AI-generated conjecture · below the evidence/publication boundary
← All conjectures · Texts, scribes & transmission
The scroll tax on the tail
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)
The scroll tax on the tail. Leishu quotation habits should encode the physical access cost of the scroll. A compiler excerpting a source before the late Tang paid a sequential-access price — matter at the head of a juan costs one motion of the roller, matter at the tail costs the whole unrolling — while codex-style bindings and print made access random. So the positional bias of encyclopedia quotation (how strongly excerpts cluster toward the front of the source juan) is not a fixed habit of Chinese compilation but a function of book format, and it should visibly relax between a scroll-age leishu and a codex-age one.
Prediction clause (verbatim)
For source works that survive independently, locate each quotation's position within its source juan by decile of graph count. Primary clause (verdict follows it): in the Yiwen leiju (624), the share of quotations drawn from the first three deciles of the source juan exceeds the corresponding share in the Taiping yulan or Cefu yuangui (10th-11th c.) by at least 10 percentage points. Secondary: within each leishu, head-bias strengthens with source-juan length, as an access-cost account requires.
Kill-dataset (verbatim)
Kill: the Yiwen leiju and Taiping yulan full texts on the Kanseki Repository (Kanripo), collated against extant source texts (Shiji, Hanshu, Wenxuan) also on Kanripo. Resolvable against Kanripo without new ingestion.
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
no prior located yet (provisional)
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.
[Independent blind re-audit 2026-07-08 (2nd pass, generator-independent), confidence medium. Double-confirmed: both the provisional pass and this re-audit located no prior formulation of the specific operationalized claim.] Leishu excerpting practice is well studied (Oxford Handbook chapter on encyclopedias/epitomes; standard accounts of Yiwen leiju and Taiping yulan), and the general scroll-vs-codex 'random access' argument is a familiar theme in book history (invoked e.g. for the Christian adoption of the codex). But I located no study testing THIS specific materialist claim: that the positional bias of encyclopedia quotation (share of excerpts drawn from the first three deciles of the source juan) is significantly higher in a scroll-age leishu (Yiwen leiju, 624) than in codex/print-age ones (Taiping yulan / Cefu yuangui), by >=10 points, with within-leishu head-bias strengthening with source-juan length. The access-cost-of-format explanation for quotation position appears to be an untested, novel operationalization.
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.