The stick outlived its sentence

Claim (verbatim)

Aboriginal Australian message sticks are the continent's portable notation: marked boards carried between groups, whose incisions fixed the count, the parties and the gravity of a communication while the message itself travelled in the messenger's commissioned speech. Notation and reading were separable by design — Howitt (The Native Tribes of South-East Australia, 1904) and R. H. Mathews (American Anthropologist, 1897) both record that the stick authenticated and prompted while the messenger delivered — and museums then performed the separation permanently: collectors acquired sticks by the hundred, and the message, resident in a person rather than the object, was recorded only where somebody took it down at acquisition. Piers Kelly's corpus work (Journal of Material Culture, 2020, and the message-stick database assembled from museum holdings worldwide) makes the documentary consequence checkable at scale for the first time. The prediction is a double signature of reader–object separation: most surviving sticks should be mute — objects without their sentences — and the sticks that do carry recorded messages should cluster hard by collector network, because documentation was a property of who asked (Howitt's correspondent circle, Mathews, Roth in Queensland), not of the objects or the regions themselves. The corpus of marks survived; the corpus of messages died with the messengers nobody questioned. Prediction: across the assembled message-stick corpus, fewer than 25% of museum-held sticks will carry any recorded message content, and of those that do, at least two thirds will trace to a small set of named collector networks of the 1880s–1900s rather than distributing evenly across acquisitions (primary clause: the under-25% documentation rate; the verdict follows it). Kill: the Australian message-stick corpus assembled by Piers Kelly from museum collections (with Kelly, Journal of Material Culture 25, 2020), checked against the message-stick documentation in Howitt (1904), Mathews (1897) and Roth's North Queensland ethnography.

Prediction clause (verbatim)

Prediction: across the assembled message-stick corpus, fewer than 25% of museum-held sticks will carry any recorded message content, and of those that do, at least two thirds will trace to a small set of named collector networks of the 1880s–1900s rather than distributing evenly across acquisitions (primary clause: the under-25% documentation rate; the verdict follows it).

Kill-dataset (verbatim)

Kill: the Australian message-stick corpus assembled by Piers Kelly from museum collections (with Kelly, Journal of Material Culture 25, 2020), checked against the message-stick documentation in Howitt (1904), Mathews (1897) and Roth's North Queensland ethnography.

Novelty / leakage triage

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

Kelly's message-stick corpus work (Journal of Material Culture 25, 2020) assembles the museum holdings and foregrounds exactly the reader-object separation the conjecture rests on, and Howitt (1904), Mathews (1897) and Roth document that the message rode in the messenger while the stick authenticated. But the primary clause is a documentation rate — fewer than 25% of museum sticks carrying any recorded message, and of those at least two-thirds tracing to a small set of collector networks — and I could not confirm that Kelly (or anyone) has published those specific figures; the corpus is assembled, the thresholds un-verified in print. Adjacent, with the caveat that Kelly's corpus is the direct neighbour and may already report a documented-message share.

• Piers Kelly, study of the Australian message-stick corpus, Journal of Material Culture 25 (2020)
• A. W. Howitt, The Native Tribes of South-East Australia (Macmillan, 1904)
• R. H. Mathews on Aboriginal message-sticks, American Anthropologist 10 (1897)

Predictions

No prediction registered yet.