Twenty-one memories of a drowned coast

Claim (verbatim)

Patrick Nunn and Nicholas Reid assembled twenty-one Aboriginal Australian traditions, from around the continent's edge, that describe coastlines, islands and features now underwater - land that was dry before postglacial seas rose (Australian Geographer, 2016). Each story, if it truly remembers the drowning, implies that its coast stood at a former, shallower sea level, and postglacial sea-level curves convert that implied shoreline into a date. The claim is astonishing - transmission across seven to thirteen thousand years - and precisely because it is astonishing it must be tested against the geology location by location, not vindicated by a single headline number. The disciplined operation asks whether the twenty-one implied inundation dates, each read off its own place's bathymetry against the regional sea-level curve, cluster in the narrow postglacial window and stay internally consistent, or whether the depths are miscellaneous and the 'dates' scatter across the whole Holocene and beyond as random matching would produce. The missing knowledge is whether oral tradition can carry factual memory across millennia; it is decidable from the fit between twenty-one named places' depths and a published sea-level curve.

Prediction clause (verbatim)

Prediction: converting each of the 21 Nunn-Reid locations to an inundation date by matching its named drowned feature's depth to the regional postglacial sea-level curve, at least 18 of 21 implied dates will fall inside the 7,000-13,000-years-ago window, and the across-site scatter will be consistent with the sea-level curve's slope rather than with depths drawn at random (primary clause: >=18/21 inside the postglacial window). Disambiguation: only locations with independent bathymetric/geomorphic control on the named feature enter the count; purely generic 'the sea came in' stories without a locatable feature are excluded, not scored as hits. Coverage guard: if independent bathymetry exists for fewer than 15 of the 21 locations the test voids rather than resolving.

Kill-dataset (verbatim)

Kill: Nunn & Reid (2016, Australian Geographer) Table of 21 locations, matched against a published Australian postglacial sea-level reconstruction (Lambeck et al.) and bathymetric/geomorphic data for each named feature - the operation is the count of implied inundation dates falling in the 7-13 ka window and the internal consistency of their scatter with the sea-level curve.

Novelty / leakage triage

already answered in the literature

Nunn & Reid (2016, Australian Geographer 47/1) is itself the operation the conjecture proposes. For each of the 21 traditions they computed the minimum water depth the story requires and matched it to a postglacial sea-level reconstruction, concluding that all 21 correspond to inundation between roughly 13,070 and 7,250 years BP. The primary clause (>=18/21 implied dates inside the 7-13 ka window) restates their published result - their thesis is in fact the stronger near-21/21 - so the count is already run and reported, not merely enabled by an available dataset.

• P.D. Nunn & N.J. Reid, 'Aboriginal Memories of Inundation of the Australian Coast Dating from More than 7000 Years Ago,' Australian Geographer 47(1) (2016): 11-47
• P.D. Nunn, The Edge of Memory: Ancient Stories, Oral Tradition and the Post-Glacial World (London: Bloomsbury Sigma, 2018)

Predictions

No prediction registered yet.