Eye-skip in sandstone

Status: Anticipated · untested

Status is derived only from the shepherd-authored triage/prediction data above -- community submissions and claims are a separate overlay and can never change it (see the participation panel below).

This is a proposed connection between two domains, generated by a language model. It is not an article and not evidence: it sits below the evidence/publication boundary. A quantitative prediction and a named kill-dataset are attached (when registered) so the claim stays falsifiable rather than merely evocative.

Claim (verbatim)

Manuscript scholars diagnose visual copying by the saut du même au même: the copyist's eye jumps from one occurrence of a word to its next occurrence, silently deleting the span between. Khmer temple inscriptions carry enormous repetitive lists — servants' names, rice quantities, field boundaries — which are the perfect trap for exactly this error, and the conjecture is that the engravers' mistakes carry the fingerprint of written exemplars: omissions and doublings clustering at repeated-element boundaries, where dictation would instead scatter phonetic confusions evenly. The mechanism is production reality — a mason copying by eye from a perishable draft inherits the draft's failure modes — so the error distribution is a structurally-necessary trace of the lost working documents, down to hints of their line format. If it holds, the existence and even the layout of Angkor's perished drafting layer becomes demonstrable purely from the geometry of mistakes in stone.

Prediction clause (verbatim)

In long list-bearing Khmer inscriptions, omissions and duplications detectable through internal arithmetic (stated totals versus enumerated items) or through parallel passages will fall at repeated-name or repeated-formula boundaries at no less than three times the rate expected if error positions were uniform within lists, and eye-skip-type faults will outnumber plausibly phonetic (dictation-type) faults. Primary clause: the threefold concentration of faults at repetition boundaries; the verdict follows it.

Kill-dataset (verbatim)

The Cœdès / DHARMA Khmer inscription inventory (K-numbers — the standard Khmer epigraphic corpus): error-typology tagging of the long enumerative Khmer-language lists.

Nobody has run this test. The kill-data is named above. If you can run it — or you know the paper that already settles it — claim the kill or submit the prior. Kills and priors are credited here, by name, as they come in.

In the atlas

This conjecture is bridged, as an L1 lead, onto these Inferpedia subject pages.

Provenance

Run: Fresh agent generation · model: claude-fable-5

Generated blind in a single Write by a fresh instance working only from the inline prompt, with no file reads, web access, or database queries.

Novelty / leakage triage

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

Eye-skip (saut du meme au meme) diagnostics for copying from a written exemplar are textbook textual criticism, and epigraphists note engravers worked from manuscript exemplars; but an error-typology census over the long Khmer enumerative lists testing repetition-boundary clustering against dictation nulls is unlocated. Thin-field flag: editorial notes on individual K-numbers may record such faults without aggregation.

Predictions

No prediction registered yet.

Weigh in

No community feedback yet.

Add your take

Posted immediately (spam is removed). Community feedback is never an adjudicated verdict and never changes this conjecture's triage label or status above.