The verses walk in order

Claim (verbatim)

The song cycles of Central Australia are, among everything else they are, geographic information systems: verse sequences track the travels of ancestral beings site by named site, and the holders could navigate by them across country they had never crossed. The fullest written record of any part of the system is T. G. H. Strehlow's Songs of Central Australia (1971), the distillation of decades with Arrernte song-holders — a record whose unpublished depth (diaries, film, sound) is held by the Strehlow Research Centre in Alice Springs under the access protocols that restricted ceremonial knowledge requires, and whose making is chronicled in Barry Hill's Broken Song (2002). The navigational reading is standard; what has never been run on the published corpus is the cartographic test: if verse order encodes route order, then plotting a cycle's identifiable sites should reproduce a traversable path — consecutive verses geographically adjacent far beyond chance — and every cycle that died unrecorded was then, quantifiably, a lost map and gazetteer, not only a lost poem. The conjecture concerns the historical record and is to be run on published texts and gazetteers; the living tradition's restrictions govern what any test may touch, and the decisive concordance must be built under them, not around them. Prediction: for at least three song cycles in Songs of Central Australia with ten or more independently identifiable named sites, the rank correlation between verse order and the order of the sites along the best-fitting geographic route will reach at least 0.8, with consecutive-verse site pairs geographically adjacent at a rate unmatched by random permutation of the same sites in 99 of 100 trials (primary clause: the ≥0.8 rank correlation in at least three cycles; the verdict follows it). Kill (not yet built, from real sources, access-governed): a geocoded site concordance for the published cycles, built from Songs of Central Australia (1971), the Tindale place-name and map collections at the South Australian Museum, and — under its protocols — the records of the Strehlow Research Centre, Alice Springs.

Prediction clause (verbatim)

Prediction: for at least three song cycles in Songs of Central Australia with ten or more independently identifiable named sites, the rank correlation between verse order and the order of the sites along the best-fitting geographic route will reach at least 0.8, with consecutive-verse site pairs geographically adjacent at a rate unmatched by random permutation of the same sites in 99 of 100 trials (primary clause: the ≥0.8 rank correlation in at least three cycles; the verdict follows it).

Kill-dataset (verbatim)

Kill (not yet built, from real sources, access-governed): a geocoded site concordance for the published cycles, built from Songs of Central Australia (1971), the Tindale place-name and map collections at the South Australian Museum, and — under its protocols — the records of the Strehlow Research Centre, Alice Springs.

Novelty / leakage triage

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

The navigational reading of Central Australian song cycles — verse sequence tracking ancestral travel site by named site — is standard, and Strehlow's Songs of Central Australia (1971), with its making chronicled in Hill's Broken Song (2002), is the fullest published corpus. But the cartographic test itself has never been run on the published cycles: no rank correlation between verse order and geographic route order (at least 0.8 across three or more cycles), with a permutation null, was located as ever having been computed, and the concordance is honestly flagged not-yet-built and access-governed. The reading is in print; the geometry is un-run. Adjacent.

• T. G. H. Strehlow, Songs of Central Australia (Angus & Robertson, 1971)
• Barry Hill, Broken Song: T. G. H. Strehlow and Aboriginal Possession (Knopf, 2002)

Predictions

No prediction registered yet.