Memory quotes don't decay

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)

Two storage systems fed scholastic quotation: verbatim memory, installed by years of chanting the Psalter in choir, and the consulted book or anthology, whose text was copied and freely re-worded. This conjecture joins a plain fact about error processes to citation fidelity: copying and adaptation degrade with quotation length, but over-learned memory does not at these scales, so fidelity as a function of length should separate the two stores cleanly, with liturgical scripture staying verbatim at any length while patristic prose drifts more the longer the quote. Masters made the split unconsciously, quoting the choir from the ear and the fathers from the desk. If this holds, we gain an instrument for mapping exactly which texts lived inside thirteenth-century heads rather than on thirteenth-century shelves, an empirical geography of the trained medieval memory.

Prediction clause (verbatim)

Primary clause (verdict follows it): in the Aquinas citation corpus, collating quoted text against critical editions, the verbatim-match rate declines significantly with quotation length for patristic prose contacts but shows no significant length effect for Psalm and other liturgically over-learned scriptural contacts; the verdict follows the significance of the interaction term. Statistical test.

Kill-dataset (verbatim)

The Aquinas citation corpus quotation texts, collated against the Vulgate Psalter and patristic critical editions.

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 by claude-fable-5 in a single Write with no file reads, web access, or database queries; all context was supplied inline in the launching prompt.

Novelty / leakage triage

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

Carruthers's memoria scholarship establishes the two-store picture — over-learned verbatim liturgical memory (Psalter in choir) versus consulted and freely re-worded book sources — which is exactly the mechanism the length-interaction test operationalizes. The fidelity-by-length collation has not been run, but the qualitative separation is established.

M. Carruthers, The Book of Memory: A Study of Memory in Medieval Culture (2nd ed.)

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.