The sister the whole world lost

Claim (verbatim)

The Pleiades are named as seven almost everywhere, yet only six are easily seen; and across Greece, Aboriginal Australia and Native North America the cluster is told as seven sisters of whom one is lost, hidden, faint, or fled. Ray and Barnaby Norris have suggested this near-universal 'seven-but-six, one lost' framing could be a shared memory tens of thousands of years old, from a time when a now-dimmed or wider-separated star made seven genuinely visible. The suggestion is spectacular and, told as a story, untestable. Made countable it is not. Independent invention has a strong null here: any culture counting a tight naked-eye cluster will often land near six or seven and will sometimes narrate a 'missing one' simply because the count wavers at the limit of vision. The question is whether the SPECIFIC 'seven, but one is lost' motif recurs across independent traditions far more often than that baseline predicts, and far more than the same 'n-but-one-missing' framing attaches to other star groups. The missing knowledge is whether a real prehistoric sky-memory is encoded; it is decidable by comparing the lost-sister motif's cross-cultural rate against an independent-invention baseline built from other asterisms.

Prediction clause (verbatim)

Prediction: across a fixed ethnoastronomy corpus, the fraction of independent traditions that name the Pleiades as seven while telling of one lost/hidden/faint member will exceed, by a factor of at least 3, the base rate at which a 'canonical count but one missing' narrative attaches to other comparably prominent asterisms (Orion's belt, the Hyades, the Big Dipper) in the same corpus (primary clause: the >=3x excess over the cross-asterism baseline). Disambiguation: only traditions with independent attestation (not descended from a shared recent source, per the corpus's own genealogy) are counted, and 'lost sister' requires an explicit narrative of a diminished/absent member, not merely a count of six. Coverage guard: if fewer than 30 independent Pleiades traditions are documented the test voids rather than resolving.

Kill-dataset (verbatim)

Kill (not yet built): Berezkin's Analytical Catalogue Pleiades motifs together with a compiled Seven-Sisters ethnoastronomy table drawn from real sources (Ray & Barnaby Norris' Pleiades work; Munya Andrews, "The Seven Sisters of the Pleiades"; regional ethnoastronomy surveys) - code each independent tradition for 'seven, one lost' versus alternatives, build the same-form baseline from other asterisms, and compute the ratio.

Novelty / leakage triage

anticipated in the literature — this exact test has never been run

The hypothesis and the cross-cultural pattern are published - Norris & Norris (2021, 'Why Are There Seven Sisters?') document the 'seven-but-six, one lost/faint' framing across Greek, Aboriginal Australian, Native American and other traditions and argue an antiquity possibly ~100 ka. But they do not build the conjecture's baseline: coding a fixed ethnoastronomy corpus for the rate at which a 'canonical count but one missing' narrative attaches to other prominent asterisms (Orion's belt, the Hyades, the Big Dipper) and showing the Pleiades motif exceeds it by >=3x. That cross-asterism excess-ratio test - the operation that would separate a real sky-memory from the generic wavering of counts at the limit of vision - was not located.

Predictions

No prediction registered yet.