← All conjectures · Jewish book cultures

The twin reconstructed from its own echoes

Iron Age & Antiquity Dated c. 100 CE – 300 CE Levant & Near East

Status: Already answered

The verdict’s fine print — quoted from the resolution record: “The verdict follows the registered primary clause exactly; the robustness range never approaches the 0.67 threshold from any handling variant.” Read the full caveats ↓

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 conjecture imagined by a language model — drawn from its trained weights and held to falsifiability, novelty, and plausibility, not to any one method: it may join two or more fields, or none. 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)

The tannaitic halakhic midrashim on Exodus come in two recensions from rival schools: the Mekhilta de-Rabbi Ishmael, which survives intact and independently, and the Mekhilta de-Rabbi Shimon bar Yohai (Mekhilta de-Rashbi), which was lost as a standalone book and reassembled in the modern era - by D. Z. Hoffmann and then in the Epstein-Melamed critical edition (1955) - out of quotations embedded in later compilations, above all the thirteenth-to-fourteenth-century Yemenite Midrash ha-Gadol of David b. Amram al-Adani. The mechanism of loss dictates the shape of the recovery: a book reconstituted from a reservoir cannot be more complete than the reservoir chose to be, so the recovered Mekhilta de-Rashbi should not mirror the verse-coverage of its intact twin but should collapse toward wherever Midrash ha-Gadol and the other citing compilations happened to draw. Because those reservoirs quarried preferentially for halakhic and quotable matter, the reconstruction should thin out exactly where the intact Mekhilta is richest in non-legal narrative. Prediction restated: the reconstructed Mekhilta de-Rashbi covers substantially fewer distinct Exodus verses than the intact Mekhilta de-Rabbi Ishmael, with the shortfall concentrated in narrative pericopes - the fingerprint of a text known only through the filter that saved it, not through itself.

Prediction clause (verbatim)

Prediction: mapping each midrash's comments to their Exodus verse-anchors in Sefaria, the count of distinct Exodus verses treated in the reconstructed Mekhilta de-Rabbi Shimon bar Yochai will be under two-thirds (a ratio < 0.67) of the distinct verses treated in the intact Mekhilta de-Rabbi Yishmael, and the coverage gap will fall disproportionately in narrative (non-halakhic) sections rather than legal ones - primary clause: the < 0.67 reconstructed-to-intact distinct-verse-coverage ratio; the verdict follows it. The test voids for coverage if Sefaria's Mekhilta de-Rashbi is itself only partially digitized, in which case the Epstein-Melamed printed edition supplies the verse map.

Kill-dataset (verbatim)

Kill (partly not yet built): Sefaria's Mekhilta de-Rabbi Yishmael and Mekhilta de-Rabbi Shimon bar Yochai, aligning each comment to its Exodus verse-anchor and comparing distinct-verse coverage; where Sefaria's Rashbi text is incomplete, the Epstein-Melamed critical edition (1955) supplies the verse map and the Midrash ha-Gadol supplies the reservoir profile.

Browse the registry: Digitized texts →

How this item is catalogued

Catalogue tags are curatorial metadata that make the conjecture findable. They are not findings, not a novelty verdict, and naming a source family is not a promise that the data exist or are accessible. Each dimension below says whether it was assigned, reviewed with none applying, uncertain, or not yet reviewed, and who recorded it.

