long stretches of relative cultural continuity. And the inundation stories are not the only candidates for deep memory. Researchers have pointed to a Klamath account in Oregon that appears to describe the eruption that formed Crater Lake about 7,700 years ago, to stories of volcanic activity in northern Queensland, and to traditions that may record meteorite falls. None of these is beyond dispute either, but together they make the general idea harder to dismiss out of hand.
The genuine disagreement
The orthodox view in folklore and history is that oral traditions rarely survive intact much beyond a thousand years. Stories get embellished to hold an audience, reshaped by outside contact, altered by memory and by politics. On that view, a literally accurate 10,000-year-old narrative is close to impossible.
The sharper objection is epistemological. The historian David Henige argued, as reported by Science News, that the claim these stories are more than 7,000 years old can neither be verified nor falsified, and the Sydney archaeologist Peter Hiscock agreed. The problem is that the dating works only by assuming the story is a faithful observation of the geology in the first place. There is no independent way to confirm that a story has descended unbroken from the event, rather than being a more recent explanation later attached to a striking landscape, or a motif that happens to fit. Henige’s own paper on the subject carried a title that captures the bind well, describing deep-time oral tradition as impossible to disprove yet impossible to believe.
Nunn and Reid’s strongest reply is the recurrence. The same basic element, the sea coming in over land, appears at sites scattered right around the continent, and they argue that independent invention of the same specific idea in so many places is less likely than shared observation of a real process. That is a reasonable argument. It is not proof, and they have not claimed it as such.
A record that comes to us secondhand
One more thing is worth saying plainly. Many of these stories survive in the written record only because colonial-era observers wrote them down in the nineteenth century, often from communities that were then being dispossessed and devastated by introduced disease. The version scholars now analyse is mediated, partial and shaped by who was listening. For the peoples who hold them, these are not puzzles about the deep past but living ties to country, which is a different relationship to the material than the one a dating exercise implies.
What to watch
This is not a claim that will be settled the way a fossil settles a question. There is no sediment layer that will confirm a story is continuous. What the work does instead is make a case that oral tradition can sometimes function as a genuine archive, and that case strengthens or weakens as more candidate correspondences are found and scrutinised, in Australia and elsewhere, including the tsunami and earthquake traditions being examined alongside Māori accounts in New Zealand.
For now the honest summary is that the fit between the stories and the drowned coastlines is real and hard to explain by chance, and that the leap from a good fit to a proven 10,000-year-old memory is exactly the step the critics will not take. Both of those things can be true at once.
