AI-generated conjecture · below the evidence/publication boundary
← All conjectures · Texts, scribes & transmission
Piecework prices its errors
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).
Claim (verbatim)
Piecework prices its errors. The Nara sutra-copying bureau ran a complete incentive scheme that survives as paperwork: scribes paid per sheet, proofreaders paid to catch faults, pay docked per uncaught error under a written deduction schedule. That is a speed-accuracy market, and it should have an interior optimum visible in the ledgers: across named scribes, output rate and fault rate positively correlated (the frontier), and the distribution of per-scribe fault rates compressed just below the threshold where the fines bite — eighth-century workers optimizing against a fee schedule, recoverable scribe by scribe from the payroll.
Prediction clause (verbatim)
From the copying-bureau ledgers (sheets assigned and completed per day, proofreaders' fault tallies per scribe), for scribes with at least 20 attested work-days: Primary clause (verdict follows it): the across-scribe rank correlation between mean daily sheet output and faults per 100 sheets is positive with rho at least 0.4. Secondary: the per-scribe fault-rate distribution shows excess mass just below the documented deduction threshold rather than a smooth right tail.
Kill-dataset (verbatim)
Kill: the Shosoin documents (Shosoin monjo) as transcribed in Dai Nihon Komonjo and databased by the Shosoin Office and the National Institute of Japanese Literature — the sutra-scriptorium payroll and proofreading records form the core of the corpus.
Provenance
Run: Fresh agent generation · model: claude-fable-5
Generated by a fresh Fable-tier instance (claude-fable-5) at maximum effort with generation-first blindness (no repo reads, no web searches, no DB queries); 248-title list supplied at launch, titles only, no verdicts/dossiers seen; prompt pre-committed in docs/generated/conjectures_1001_wave_ledger.md and docs/generated/conjecture_fresh_fable_w04_prompt_20260708.txt. Wave W04 of docs/GOAL_CONJECTURES_1001_CAMPAIGN_20260708.md. Novelty unverified by construction. An output-token limit interrupted the first response turn before any tool call was made; the packet was still produced in a single Write with no information ingress.
Novelty / leakage triage
provisional — model-triaged, shepherd review pending
A provisional first pass authored by the model (Opus), not yet confirmed by the shepherd. It carries the same dated-search requirements as an authoritative verdict but is excluded from every headline figure and cannot underwrite a prediction until a shepherd confirms it. Provisional reading: Adjacent (closely related prior work exists).
The incentive scheme is documented in detail: the Nara Office of Sutra Transcription (shakyojo) paid scribes by the sheet (the fastest exceeding thirteen a day), had proofreaders check against the exemplar, and docked pay for uncaught errors under a written schedule -- Bryan Lowe's Ritualized Writing reconstructs exactly this speed-accuracy market from the Shosoin documents. Not located: the item's specific ECONOMETRIC signature -- an across-scribe positive rank correlation (rho >=0.4) between mean daily sheet output and faults per 100 sheets (the frontier), and a per-scribe fault-rate distribution with excess mass just below the deduction threshold rather than a smooth right tail. The payroll-and-penalty scheme is established; the interior-optimum measurement scribe-by-scribe is unrun.
- Bryan D. Lowe, Ritualized Writing: Buddhist Practice and Scriptural Cultures in Ancient Japan (University of Hawaii Press, 2017); 'Scribes and Purity in Eighth-Century Japan' — The Nara scriptorium's pay-per-sheet, proofreading, and error-docking scheme -- the established incentive market
- Shosoin monjo transcribed in Dai Nihon Komonjo and databased by the Shosoin Office / NIJL (scriptorium payroll and proofreading tallies) — The kill dataset for the output-vs-fault correlation and threshold-clustering test
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.