Place & era tags — Assigned: Levant & Near East; Iron Age & Antiquity Tagged by claude-fable-5
Dating — Phenomenon: Dated c. 100 CE – 300 CE Tannaitic halakhic midrash Mekhilta de-Rabbi Shimon bar Yohai, composed in the tannaitic redaction era (standardly 2nd-3rd c. CE); Midrash ha-Gadol (13th-14th c.) is the reconstruction instrument. Tagged by claude-sonnet-5 + claude-fable-5 (gate)
Subject — Assigned: Philology, editions & stemmatics a lost legal-midrash reconstructed from quotations embedded in a later compilation Tagged by claude-sonnet-5
Finer subject topic — Assigned: Midrash The object is two recensions of a tannaitic halakhic midrash on Exodus, squarely the midrash page. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Claim level — Assigned: Record claim The prediction is the verse coverage of a midrash reconstructed from quotations against its intact twin, a property of the reservoir that saved it. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Method — Assigned: Taphonomy and survivorship bias The intact twin is the production baseline and the reconstruction is the filtered survivor, so the registered coverage shortfall is read as the named reservoir's selection bias rather than as the lost book's own shape. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Evidence source — Assigned: Digitized texts; Not yet built; Published editions The row's own kill line says partly not yet built: verse-anchor alignment runs on Sefaria, with the Epstein-Melamed edition supplying the verse map where the digitization is incomplete. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Research stage — Test resolved
Dataset family of the resolution — Assigned: external the resolution loaded a named external corpus or database, not the project's manuscript census Tagged by claude-fable-5-1 (driver, lane 1da5; requested max effort, runtime not independently attested)

On Inferpedia

This conjecture is linked to the following pages on Inferpedia, an encyclopedia of the missing — working atlas pages, some still early scaffolding.

Provenance

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

Fresh blind generation by claude-fable-5, 2026-07-17, Jewish text-culture wave instrument-anchored on the open Sefaria corpus and its cross-reference link data, with standard critical editions as controls: every kill names a real corpus and a countable operation (coverage maps, citation-formula counts, link-orphan shares, digest-fraction, recension divergence, citation-decay), thresholds far from 1 with explicit coverage guards distinguishing what Sefaria holds from what existed. Ground is CITATION-GEOMETRIC and disjoint from the 2026-07-10 w19 Jewish wave, which was material-culture-of-loss (colophons, parchment, genizah, masora, binding fragments): no material-culture re-posing here. Two candidates were dropped after a grep of all fresh packets: a Tosafot-density-by-tractate survival item (pre-empted by w02-philosophy #18, which already correlates per-folio Tosafot density with manuscript survival), and a piyyut liturgy-vs-anthology survival item (pre-empted by w01-literature #28 'Liturgy out-survives fame', keyed to Davidson's Thesaurus). The Mishneh Torah item here tests EXTERNAL source-attribution erasure vs the Tur (loss of source-geometry), deliberately distinct from minds-w02 #25, which tests the code's INTERNAL recall/promise cross-references. Two items are marked Kill (not yet built) where the deciding corpus (Kohut's Arukh apparatus; Lewin's parallel Iggeret recensions) is not yet digitized.

Novelty / leakage triage

already answered in the literature

The printed verse-maps already answer the primary clause - against it. The Epstein-Melamed edition (which rests substantially on Geniza leaves of the work itself, not only on Midrash ha-Gadol quotations) and the standard reference's coverage lists give the reconstructed Mekhilta de-Rashbi a span reaching Exodus 3 and 6 and running 12:1-24:10 plus the sabbath pericopes, while the intact Mekhilta de-R. Ishmael begins only at Ex 12:1 and ends at 23:19 (plus 31:12-17, 35:1-3); the reconstructed twin's distinct-verse extent is comparable to or greater than the intact one's. A competent reader derives from these printed maps, without touching Sefaria, that the < 0.67 reconstructed-to-intact coverage ratio fails - guaranteed list-arithmetic on published editions, in the exemplars' sense, with the leak running against the conjecture's expected direction.

Sources cited by the triage

J. N. Epstein & E. Z. Melamed (eds.), Mekhilta d'Rabbi Sim'on b. Jochai (Jerusalem, 1955), edition and introduction (coverage and Geniza basis)
H. L. Strack & G. Stemberger, Introduction to the Talmud and Midrash (2nd ed., 1996), sections on the two Mekhiltot with their verse-coverage lists
W. D. Nelson, Mekhilta de-Rabbi Shimon bar Yohai (Philadelphia: JPS, 2006), introduction
D. Z. Hoffmann, Mechilta de-Rabbi Simon b. Jochai (Frankfurt a.M., 1905)

