A bird on a Neolithic pillar. Birdmen on a Pacific statue. What are we seeing—and what could explain it?
Explore the collection →The pillar, the moai & the bird ↗
Follow motif families across cultures, then see which traditions keep meeting — with objects, dates and sources.
Enter the collection →
Explore drowned landscapes, underwater discoveries and data we can test.
Explore submerged worlds →Images: Met, Open Access; Routledge, public domain. Relief: NOAA ETOPO1, modern depths.
First describe what is alike. Then ask how unusual it is, and what history could explain it.
A bird, an outstretched limb, a rounded form. Explore the pillar, moai and Orongo boulder alongside local examples and museum objects, with a history of what can be dated.
Investigate the resemblance →A dated chain of unusual shared details would strengthen transmission. Local predecessors and widespread equivalents would strengthen convergence. Selection and presentation can also amplify a resemblance.
What would distinguish these explanations? →Hancock proposes that survivors of an Ice Age civilisation carried knowledge to later societies. The test is to find distinctive inherited forms, establish their chronology and trace their transmission.
His account of Magicians of the Gods. Guest articles on his site have their own authors.
Fenton proposes a maritime tradition linking northern Australia and Island Southeast Asia with distant imagery. His specific comparisons need secure identities, dates and intermediate examples.
Better labels are progress. Correct locations are the next test.
Some earlier responses described a face on the moai’s back. The controlled follow-up improved feature labels, yet placed scorpions in the bird area. Similarity scores need these errors checked first.
Results and the next tests →Rankings change with drawing style. The combined metric performed worse than a simpler one in the rendition check.
Broad motif categories recur widely. They do not test the specific bird-and-disc scene.
Models disagreed and sometimes misread the carvings. Both high and low similarity scores need perception checks.
Claimed partners ranked first in 62.5% of lineups. Source quality, repeated pairs and recognisable summaries limit the inference.
The selected summaries show regional patterns. Those patterns alone cannot distinguish borrowing from older inheritance.
Browse motif families, see which cultures keep meeting, inspect the input audit, or read the original illustrated atlas. New photographs join existing objects; unresolved identities stay visible.
Photographs, source passages and saved results remain inspectable. The people who made these objects, and living communities’ knowledge, belong at the centre of their interpretation.
Sources and image credits · Research review · Editorial record