Predictions

Killed registered 2026-07-17 calibration prediction (parent triage: leaked/adjacent)

Resolution: Killed

Caveats: The verdict follows the registered primary clause exactly; the robustness range never approaches the 0.67 threshold from any handling variant. Two scope caveats stand: (1) Sefaria's MdRSbY text is the Epstein-Melamed reconstruction, whose own sources include Genizah leaves of the work itself, so 'coverage' here measures the modern reconstructed corpus, exactly as the conjecture's kill specified; (2) verse-level coverage is not depth - MRI may still carry more text per covered verse (word-count depth was not the registered clause). Verdict authored by the shepherd (Fable) from the firewalled compute; independent VERIFY per the no-self-grading rule remains open for a separate session.

Registered against Sefaria's open corpus (Sefaria-Export / API snapshot to be frozen at compute time). Claim under test (primary clause of the conjecture): the distinct-Exodus-verse coverage of the reconstructed Mekhilta de-Rabbi Shimon bar Yochai is LESS THAN 0.67 times the distinct-Exodus-verse coverage of the intact Mekhilta de-Rabbi Yishmael.

Resolution criteria — the registered fine print

Resolution criteria: DATA: the two Mekhilta indices in Sefaria's Midrash Halakhah category (Mekhilta d'Rabbi Yishmael; Mekhilta DeRabbi Shimon Bar Yochai), fetched from the Sefaria API or the Sefaria-Export GitHub mirror; freeze snapshot date and sha256 of the fetched JSON. POPULATION: distinct Exodus verses (chapter:verse) at which a text carries non-empty structural content (PRIMARY: the work's own chapter/verse anchoring as structured in Sefaria; SECONDARY variant: verses reachable via Sefaria's link data). R = distinct_verses(MdRSbY) / distinct_verses(MRI). CLAUSE PRECEDENCE: (1) INCONCLUSIVE_BY_DESIGN if either index is absent from Sefaria or its structural anchoring cannot be extracted; (2) SUPPORTED if R < 0.67; (3) REFUTED if R >= 0.67. Compute is epistemically firewalled: the computing agent receives only the counting operation, never the conjecture text, the threshold, or the expected direction.

Known-priors disclosure — what the registrant already knew

Known priors disclosure: Shepherd triage (2026-07-17, this corpus) judged this conjecture leaked-toward-refutation: the printed verse-span maps (Epstein-Melamed 1955 edition of the Mekhilta de-Rabbi Shimon bar Yochai; Strack-Stemberger's introductions) indicate the reconstructed text's Exodus coverage EXCEEDS 0.67x the intact Mekhilta's, i.e. the registrant expects REFUTED. The compute agent is firewalled from this disclosure and from the threshold; the computation decides.

Method and dataset — how it was measured

Register-before-compute (registration packet conjecture_prediction_sefaria_mekhilta_20260717.json, imported before compute began; known prior DISCLOSED there: triage anticipated refutation from the printed verse-span maps). Compute by a firewalled Sonnet agent given only the counting operation - no claim text, no threshold, no direction. Hebrew field used for non-emptiness (MdRSbY's English is ~7% complete and would have been a translation artifact); links-API secondary abandoned on 504s, structural count primary as registered. computed_at postdates registered_at (rule 6).

Dataset: Sefaria open corpus, structural Hebrew-text coverage: distinct Exodus verses (chapter:verse) at which each Mekhilta carries non-empty content. Exact Sefaria index titles: 'Mekhilta DeRabbi Yishmael' (via its alt_structs Chapter addressing, 9 tractate fetches) and 'Mekhilta DeRabbi Shimon Ben Yochai' (native chapter/verse schema, default node; the Additions appendix excluded from primary). 16 raw payloads sha256-logged in the compute artifact.

computed 2026-07-17

Weigh in

No community feedback yet.

New here? Create an account first

Create an account or sign in and your feedback is tied to you — you can track it, get replies, and claim this conjecture so others know you’re working on it. Prefer not to? Just leave your take below as a guest — only the name you type is shown.

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.